AI Free Agency behavior weaknesses

[Electricbassguy]

As my giants went into debt (i started in the 1997 year) I was forced to dump my starting third baseman bill mueller, who was demanding a 6 million dollar salary in 1998

He was picked up by the Red Sox for that amount and has been hitting .350 at AAA.

The Red Sox are paying him 6 million a year to be a star minor leaguer, even though he is an above-average starting third baseman.

In reality, what team would do this, and what player would accept a contract to sit in the minor leagues when they are of starter quality?

[GM24]

I agree that this happens all too often... couldnt teams look in to the future and see where a player fits in with a team, and then put a number figure on it... and update this everytime that team signs a FA?

[JrCrowley33]

I also agree, once a player reaches a certain age, he should not be allowed to be signed into AAA. I dont know how many times ive found frank thomas, juan gonzalez and other aging sluggers in the minor leagues. And its not like they spend a season down there, sometimes you find that they havent played a game in the majors in 4 or so seasons, they should be forced to retire.

[Clay Dreslough]

-----Original Message-----
To: Clay@sportsmogul.com
Subject: Free Agents to be

Why when a player on my team, that is in his final year of his contract, demand 12.0 mill from me so I trade him and he demands 6.0 mill from his new team?

---

There was a bug that caused players to ask for way too much money if you had a lot of cash in the bank. I have addressed this with version 7.14 (which I'll be posting soon).

Clay

[HoustonGM]

It's really making the game unplayable for me because its just so unrealistic. Boston signed 5 starters, and now has Curt Schilling and Joel Pineiro in their bullpen, with 5 85+ starters in their rotation. San Francisco signed both Mark Teixiera and Justin Morneau, when they already had Lance Neikro. Boston also signed 1B Dmitri Young (now in AA, making 3.9 a year) and 1B Phil Nevin (also in AA making 4.7 a year), when they have Heo Seop Choi, David Ortiz, and Kevin Youkilis (82 rated). (And when trading, they still want a starting DH and a starting 1B!!!!!!!!!)

[irish07067]

I totally agree and I have shelved the game until this issue is resolved. It's not accurate and doesn't play like a "sim" should.

[HoustonGM]

I've resorted to simming one day at a team, and turning on Commish mode and releasing the players if they don't make any sense. If I sim again, and in a few days, the same team signs the player, I release them, go through the teams for a team that it owuld make sense to sign the player, and sign the player to that team.

It's annoying

[Gaedel]

OK, here comes the old problem again. I began play in 1951 with the St Louis Browns. I am now the Baltoimore Orioles and it is March 1, 1961.

In the fall of 1958, the AI General manager of the New York Yankees went nucking futs. During the off-season he began to sign and release on the same day a number of 25-28 year old pitchers with like 65/70 ratings. He has signed and released the same guys so much their history stacks overflowed and blanked out (see a previous thread of mine on history stack blanking out).

Here is a list of his activity for the first eighteen days of January 1961:

Signed and released on same day:

Jan 1: Ed Donnelly
Jan 2: Ramon Monzart, Hy Cohen, Donny Saunders
Jan 3: Kevin Walsh
Jan 4: Walsh
Jan 5: Walsh, Donnelly, Seth Morehead
Jan 6: Walsh, Pete Hillon
Jan 7: Monzart, Don Williams
Jan 8: Charlie Rabe, Walsh, Williams
Jan 9: He rested for a day
Jan 10: Pat Borlen, Joseph Boetcher
Jan 11: Tom Qualters, Morzant
Jan 12: Saunders, Qualters, Jay Duck
Jan 13: Day Off
Jan 14: Pete Dillon (been signing this guy since 1958)
Jan 15: Took day off)
Jan 17: Derek Hardcastle
Jan 18: Hardcastle, Ken Rowe, Cam Homan

He doesn't do this during the season, but he doesit constantly in the off-season. Only the Yankees and only for the past two years (Oct 58-Mar 61).

Here is an example:

Derek Hardcastle, pitcher, 22 years old, 68 overall, 73 peak, 78 health

Signed in free agent draft by NYY 01-17-56.

Played in rookie leagues in 56, 57, 58, and 59.

Began 1960 in rookie league and moved to A.

Coaching a new pitch 03-22-57.

Released by NYY 10-17-60.

Since that release, NYY has signed Hardcastle to two year contracts at six figures per year, then released him on the same day on the following dates:

1960: 10/26, 10/29, 11/14, 11/17, 11/23, 11/30, 12/02, 12/09, 12/10, 12/20, 12/30.

1961: 01/17, 01/18, 01/22, 01/23, 01/24, 02/09, 02/18.

It is now Mar 1, 1961.

[HoustonGM]

The main cause of this is injuries.

This guy attached is currently approaching one month on one team's disabled list - a season high for him.

Players who are injured are consistenyl dropped unless they have star ratings (85+). Players who are injured for the whole year are dropped even more often.

The problem I think is that when signing, a team does not look at a player's injury status to determine the usefulness of the player to the team. Then, once that player's on the team, it looks at the injury, and sees that an injure dplayer sint useful, and releasesh im.

However, injured players are never released like this in real life. The only time they are released is if they actually deserve to be released. The Devil Rays signed Al Reyes, who they knew would miss the whole 2006 season, and they're going to keep him on their roster all year - in real life. Players who miss a season in real life don't change teams like people change underwear. In mogul they are.


It needs to be fixed.

[Gaedel]

HGM: With due respect, these guys aren't injured. These multiple signings and releases by one team, all pitchers, all in their twenties, and the signing and releasing all occur on the same day (multiple times).

These are all occuring in the off season when no one is getting injured.

Doesn't the attached look a trifle insane??
4300
In the example of Derek Hardcastle noted above, the guy has never been injured.

[HoustonGM]

I haven't played tghe 2006 offseason yet, but when I do, we can figure out whether or not this is a problem with historical sims, or all sims.

There are obviously mutltiple problems with the free agent signing/releasing AI. I was pointing out one major one I've noticed, taht needs to be fixed alnog with your examples.

[Gaedel]

I was wondering if it was just that play and something happened, so i loaded up another mogul run I did with the Red Sox 1951-1967. Same thing there. The Yankees suddenly went on a tear in the offseason signing and releasing marginal (uninjured) pitchers.. The Doodgers and Giants had a few episodes of doing it, but the yankees kept right at it after they started. From free agency in October right up to opening day, they sit there signing and releasing.

[dolfanar]

i have seen BOTH of these issues.

Problem 1: Injuries. HGM is absolutely correct, the game does do this. I saw Cleveland release the eventual leage batting champion (.326) in august because he had a 2 month injury. Ridiculous! For the time being, the AI should simply NOT RELEASE INJURED PLAYERS. Period. Yes, there are exceptions, but for the most part injured players *don't* get released. For one thing, the MLBPA would be all over a team that operated the way BBM's GMs do. Second of all it leads to teams discarding perfectly acceptable starters and bench men, which weakens them forcing them to play wildly inapropriate players full time (Like Rookie ball call-ups!).

Problem 2: Teams with money, and holes in their lineup. My guess is that teams in BBM don't really have a minor league "strategy". They just identify a need then go into a kind of algorythmic bidding war, which the rich teams win more often then not.

Example: The AI identifies that the Yankees are a bit weak on pitching, so it puts them in aht ebidding war for ALL of the current free agent pitchers. On Day 1 the Yankees win the bidding on 3 of these guy's, all of them middle of the road lifelon AAA guy's. The AI signs them, and puts them on the roster and does an auto-sort. Oops, it realizes that that it has too many of these guy's clogging up its system, so it releases 2 of them. On day 2, still weak on pitching the AI rolls the dice again, signing 2 pitchers (1 of which was the same as on day 1). Again it realizes it doesn't need THESE particular pitchers, and releases them again. Wash, rinse, repeat.

The AI doesn't really go after SPECIFIC things. For example: If I need a long reliever, I'll try to look to my farm system *first*, if there is no one suitable there, then I'll look at trading for someone. If there is no one available that I'm willing to pay the price for, I'll look at free agents... but ONLY the pitchers who I think would be suitable for the slot, and if no one is available, I'll stick with what I have, and try to make due. The AI, just doesn't do this for ANY of its signings. It isn't selective at all (which is why we end up with rich teams stockpilling middle of the road Major leaguers in AAA and AA).

[Gaedel]

George Steinbrenner must have really had a "love-hate" relationship with this player:

[Mazam33]

*shrug* Well, at least it's been an improvement over last/previous years, imo.

Setting revenue at +40 and contract demands at -20, i just played out the entire 2006 season (checking the FA list every day) and didn't see a single 80+ rated player get released. In fact, there were only 5 players rated above 75 that got released, and all were marginal 30 year old+ bench players.

Since i strongly suspect that the AI is not actually having to finacially pay for these small releases, it personally dosn't bother me or effect play all that much, again imo.

But yeah, it should deffinately be addressed.

[ohms_law]

Since i strongly suspect that the AI is not actually having to finacially pay for these small releases, it personally dosn't bother me or effect play all that much, again imo.
That's the second time that i've seen this said reacently. However, AI controlled teams do go into the read and the couple of times that i've looked at them or gone and switched teams to one of them, there stuck with $0 spemding on farm, scouting, and medical as well as the fact that they must release every free agent at the end of the season and cannot resign any of them. I just don't see it.

[Gaedel]

The Yanklees seem to be the major offender on the sign and release same day kick. The Yankees have always had money in whatever sim I played. They usually have the highest payroll, but still are always rolling in cash. Eating a couple of hundred thousand per sign and release doesn't really hurt them.

[HoustonGM]

The smaller market teams don't usually participate in this insanity. It has to have a small root in the amount of money available to a team.

[Gaedel]

The smaller market teams don't usually participate in this insanity. It has to have a small root in the amount of money available to a team.

Maybe it is the yankees version of "build stadium" to prevent excessive cash buildup and loss of fan loyalty. Whenever the yankees get too rich, they just sign and release blowing off six figures with each transaction (assuming the AI teams have to eat contracts).

[adamsin]

That oddly enough sounds like it makes sense, you never see a Yankess team with 2-300 mill in the bank???

[Gaedel]

OK, in my 1951 historical sim, the AI Yankees began to sign and release marginal pitchers on the same day right after free agent period began in 1957. They continued this in the off-season right up until beginning of season 1962. I am now at Jan 17, 1963 and they quit doing it after the end of season 1962. They just switched off doing it. Their cash is now aboyt $37 million.

[ohms_law]

I have a theory that the AI is burning cash on purpose. how much cash did they have before the season?

[Gaedel]

I have a theory that the AI is burning cash on purpose. how much cash did they have before the season?


I don't know. In 1961 they lost big money (25.4 megabucks), but still did it in the 1961-1962 winter. In 1962 they made $3.5M and quit doing it in the winter of 1962-1963.

[ohms_law]

ah, I see (reading the first page) that you already mentioned what I'm talking about. Rather than signing and releasing players, the AI ought to actually spend the cash somehow. Raising expenses to insane amounts would make alot more sense than signing and releasing players for no reason.

[Gaedel]

I have a theory that the AI is burning cash on purpose. how much cash did they have before the season?

Okay, I just simmed from Amatuer Draft (when uit is supposed to take place) on Jan 17, 1963 and Spring Training on Mar 1, 1963. The Yankees began an orgy of signing and releasing pitchers from 1-17 to 3-1 and their cash stayed the same at $37.1 million. This means that the Yankees are not eating those six figure contracts when they sign and release on same day. That puts a hole in the "eating cash" theory that you and I thought might be happening.

If the AI changed to where this was costing cash, maybe the AI would quit doing all of this "sign and release". There might be a bug and only the player team eats a contract on release.

[ohms_law]

yea, that's possible.
I think at this point we'd need input from Clay before making any other assumptions. He's the only one that would really know, you know?

[Gaedel]

yea, that's possible.
I think at this point we'd need input from Clay before making any other assumptions. He's the only one that would really know, you know?


I fully agree with that. I am just trying to provide Clay with as much evidence as possible of what it is that is going wrong so that he can figure out what glitch in the program is causing this. I am not complaining that it is wrong, it is just something i would like to see fixed and want to provide as much info as possible.

[jonnymo]

Problem 1: Injuries. HGM is absolutely correct, the game does do this. I saw Cleveland release the eventual leage batting champion (.326) in august because he had a 2 month injury. Ridiculous! For the time being, the AI should simply NOT RELEASE INJURED PLAYERS. Period. Yes, there are exceptions, but for the most part injured players *don't* get released. For one thing, the MLBPA would be all over a team that operated the way BBM's GMs do. Second of all it leads to teams discarding perfectly acceptable starters and bench men, which weakens them forcing them to play wildly inapropriate players full time (Like Rookie ball call-ups!).
Excellent suggestion, Dolf. We've been trying to figure out how to make the AI think like real-life GMs. I have faith that someday Clay & Ian will get that down, but until then, this is the simplest, and best, fix.

[ohms_law]

i've done three 10 year long sims that started in '06, using 9.41a. Looks to me like Clay squashed this bug...

[jonnymo]

i've done three 10 year long sims that started in '06, using 9.41a. Looks to me like Clay squashed this bug...
9.42A?

[ohms_law]

uh, yea. sorry.

[boomboom]

Below is a typical trade that the AI purposes me during the season....I usually turn it off, but my desktop is giving me trouble, so I am on my 2nd computer, and I didnt turn it off...well anyway, I dont think that trading two 80 rated players for 82/82, and one in its 70s already at its peak is a bug...other times, I get a trade for one 85 reliever for 7 players that are rated at 75.....sure, that 85 reliever would really help my team, but not at the expense of losing 7 elements to my team, bench players, my depth if a player gets hurt...ect...

This is with trade offers turned on.

[ohms_law]

The thing about the computer GM's offering you trades is that what is originally offered is the best deal possible for the team offering the trade. If you blindly accept any trades offered to you without modification, the computer GM's will quickly pull your team completely apart.

What i've found is that when the computer GM's offer you a trade, it's simply showing you what that GM is interested in. In the example that you've shown above it looks to me like the Card's GM probably believes (has calculated) that:

You need starting pitching.
It may need to cut salary for the Cardinals.
It's probably looking for RP and/or 3B.

Most of the time, the computer GM's will offer one player of equitable value for one or more of your players. Often, they will also throw in a 70's rated also-been or never-been to "even up the trade". If you talk to the GM (go to the trade dialoge), you can often take one or more of these players off of their offer and substitute $1-5 million instead.

The real trick though is in interpreting what they want from you and/or what they need or need to get rid of. I don't see the AI trade offers as being useless; they're a window into how the AI thinks. You'll hardly ever want to accept a deal at face value, but it does give you a starting point to work from if you do want to make a trade.

[HoustonGM]

That trade is one of the fairest I've ever seen. lol. I was playing the Reds once, and yeah, Ken Griffey's rated 85 or somtehing, but some team offered me 2 70 rated nobodies for Ken Griffey and my top set up man.

[ohms_law]

oohh, ohhh; brainstorm!

here's my idea. It'd be great if, in these trade offer dialoges (and in the trade dialoge itself maybe?) if a text box could be added with the computer GM's "thoughts about the trade". In the example above, it could say something like "I need a releiver" or something like that. Then, in the actual trade dialoge it could evaluate trades as you make changes to the players and money offered and display both the new sayings (basically pulled form what you get when you currently press hint after a failed trade) and the standard saying that you get when you do fail a trade (extended to include saying for when the computer wants the trade, of course).

It certainly won't solve the "problem" that you posted about boomboom, but I think that it would increase the feeling of being in the game. That and you would have a better sense of what you're working with (or rather, against?).

[HoustonGM]

ohms_law, in the trade offer dialog, if you submit a trade and they decline, you can click hint, and it'll tell you what the team is in the market for (although, sometimes they make no sense, which is in my bug compilation, so Clay has to adjust that AI there. Like, the 2006 Red Sox were in the market for a left fielder, a catcher, a first basemen, a third basemen, and a designated hitter :rolleyes:)

[ohms_law]

I know. I was just proposing moving that from where it is now (currently somewhat hidden) to being almost constantly displayed. I imagine that some sort of highlighting of the list could even be added, to highlight what wants the AI is trying to fulfill with the current trade offer.

[HoustonGM]

I know. I was just proposing moving that from where it is now (currently somewhat hidden) to being almost constantly displayed. I imagine that some sort of highlighting of the list could even be added, to highlight what wants the AI is trying to fulfill with the current trade offer.
True.

[T Rip]

I also try to see the offers as a starting place. However, some are hilarious at first. One offer was some "Ed Shlabotnik" kind of guy for rookies Tony Oliva, Lou Brock and Mike Cuellar. "No thanks."

[Gaedel]

I have 9.43 installed and the Yankees (and to a lesser degree the Mets and Cards) are spending the latter part of november and all of December and january signing and releasing (same day or a few days later) whole flocks of free agent marginal pitchers with current and peak values in the 60s.

[HoustonGM]

I've never seen this while simming into the future. However, from screenshots people have posted, I've noticed that this is a major concern in historical leagues especially.

[Gaedel]

I've never seen this while simming into the future. However, from screenshots people have posted, I've noticed that this is a major concern in historical leagues especially.


I usually do historical sims. There appears to be a subroutine in the AI where the teams with money review their rosters decide to add to their roster of pitchers. They then sign some marginal minor league hangers on. Then there is another review of the roster and they get rid of "schlubs" who have no potential. There must be some threshold for signing a marginal player which is set lower than the threshold for keeping a player. This seems to be pretty much a pitcher problem as I have not seen it with position players. This should be a relatively easy tweak.

[ohms_law]

hopefully it is a relatively easy fix. Clay has been working on similar issues in the last several patches as well, so I'm fairly confident that it's getting it's fair share of attention.

these things are as likely to happen during a 2006-2007 sim as during any other time. The thing is, is that they are much more noticeable if you play through several seasons (10 or more), since there is much greater opportunity for the AI to exhibit this behavior.

[HoustonGM]

these things are as likely to happen during a 2006-2007 sim as during any other time. The thing is, is that they are much more noticeable if you play through several seasons (10 or more), since there is much greater opportunity for the AI to exhibit this behavior.
In modern sims, I've noticed sign/release things, but nothing anywhere near as severe as some of the posts I've seen. I've never seen a guy signed/released on the same day for the entire offseason. Teams often sign a bunch of guys, then a few weeks later, release that whole bunch. While that still should be fixed (along with the topic of this thread), it's not as severe as I've heard about in historical leagues.

[ohms_law]

my only point though is that it really has nothing to do with what year the sim is in. The program just reaches some point based on some set of conditions where the decision making function ends up oscillating between "i gotta sign someone" and "i gotta release someone", for whatever reason.
Sometimes it seems to fall into a really severe problem with this. most of the time it's relitively minor though, and the team will only sign and release a player a handfull of times.

An over-abundance of what may seem to be minute and unaffiliated information about what else is going on in the league at the time this bahavior is exibited would probably be really helpfull to Clay and Ian, I'm sure. Doing so is probably the only chance that this problem will ever get solved as well.
...unfortunately, I can never catch the AI "red handed". I notice it after it's occured, but I have yet to see it while it's happening, which is why I haven't posted a report about it myself yet.

[Gaedel]

An over-abundance of what may seem to be minute and unaffiliated information about what else is going on in the league at the time this bahavior is exibited would probably be really helpfull to Clay and Ian, I'm sure. Doing so is probably the only chance that this problem will ever get solved as well.
...unfortunately, I can never catch the AI "red handed". I notice it after it's occured, but I have yet to see it while it's happening, which is why I haven't posted a report about it myself yet.

While it does occur randomly through the year where a player will be signed anbd released after a month or so, it becomes endemic in the off-season. It usually starts right after the free agent period. the teams pick off the "plum" free agents. Then some teams (mostly the Yankees) begin signing and releasing pitchers right on through November, December, January, and February. Once spring training begins, there is some signing of position players.

I haven't fiddled with any of the settings, but I get this:
4802
4803
4804
Some representative players:
4805
4806
4807
4808

Having gone through four years of Civil Engineering using a slide rule, I am far from being an expert computer programmer. I did however look at the logic of the issue.

1. It appears to be primarily a phenomena dealing with the financially better-off teams.

2. The Yankees are the prime offender and the Mets are second. They are the two most obscenely rich teams.

3. The richer teams are more active in the free agent market. The free agent market for pitchers is relatively shallow and the good ones go quickly.

4. Once you are into the elderly has beens and the youthful never-will-bes of the pitchers market, the ratings are rather low.

5. If a team has a certain amount of cash (and is below projected payroll), the off-season AI periodically says "sign free agents" which the AI then does, picking up the cream of the crapola with some sort of threshold (player > T1).

6. The off-season AI periodically reviews its roster looking for deadwood which it prunes. Deadwood is determined by another threshold (player < T2).

7. If T1 < T2, AI will have the current sign-release-sign-release condition.

8. If T1 > T2, AI will retain the players it signs (unless their rating drops).

It doesn't appear to be a major programming tweak.

[ohms_law]

7. If T1 < T2, AI will have the current sign-release-sign-release condition.

8. If T1 > T2, AI will retain the players it signs (unless their rating drops).

you're right, it really isn't something that would be considered "major", in programming terms. This is a bug fix, not some sort of major rewrite.
That being said, knowing what "T1" and "T2" actually are, and then finding them in thousands of lines of code isn't exactly easy.

Anyway, good post. My only point when I say all this stuff is that it takes time. there isn't (usually) going to be a patch overnight. the problem has been reported and some details have been given so at this point, unless Clay or Ian comes and asks for some more information we just need to be patient is all.

[Gaedel]

Anyway, good post. My only point when I say all this stuff is that it takes time. there isn't (usually) going to be a patch overnight. the problem has been reported and some details have been given so at this point, unless Clay or Ian comes and asks for some more information we just need to be patient is all.

OL. I really have no sense of urgency in having bugs fixed (unless they cause the game to lock up or crash). I do enjoy finding glitches, reporting on them, and giving suggestions. I could care less whether my ideas and findings appear in the next fix, in BBM2008, or in BBM2025.

I have gotten far more in entertainment value out of my BBM2004 and BBM2007 games than I spent on them. Clay and Ian owe me nothing and I hope what I say here is constructive criticism and will assist them in achieving BBM nirvana. It is their business and how they prioritize their effort is certainly up to them and I would not presume to order them around.

[ohms_law]

yea, me too.
:)

honestly, that was directed less at you than, well you know, other people...

[RAW]

Yes, many teams, especially the Yankees, sign players one day and release them the next day or so later. WHY? This occurs throughout the season and really makes no sense. I also have a problem with all other teams being able to sign young free agents for $13,000, $18,000, etc, while I can only sign them for $50,000 minimum. Someone suggested that if I sign them for $50m and assign them to a minor league team their salary goes down to that rate..not so, it stays at the $50m even if they are sent to rookie league. I've decided to just not sign players to complete minor league rosters letting my farm teams go without enough players even though I would like to have full teams, or nearly full, at all levels.

[ohms_law]

not so, it stays at the $50m even if they are sent to rookie league.
That's what it says, but as far as I can tell, it's not what you're actually charged. I once signed 20 rookie "free agents" at $50k/year and stuck them all in rookie ball. My payroll didn't appear to change, which tells me that the total charge for those players was < $1 million, which 20 x $50k ought to add up to.

[RAW]

Thanks. I go back and double check. Hope you're correct as I am well over payroll budget. If I have negative cash I don't think I can sign anyone anyway. Got myself in this fix by trying to win my division and world series. If that happens I'll probably have a fire sale and then play for the future. I'm currently managing the Phillies. Started in 1991 and am now in 1997 with a great lineup and staff, but suffering financially. That's the great challenge in this game, isn't it?

[Gaedel]

Yes, many teams, especially the Yankees, sign players one day and release them the next day or so later. WHY? This occurs throughout the season and really makes no sense. I also have a problem with all other teams being able to sign young free agents for $13,000, $18,000, etc, while I can only sign them for $50,000 minimum. Someone suggested that if I sign them for $50m and assign them to a minor league team their salary goes down to that rate..not so, it stays at the $50m even if they are sent to rookie league. I've decided to just not sign players to complete minor league rosters letting my farm teams go without enough players even though I would like to have full teams, or nearly full, at all levels.

You don't need to sign schlubs just to fill up your rosters. The AI creates "dummy players" to fill the holes. You just need to do the following:

1. Have enough players at each position in the system too overcome a bad string of injuries.

2. Be developing enough prospects to fill in for forecasted losses.

3. Make sure you don't have MORE than one player at each position in AAA, AA, and A (otherwise they won't get playing time).

[RAW]

Thanks, Gaedel. I'll start dumping players at the end of the season. Maybe make some two for one trades, etc.

[ohms_law]

looks like this is back (or still around), although not quite as bad as I remember it being when this was first posted.

If I could offer a suggestion, the players that I see this occuring to are primarily the marginal, journeyman type players. guys with ratings in the mid 70's, mostly. My suggestion would be to adjust the threshold that causes teams to sign a player, so that most of these guys don't get signed as free agent's unless the team really needs them to fill a position.

[jonnymo]

It's the ones that I signed myself that the AI won't let me trade.
I feel your pain, Beerchaser, but this isn't a bug. Take a look at Clay's description of the trade engine, Tradezilla (http://www.baseballmogul.com/baseball2k7/tradezilla.htm). In the second-to-last paragraph on the first page, he talks about AI GMs determining player value. While $36M over 5 years seems like a reasonable offer for the D-train, the other GMs think that you overspent. My proof is that he was still available at his asking price when you signed him. It doesn't matter whether you sign a staff ace or a career minor-leaguer; all 29 AI GMs would eventually have had the chance to sign him for less if you didn't.

You're not screwed, however. If he gets off to a fast start to the season, his value will increase. Once one of the AI GMs decides that Willis is worth more than his contract, you'll get trade offers. In my experience, however, I wouldn't expect anything good until around the trade deadline, and maybe not until next off-season.

You could bundle him w/ a highly-touted prospect, and then you would get offers, but they likely would not be any better than if you just offered the prospect. I know that if I was the AI GM, I would assume that you were trying to pull a Jim Bowden (http://news.enquirer.com/apps/pbcs.dll/article?AID=/20060813/COL09/608130419/1082/SPT) on me. :p

[HoustonGM]

Actually, I think the reason you can't trade him is becuse you just signed him. That's intentional, so that you can't sign players and immediately trade them for something better. After a certain amount of time, you should be able to trade him. (PErhaps by midseason of the next season...not sure on that though)

[beerchaser]

jonnymo,

Ah, but they didn't get Willis, did they? I did!!! I got him, on the 1st day of the FA season. And at a very reasonable price, relatively speaking compared to the market I think, for a 25 yr old 92/97 rated player. And now that I have him (and they don't), that does not by any stretch of the imagination make him completely worthless to every other team in baseball. He's gotta be worth something, in fact a pretty nice something I have to believe, to virtually every team in baseball and a lot more to some of them. I mean, the Padres gave up a very nice catching prospect in George Kottaras in August, just to rent a 43 yr old beerbelly softball pitcher in David Wells for a month or so.

And it's not just Willis, either. I signed Chris Capuano, Chad Tracy and a couple other highly rated players to decently priced contracts and I'm being offered zippo for all of them. Shoot, I even put all three of those guys on the Trading Block together just to see what would happen, and still got offered NOTHING. Sorry but I just don't buy that logic.

While on the other hand, as I mentioned earlier the AI signed Morgan Ensberg and Vicente Padilla to what I thought were overpriced contracts and assigned them to my expansion team. That in itself seems like a bug, since the AI did something to my team that I did not ask for or want. However, the AI DID let me trade those two guys. I didn't get a whole lot for them, which didn't surprise me as I didn't think they were worth that much, but at least I got something.

I'll expound more tomorrow, but right now it's 12:30AM ET and way past my bedtime.


Actually, I think the reason you can't trade him is becuse you just signed him. That's intentional, so that you can't sign players and immediately trade them for something better. After a certain amount of time, you should be able to trade him. (PErhaps by midseason of the next season...not sure on that though)

Houston, that thought occured to me too and I can buy (no pun intended) into that logic if true. However if that is the case then it needs to be spelled out somewhere in the documentation, or if it's there then someone please point me to it (thanks). And it still doesn't explain what happened with Ensberg and Padilla.

And I really do have to go to bed.

[HoustonGM]

Houston, that thought occured to me too and I can buy (no pun intended) into that logic if true. However if that is the case then it needs to be spelled out somewhere in the documentation, or if it's there then someone please point me to it (thanks).
I don't believe it's documented anywhere, but I know its there.


And it still doesn't explain what happened with Ensberg and Padilla.
Whenever you get a chance, can you explain that situation more thoroughly? I'm a bit confused with it.

[beerchaser]

Whenever you get a chance, can you explain that situation more thoroughly? I'm a bit confused with it.

OK, here you go. More details than you want probably.

I had originally started out playing a game as the 2006 LA Dodgers, using the default rosters that came with BM 2007. Immediately following the 2006 World Series, on the first day of free agency I think, I decided to start an expansion team in Raleigh, NC. I updated everything in League Editor under Commish Mode, then took ownership of the Raleigh team and relinquished control of the Dodgers back to the AI. I then saved the game under a different filename in case I want to go back to the Dodgers, who BTW won the 2006 WS. Beginner's luck, I guess.

Since I gave myself a pretty decent budget and was starting basically from scratch with no expansion draft to work with, and only the rookie-filled initial roster that BM assigned to my Raleigh team, I decided to start by signing some FA's. Hence the aforementioned Dontrelle, Capuano, Chad Tracy and a couple others, all of whom I signed on the first day of FA before another team could grab them. Note that Morgan Ensberg and Vicente Padilla were on the FA list but I was not interested in them at the time as I felt they were overpriced and did not enter into any negotiations with them. I don't think I even clicked on their scouting reports. However, I did notice that the Rumors box at the bottom of the screen showed both players as being in negotiations with Raleigh, but I ignored that since I figured rumors are just that, rumors.

After those early signings I decided to wait awhile before signing any more, knowing that some guys would get snapped up but the price would come down on the rest. I also saved my game at this point. I then simmed a week ahead to see what that did to the FA list. It was at this point that Ensberg and Padilla showed up as having been signed by Raleigh, on the first day of FA, for $5.0 and $5.8M respectively. I thought, WTF? and went back and restored to where I had last saved, and then simmed ahead one day instead of one week. Same thing happened. Well I didn't really want either one, certainly not at that price, so I said what the **** and went ahead and traded both guys for whatever I could get for them. Couldn't get much because of their price tag, but I got Scott Elbert from LAD (74/84 rating) for Ensberg and Humberto Sanchez (77/80) from DET. Elbert I like since he's a Dodger, even though I know he's a year or two away and the AI scouting said Sanchez would be an A+ fit for my team, so I went along with it.

Since I was able to trade those guys, I decided then to go ahead and trade Dontrelle also to try and fill out my roster with several decent players instead of one or two really good ones. That's where I ran into the problem of not being able to trade those guys. I had actually run into this issue once before in a different scenario, but I had assumed then that that was because the guys I was trying to trade actually had very little market value, like a Richard Hidalgo or a Brad Fullmer. Obviously that's not the case with Willis or Tracy. While I can understand (to some extent) the logic of placing a restriction on trading FA's you just signed, it sucks here because I'm working with an expansion team and have basically nothing of trade value other than some cash and those FA's. And to my knowledge there's no restriction on real-life GM's working such a deal.

I guess the fact that the AI signed Ensberg and Padilla and put them on my team without my consent is why it let me trade them, but it doesn't answer why it did that in the first place.

Sorry for the long post.

[beerchaser]

I've still got some work to do with the free agents, but there's enough of them left that I was able to sim ahead to the amateur draft. Not only did I get in but I went off first in every round. Sweet. Got an 83/96 lefty SP with the 1st pick. The only question now is do I let him season for a little while or put him straight on the major league roster. With an expansion team I think that's probably an easy call, might not be what's best for the kid but the team needs him.

[ohms_law]

You probably will be able to put him on your roster on opening day, but after spring training opens his ratings will change to something more realistic. Probably around 75/90 or so. In about 5 seasons, barring serious injury and him just flamming out, he will probably end up being a 90's overall guy though.

You're probably correct though, in that since your an expansion team you'll want him on your MLB roster regardless.

[jonnymo]

I'm not positive on the date, but I believe if you sign a free agent over the offseason, you cannot trade him until May 1st of the following season.
You make a good point, Houston, and it would seem that there might be a rule about not trading a newly-signed free agent until May 1. It seems to fit w/ the rule whereby a team cannot resign a FA who refuses arbitration until May 1. However, I don't think the rule exists. I've never heard of it, and I can't find any reference to it. Can you point me toward more information on this rule?

BC, as far as not being able to trade a newly-signed FA in BBM, as I said, you just need to make it worth the other team's while. In the example below, I signed Barry Larkin on the first day of free agency. Then, I turned around and put him on the trading block and added cash (in effect paying a portion of Larkin's salary). As an expansion team, I know that you wouldn't have much cash, but the principle is the same for adding prospects.
5575
5576
5577

Going back to the Willis example, let's assume that you didn't overpay for him. You could not have underpaid, since you gave him his asking price. Sure, you may have talked him down, but you still signed him on Day 1 for as much or more than you would have been able to sign him for on Day 2. Thus, best case scenario, you paid him market value. In this scenario, if you hadn't signed him, let's say Team X would have signed him instead, for the same amount. So, when you try to trade him, his value is exactly the $36M for which you signed him -- and no more. When you trade Willis, you are trading Willis (worth $36M) and Willis' contract (worth -$36M). In order to get a 60/85 player (worth, say, $3M) in return, you would need to add something -- either cash or players -- worth $3M to your side of the offer.

Make sense?

[ohms_law]

Good points Jonnymo. This issue regarding weather you can trade a player after signing him as a free agent has been dabated before. I'm never seen Clay actually say that you can't trade a newly signed FA, so i've always assumed that by signing them your simply reducing their perceived value to the AI. It looks like that's actually true, as well.

Good analysis here.

[HoustonGM]

I can't find anything on the RL life, although I've heard it somewhere before, more than once.

[SirKodiak]

Interesting, found this in the current CBA:

(5) Quota

(a) Clubs shall be limited in the number of Type A and B Play-
ers, as defined below, they may subsequently sign to contracts. The
number of signings permitted shall be related to the number of Play-
ers electing free agency under this Section B. If there are 14 or less
such Players, no Club may sign more than one Type A or B Player.
If there are from 15 to 38 such Players, no Club may sign more than
two Type A or B Players. If there are from 39 to 62 such Players, no
Club may sign more than three Type A or B Players. If there are
more than 62 such Players, the Club quotas shall be increased
accordingly. There shall be no restrictions on the number of
unranked Players which a Club may sign to contracts.

(b) Irrespective of the provisions of subparagraph (a) above, a
Club shall be eligible to sign at least as many Type A and B Players
as it may have lost through Players having become free agents under
this Section at the close of the season just concluded.

I think I found the rule about trading free agents that you have signed, it is in the CBA (right after the Quotas I quoted above) in :
ARTICLE XX--Reserve System
B. Free Agency

(6) Miscellaneous
(a) Any Club signing a contract after the expiration of the elec-
tion period with a Player under this Section B may not assign his
contract until after the next June 15. However, notwithstanding the
foregoing, such contract may be assigned for other Player contracts
and/or cash consideration of $50,000 or less prior to the next June
16 if the Player gives written consent to such transaction.

Looks like 'assign his contract' means 'trade him'

[ohms_law]

That's interesting, there's the $50,000 figure as well...

p.s.: I found the last thread where the no trade issue came up.
http://forum.sportsmogul.com/showthread.php?t=115860

note that that thread is in ejections now. what a hoot.
:)

[FRENCHREDSOX]

Still in BM & BMO(I assume) you can sign & trade which is illogical when looking at what is writen in the CBA .Thus it should be introduced,& I doubt that it requires a massive overhaul of the programming,either a newly signed FA be:

1)'untradeable' till the 15th June of the the current year (basically allowing you 15 days to trade him if wanted) or

2) cut AT full SALARY over the length of the contract unlike the infamous 1/2 price that exists


This would affect AI/player signings especially during the infamous first 24/48 hours where top talent just disappears at "lightning" sped

[beerchaser]

Still in BM & BMO(I assume) you can sign & trade which is illogical when looking at what is writen in the CBA .Thus it should be introduced,& I doubt that it requires a massive overhaul of the programming,either a newly signed FA be:

1)'untradeable' till the 15th June of the the current year (basically allowing you 15 days to trade him if wanted) or

I would have no problem with this rule as long as it was clearly stated in the Help documentation. Right now there is no such information about signing and trading FA's, you just simply can't do it. That sucks, especially with an expansion team with few other options available to it. And I think we all agree that an expansion draft is a needed feature for future versions of BM.

I believe I understand jonnymo's logic about the trade value of free agents and have read Clay's take on it, but I don't believe I agree with the notion that recently signed free agents have zero trade value. Especially if you were able to sign them for 300-400K less than what they were seeking, without having to add sweeteners like player options or any of that ****. But then, the idea of what constitutes market value for any given player is going to differ from one person to the next, and two people could debate this all day without either side changing their position one iota. Much like politics.

[HoustonGM]

Still in BM & BMO(I assume) you can sign & trade which is illogical when looking at what is writen in the CBA .Thus it should be introduced,& I doubt that it requires a massive overhaul of the programming,either a newly signed FA be:

1)'untradeable' till the 15th June of the the current year (basically allowing you 15 days to trade him if wanted) or
I don't think it's something that needs to be added. Some people might want to be allowed to do it, and since it's their game, they should be allowed to. For users who want to play with the realistic rules, they can just not trade recently signed free agents until June 15. It works out fine the way it is now - if you dont like the rule, dont use it. If you do, use it.

[ohms_law]

2) cut AT full SALARY over the length of the contract unlike the infamous 1/2 price that exists


This would affect AI/player signings especially during the infamous first 24/48 hours where top talent just disappears at "lightning" sped

The 1/2 salary thing only affects the user, not the AI.
the user get's first pick at free agent's, so i'm not sure what your point about talent disappearing at lightning speed is. If you want to sign a player before one of the AI gm's does, then do it.

HoustonGm is correct though, there's no reason to hobble user's game play just for "realism". It's a game not a sim, so being forced to abide by rules simply "because its more realistic" isn't any fun. Read beerchaser's reply, that's what i'm talking about.

[SirKodiak]

For many users, I would think the more realistic a baseball game is, the more they would enjoy it, and more likely they would be to buy it. On the other hand, there are users that just want playability. It seems to me it would be nice to have a Realism Options Page, where users have the option to include rules or not (certainly would be a boon to those that play during historical eras).

The "Realism Options Page" could include:
allow free agency

allow arbitration
do not allow Free Agents to be immediately traded
allow only X Free Agents above Y skill to be signed by one team during off-season


Player Development
a slider that goes from "random" to "consistant"

etc.

[HoustonGM]

Eh, I don't think that there needs to be a "do not allow Free Agents to be immediately traded." That's a rule that you can iether follow if you want, or not follow. It doesn't need to be enforced by the game. Say somebody wants to play one way, and then switch to the other, it's much simpler to just...change hwat you do, instead of having to go switch an option.

[SirKodiak]

I was thinking it would be useful for commissioners of offline leagues.

[ohms_law]

Actually, I would think that it wold be pretty easy to deal with in an offline league. If it's all human beings playing in a league, you can use any rule set that anyone could dream up...

Anyway, I'm hardly against realism. It's jsut that, for the most part, most people don't even know that these rules exist. heck, we had to look them up to even know about it. Additionally, think about what adding some of these sorts of rules are attempting to acomplish. what's the point? If the actual problem that we want addressed has to do with AI intelligence, then why not address that issue rather than ask for arcane rules that limit a user's ability to play the game the way they want.

I'll never be supportive of anything that limits a user's ability to do anything that they want.

[SirKodiak]

I tested it myself, signed 27yo Adam Dunn (93 Overall) as Free Agent (he was the best one), then lowered his contract to $350,000/5 yr. Tried both Trading Block and trading him directly to teams for older players rated 72 or lower: right after, after amateur draft, before spring training, after spring training, one week into the season, June 14, June 21. No one would take him straight up (I hit balance cash once and they wanted $21mil, which was all the cash I had). So it seems Clay at least knows somewhat about the rule, and he has decided to impliment it in some manner, and it is default and not an option. By asking that we have a way to turn it off, it would seem to me that I am asking to increase "a user's ability to do anything that they want." /shrug

[ohms_law]

well, humm... now i'm confused. That directly contradicts what jonnymo said earlier (more or less), so I guess i'll have to try it all out myself.

I wonder if the results are er... tainted? by editing the contract at all.

[FRENCHREDSOX]

It's a game not a sim, so being forced to abide by rules simply "because its more realistic" isn't any fun. Read beerchaser's reply, that's what i'm talking about.

Firstly,Ohm,my reply was before Beerchaser's.


Secondly, rules are rules - so if you abide by only half of them what is the point to abide by any? I know I am being extreme but why apply the trade deadline ,it may be more "fun" as you say to trade after,or why a draft order or .... see the point.

Thirdly,some USERS do not know the real MLB rules & thus can & will usually ,unknowingly,"cheat" - heck do you know all the rules ? On another thread there is a debate of who or who shouldnt get a WS ring.


Fourthly,& this joins SK's point,each user is different some are more or less into the 'sim side' of the game others into 'fun' - why should one be penalised for the BELIEF of one user(even if he is a moderator)?

What you describe as fun may not be MY (or other users) definition;


Eh, I don't think that there needs to be a "do not allow Free Agents to be immediately traded." That's a rule that you can iether follow if you want, or not follow. It doesn't need to be enforced by the game. Say somebody wants to play one way, and then switch to the other, it's much simpler to just...change hwat you do, instead of having to go switch an option.

Yes truein a utopic world - but the ability to do so should be "a game option" NOT ALL users are knowledgeable as you - heck - to switch systems or methods you must have the ability to do so - like in the League options.You should not expect the user to do the adjustment because he has certainly flawedinformation -some users are men who have been wathing the game for years but others are learning & use BM as a guideline so if hte info is incorrect there understanding & playing will be the same until someone/something changes that


By asking that we have a way to turn it off, it would seem to me that I am asking to increase "a user's ability to do anything that they want." /shrug

Exactly:)

[SirKodiak]

well, humm... now i'm confused. That directly contradicts what jonnymo said earlier (more or less), so I guess i'll have to try it all out myself.

I wonder if the results are er... tainted? by editing the contract at all.

I tried it before I edited his salary, and same thing, but since $14.5mil/year for 5 years is what I signed him too, it was possible no one wanted to pay that amount. The more tests, the better.

[FRENCHREDSOX]

Eh, I don't think that there needs to be a "do not allow Free Agents to be immediately traded." That's a rule that you can iether follow if you want, or not follow. It doesn't need to be enforced by the game. Say somebody wants to play one way, and then switch to the other, it's much simpler to just...change hwat you do, instead of having to go switch an option.

I said BM & BMO (I just did it yesterday in BMO signed a guy & traded him within 3 hours to obtain a better player)


I tried it before I edited his salary, and same thing, but since $14.5mil/year for 5 years is what I signed him too, it was possible no one wanted to pay that amount. The more tests, the better.

I did 3 tests using:
1) Kearns(90) at 10m$/3 years then dropping him to 500k/3 years;
2) Webb(97) at 13/3 ---------> 500/3
3) Clemens (92) at 12.5/2 -----------> 500/2



The only one I got offers, other than CASH ,was (2) & that was in July so it seems that the "no trade" rule is implicit in the AI because who wont trade for a 97 pitcher at 500,000$ per year - or the AI "remembers" the original contract,is that possible ?

[SirKodiak]

Eh, I don't think that there needs to be a "do not allow Free Agents to be immediately traded." That's a rule that you can iether follow if you want, or not follow. It doesn't need to be enforced by the game. Say somebody wants to play one way, and then switch to the other, it's much simpler to just...change hwat you do, instead of having to go switch an option.

Same could be said of the Rule 5 Draft, waiver wire, 40-man roster, etc by the people that don't care if they exist, and don't want to have to go switch an option to turn them off and certainly don't want the extra work 'forced' upon them.

Free agency is enforced for all years even though it is rather new (the 70s), users get no choice. Amateur draft is enforced for all years even though it is rather new (the mid 60s), users get no choice. Profit sharing and luxury taxes don't exist. FAs don't cost draft picks (which is the case in MLB but soon won't be). This doesn't matter to some users, but it ruins the experience of others, and has various effects in between. Some people see the game as a failure if it can't get saves or wins figured out correctly, others don't care as long as it doesn't happen all the time. What rules of the game of baseball (regardless if they are well known by fans or not) that BM includes, makes optional, or ignores is a design choice.

As someone points out above, I agree it would be nice if how the rules of the game are interpeted and implemented was included in documentation. For example, why free agency exists in 1927... (the reason is probably very good, but it would make it easier for those that play in the early years and want a game without free agency to swallow if they knew why)

I, for one, could care less about an almanac feature, but I am not going to say that implementing it is not needed, unless it clutters the interface, which I see as the strength of this game.

Obviously, this is Clay's game, he can figure it out in whatever manner he wishes. I just want to throw out some ideas that perhaps he may want to include. Heck, some of the things I come up with I don't much care if they are ever used, but I see where others would like it. It all comes down to Clay and his vision for the game. We have no idea what it is, or which types of users he will choose to target, but that will decide what he does to the game.

[FRENCHREDSOX]

Same could be said of the Rule 5 Draft, waiver wire, 40-man roster, etc by the people that don't care if they exist, and don't want to have to go switch an option to turn them off and certainly don't want the extra work 'forced' upon them.

Free agency is enforced for all years even though it is rather new (the 70s), users get no choice. Amateur draft is enforced for all years even though it is rather new (the mid 60s), users get no choice. Profit sharing and luxury taxes don't exist. FAs don't cost draft picks (which is the case in MLB but soon won't be). This doesn't matter to some users, but it ruins the experience of others, and has various effects in between. Some people see the game as a failure if it can't get saves or wins figured out correctly, others don't care as long as it doesn't happen all the time. What rules of the game of baseball (regardless if they are well known by fans or not) that BM includes, makes optional, or ignores is a design choice.

As someone points out above, I agree it would be nice if how the rules of the game are interpeted and implemented was included in documentation. For example, why free agency exists in 1927... (the reason is probably very good, but it would make it easier for those that play in the early years and want a game without free agency to swallow if they knew why)

I, for one, could care less about an almanac feature, but I am not going to say that implementing it is not needed, unless it clutters the interface, which I see as the strength of this game.

Obviously, this is Clay's game, he can figure it out in whatever manner he wishes. I just want to throw out some ideas that perhaps he may want to include. Heck, some of the things I come up with I don't much care if they are ever used, but I see where others would like it. It all comes down to Clay and his vision for the game. We have no idea what it is, or which types of users he will choose to target, but that will decide what he does to the game.

I wont add anything because this is an excellent summary of the points I was
trying to make

Only one thing - why add roster management (40 man etc) if the dont play by the rules ?(even if its optional which I believe is the best solution to obtain each person's/user's personal level of "fun")

[beerchaser]

I don't see why allowing the signing and trading of FA's as an option would be a big deal. After all, IIRC there are some other things in BM which are posed as yes/no items which may or may not fall in lockstep with MLB rules. For example, I believe you can check/uncheck whether you want to allow trades after July 31.

But y'all are also missing part of my point. I said in an earlier post that I would have been better able to accept the no-trade lockout period if it were explicitly stated as such in the documentation. It isn't. Believe me I looked. I'll be honest, I had no idea that MLB had any such restrictions in place, so I thought was I was trying to do was perfectly acceptable. Had I been able to see that it was not, it would have been factored into my thinking. I might very well have decided to sign Dontrelle Willis anyway, but I would have thought more about it.

[FRENCHREDSOX]

I don't see why allowing the signing and trading of FA's as an option would be a big deal. After all, IIRC there are some other things in BM which are posed as yes/no items which may or may not fall in lockstep with MLB rules. For example, I believe you can check/uncheck whether you want to allow trades after July 31.

But y'all are also missing part of my point. I said in an earlier post that I would have been better able to accept the no-trade lockout period if it were explicitly stated as such in the documentation. It isn't. Believe me I looked. I'll be honest, I had no idea that MLB had any such restrictions in place, so I thought was I was trying to do was perfectly acceptable. Had I been able to see that it was not, it would have been factored into my thinking. I might very well have decided to sign Dontrelle Willis anyway, but I would have thought more about it.

That's the point if MLB rules were TOTALLY implicated or not at all then there would be a certain logic - that is the thing I wanted to point out.


As said by other posters users can do or not do things which are marginal to MLB rules but if you cant trade players after signing them well then a pop up should occur saying so.That way users KNOW what their action limits are.Same occurs when you try to trade after the deadline so it implication cant/shouldnt be that difficult as long as you know in advance

[ohms_law]

Well, I don't support having Free agency prior to the 1970's either.
*shrug*

I've said my piece. If there's some sort of restriction, beerchaser is correct that it should be documented.

[Gaedel]

The Yankees and the Cubs are still spending the winter signing and releasing 50-65 rated pitchers with no upside. Some pitchers will be signed and released several times in the same month.

[HoustonGM]

Is this in historical leagues? I haven't noticed anything of the sort starting in 2006 with my rosters.

[ohms_law]

I have. well, not with your roster's, but starting in 2006. While overall, the number of occurances has been reduced, this is one issue that simply won't go away.

[RAW]

Yes, its happening to me not only in the off season, but also during the regular season and usually to teams that are making a lot of money. I figured its just a way to keep the rich teams from making too much money by signing players, then having to buy out their contracts to release them. It does clutter up the "Transitions" catagory, though. Other than that, so what?

[midwestjw]

Is this in historical leagues? I haven't noticed anything of the sort starting in 2006 with my rosters.

I've seen it with you rosters, but it's not always pitchers. Sometimes thel'll sign a player for 3 weeks and release them. very unrealistic.

[FRENCHREDSOX]

Is this in historical leagues? I haven't noticed anything of the sort starting in 2006 with my rosters.

Had it with the latest version the "bug" (if it is that) kicked in around 2009/10 except it was NYY & NYM :eek:

[ohms_law]

Yes, its happening to me not only in the off season, but also during the regular season and usually to teams that are making a lot of money. I figured its just a way to keep the rich teams from making too much money by signing players, then having to buy out their contracts to release them. It does clutter up the "Transitions" catagory, though. Other than that, so what?

Thing is, the computer run team's do not actually spend money on releasing players.

I'm sure that this problem does have some relationship to how much cash the team has. There's probably something deeper in the AI causing it to occur than simply having over $x amount though.

The "so what" part is two fold. One, there's definately something 'wrong' going on inside the AI, so that should be fixed just on general principles. two, it does clutter up the transactions screen which is more of a problem for some than others. Most importantly though, when this is going on for some team and player(s), it slows the sim down. It's not nearly as significant of a problem as it aparently was in the past, but it's still there.

[Gaedel]

I started a league in 1960 with historical rosters. As noted, other than cluttering up transactions (and the individual player histories), it doesn't really cause any damage. The AI teams doing serial signing and releasing do not seem to get charged for "dead money" when they release these guys after signing them.

Obviously, the problem is in the AI. Possibly having the "review free agents and sign AI" check with the "review roster and release AI" before doing the signing might fix the problem. Another possible fix would be to have a ukase on AI signing a guy twice in one year.

[dolfanar]

Another possible fix would be to have a ukase on AI signing a guy twice in one year.

Or a minimum time needed for a player to be part of a roster before the AI will cut him.

[RHPT]

I'm seeing this also. Very strange.

[Timofmars]

I been playing BM 2005 and found some ways to exploit the behavior of the AI. I just starting playing BM 2007, thinking the AI would be improved, but it seems even easier to exploit now.

In BM2005, opposing teams didn't seem to take into account the number of years remaining on a contract, and what potential that has to save or cost lots of money. And easy strategy is to aquire good players earning the minimum 300k with lengthy contracts, especially newly drafted players with high peaks. Then in their last year of contract right before the trade deadline, while they are asking for millions, I sell/trade them sometimes for 20 million or so depending on ability. The opposing team gets to keep the player for less than half of the season and can't afford to sign the player to a new contract. Then the player is back on the Free Agent list.

Another thing was that I could even sign that player and other highly rated Free Agents even if they asked for large amounts, and it would essential be "free" or very discounted for me to do so. This is because after 1 year passes, teams are willing to pay for the player in a trade, often for quite a bit more that the cost of their 1 year of salary. Before 1 year passes, the AI doesn't even seem to value the player at all, so won't fetch any price.

And this last point also made it possible to unload unwanted, newly signed free agents on a team on top of any small trade. In fact, I once signed as many of the most expensive free agents as I could and then was unloading like 8 of em at a time onto teams along with just a small trade. For example, I'd sell a weak backup player for 1 million, and add to that trade about 7 or 8 players who I just signed that same day with combined salaries of 50 million, and the AI would take the deal. Of course, you'd then see them lose a ton of money over the season :eek:



Now in BM2k7, I notice there is the addition of players options for contracts. And it seems this is a huge exploit. I can offer a player option of 50 million for 3 years and they player will often let me lower their base salary to the minimum possible, much lower than the maybe 5 million a year they were requesting. And as long as I trade the player, release him, or he retires before the option can be used, I have no penalty except for the fact that I can't keep the player for more years. This undermines the whole personality aspect of BM2k7. It really doesn't matter if the player is unhappy or greedy, etc. You can get good deals no matter how poorly you managed player personalities.

Also just as in BM2k5, you can offer older players and even peak players a contract for 7 years and you'll be able to sign them for much less, and it easy to never have to deal with them for the full 7 years. I mean, a good 38 year old guy should not be very interested in a cheap 7 year contract with a 50 million 3 year option if he plans on retiring in the next year or so. Or if he is offered that deal, he should continue to play for the next 10 years even if has to use his bat as a cane to run the bases. :p




So anyway, these exploits can be fixed by first making players smarter and more ruthless in their persuit of money. They should be able to estimate what they will be worth in a few years and should measure the total income of a lengthy contract compared how much they would have earned on a more expensive short term contract with a likely smaller 2nd contract after it expires. Of course, personality should be a factor as it is now. And if you sign them for big money when they should be retiring, they should hang around just to collect that money. Either that or refuse the deal, demanding more up front.

2nd, the teams AI should do the same. They should estimate a players value in the future and should hesitate to accept a trade of an old player in decline who has a long, relatively expensive contract. If the player is only going to be decent as a started or backup for maybe 1 year, the AI should look at the entire cumulative cost of the 7 year contract as being the cost of 1 year of usefulness.

[GreenDiamond]

A lot of these issues still exist, as far as I can tell. (using 9.43)

Personally, I think addressing these issues should be the highest priority, as they have the greatest overall effect on the game. The AI needs to look at the contracts as a whole and calculate their usage value over the length of the contract.

Free agents traded in their first year should depreciate in value rather than be essentially untradeable, you should still take a heavy loss if you attempt to sign and then unload, but it should be pro-rated to a degree depending on how far into the season it is, team need and player performance to that point.

Long term contract implications should definitely be considered by the AI, particularly with aging players or players with exceptionally favorable contracts.

Absolutely agree about the final year of a young players contract as well. The AI should consider the "estimated asking price" at renewal and factor that into their trade bids.

Several good points, and I think a vital part of ownership strategy. Probably worth investing some AI development in this area.

Trades involving large amounts of cash should probably have a percentage chance of a veto occouring as well, although this may be tricky as it would probably require the AI to remember vetoed trades and the ability to recognize a slightly modified offer. Possibly too complex initially.

Definitely like the idea of fixing trade exploits as a high priority, however.

[ohms_law]

Yes, agreed, absolutely.
:)

[pacers721]

the 50 million dollar player option is worse than you described. It's literally impossible for the player to use it (i think the max is 30 mil). So when their contracts are up, they just ask for a new one.

[Timofmars]

I thought all of this was going to be fixed in BM2k7. I read the "Tradezilla (http://www.sportsmogul.com/baseball2k7/tradezilla.htm)" stuff that raved about the depth and competitiveness of the AI in BM2k7. That seems like false advertising after what I've seen so far. Maybe it's there, but just badly bugged?



Free agents traded in their first year should depreciate in value rather than be essentially untradeable, you should still take a heavy loss if you attempt to sign and then unload, but it should be pro-rated to a degree depending on how far into the season it is, team need and player performance to that point.


I think the free agents should be asking for a salary as high as any team is willing to pay them. I know in BM2k7, I picked up a 26 year old 97-rated pitcher for the 9 million a year he was seeking. Teams probably would have been willing to pay well over 20 million for him if he asked. By the time the next season rolled around, people were offering like 20 million for him in a trade, plus they'd have to pay 9 million a year in salaries on top of that. Clearly, he was worth more that 9 million a year.

The issue with free agents having trade value or not would be fixed if the free agents would consider what other teams were willing to pay, rather than just what they themselves would demand. It would be like a bidding war. You have to beat the highest bid to get the player. The highest bid is representative of the most a team is willing to pay, so no team is going to be willing to pay cash to get the player in a trade in addition to having to pay his salary, because his salary alone is the most anyone is willing to pay.

With this method of doing things, there is no need to make an artificial trade barrier such a 1 year period where the player is worthless, or a pro-rated system based on how long the player has been on the team. The player naturally would have no trade value or even negative trade value.

Also, this might help fix some imbalances I've heard mentioned, where the top teams are extremely good all the time with 100 win seasons, etc., while the bottom teams are extremely bad. Right now, a top rated player often has nearly the same salary requested as just a decently rated player. But if the top rated players had huge salaries, while the decent players are more reasonable, then the top teams will pay huge amounts for just a slight advantage. Low budget teams could collect players from the bottom of the barrel for cheap. If everyone wants the best player, it should drive his salary up.

[ohms_law]

All true. The problems have been steadily being fixed and improved upon as time goes by, and things only look to be getting better as we get into the Mogul 2008 Beta period, which is about to start soon and will be an open beta for everyone that owns 2007 and wishes to participate (basically, it'll be a series of patches to 2007 until the game is ready for release).

One of the primary problems, and the reasons that the salaries are so close, is that the AI teams are simply unwilling to compete for most highly rated players on the FA market each season. As time goes by, one or two players are signed immediately, and the rest simply whither away on the market dropping their asking price until everyone's in similar price ranges. The Ai simply is not aggressive enough, in many cases.

Still, the game isn't completely broken. There are definite problems, but it's not as if it hasn't been addressed at all.

[Timofmars]

I think this should be a priority. The new personality things are an interesting idea, but they mainly affect players' salaries, not player performance. So this is almost completely undermined by the existing problems with the AI that makes the effect of personalities something to ignore.

Lots of the nice touches added to the game have little imporantance compared to the broken AI. I don't care about the minor details of a player's ability to pivot and throw on defense because details like that provide such miniscule effect compared to learning how to deal with the AI.

I'm just saying I'm completely bummed out after finding the AI is even easier to take advantage of than in BM2k5, while after reading the "tradezilla" rave about the intelligent AI, I thought the AI would be very competitive. I just got the game, but I don't care to play now. All the bells and whistles and realism is undermined by the completely unrealistic my team is gonna look in a few seasons. I'll probably have all the best players in the lineup plus all the best prospects in the minors!

[ohms_law]

It is actually a priority. More work has been done on the AI this year than anything else, really.

It is easy to game the system right now, if you know what you're doing. The biggest thing I can offer to you is ways not to do so, as the AI does a fairly good job (although there are certainly still large problems) on it's own. If you're totally bummed though, you could just wait until beta starts, and test it out and offer feedback with the rest of us. The game is a living thing, so just because there's a problem with it right now doesn't mean that the game is worthless or anything.

[GreenDiamond]

I'm just saying I'm completely bummed out after finding the AI is even easier to take advantage of than in BM2k5, while after reading the "tradezilla" rave about the intelligent AI, I thought the AI would be very competitive. I just got the game, but I don't care to play now. All the bells and whistles and realism is undermined by the completely unrealistic my team is gonna look in a few seasons. I'll probably have all the best players in the lineup plus all the best prospects in the minors!

Its tougher than it used to be. Playing a small market team on mogul level is still quite challenging, so far, for me. The trading AI is a lot stronger than it was in earlier editions, and better than most games of the genre. Some exploits are possible, but I wouldn't call the game broken by any means. Just looking to improve it to the next level.

[jonnymo]

I thought all of this was going to be fixed in BM2k7. I read the "Tradezilla (http://www.sportsmogul.com/baseball2k7/tradezilla.htm)" stuff that raved about the depth and competitiveness of the AI in BM2k7. That seems like false advertising after what I've seen so far... The highest bid is representative of the most a team is willing to pay, so no team is going to be willing to pay cash to get the player in a trade in addition to having to pay his salary, because his salary alone is the most anyone is willing to pay... With this method of doing things, there is no need to make an artificial trade barrier such a 1 year period where the player is worthless, or a pro-rated system based on how long the player has been on the team. The player naturally would have no trade value or even negative trade value.
There is no artificial trade barrier. (http://forum.sportsmogul.com/showpost.php?p=630927&postcount=25)


I'm just saying I'm completely bummed out after finding the AI is even easier to take advantage of than in BM2k5, while after reading the "tradezilla" rave about the intelligent AI, I thought the AI would be very competitive. I just got the game, but I don't care to play now. All the bells and whistles and realism is undermined by the completely unrealistic my team is gonna look in a few seasons.

Its tougher than it used to be... The trading AI is a lot stronger than it was in earlier editions, and better than most games of the genre. Some exploits are possible, but I wouldn't call the game broken by any means.
I agree with G.D. In fact, after Ohms' argument against using cash in trades, I modified my approach to be more representative of reality: no more than $1M may change hands unless one of the teams is paying a portion of the salary of one or more of the players being traded away.

Tim, it seems like your biggest complaint is that it's too easy for you to cheat. I have any easy fix. Feel free to call it "9.46." Ready? Just don't cheat.

Let me know how that works for you. :p

[GreenDiamond]

The highest bid is representative of the most a team is willing to pay, so no team is going to be willing to pay cash to get the player in a trade in addition to having to pay his salary, because his salary alone is the most anyone is willing to pay... With this method of doing things, there is no need to make an artificial trade barrier such a 1 year period where the player is worthless, or a pro-rated system based on how long the player has been on the team. The player naturally would have no trade value or even negative trade value.

There is no artificial trade barrier.


I agree with G.D. In fact, after Ohms' argument against using cash in trades, I modified my approach to be more representative of reality: no more than $1M may change hands unless one of the teams is paying a portion of the salary of one or more of the players being traded away.
I think the engine could be modified to address this situation.
In determining the players trade value the AI should consider in making its bid.

1. What that team would be willing to pay if the player were a free agent, for a contract of the existing length (pro-rated for percentage of active season completed).

2. Modify by any expected depreciation/appreciation of players value due to age/experience factors during the contract period (ideally this should be built into the free agency engine to begin with).

3. Add desirability modifier (probably a multiplier) for current season need to the prorated value of the current season. This would consider:

a: Teams likelyhood of making playoffs (if established player).

b: If team is out of contention, potential to reduce salary by removing an equivelant player of higher salary from lineup. (set trade flag for replaced player)

c: Improvement of team rating by replacing existing player with offered player.

d: Players increase to bench rating (if non-starter)

e: Players increase to farm rating (if prospect).
* add these factors up (contention + increase) to set multiplier.

4. Subtract the known contract value (anticipated expense of owning the player). This should be calculated for time remaining on the contract.

This should set a base value for the player.

The team should consider whether the player acquisition is within their budget, and modify their offer appropriately.

To further increase realism. Players with contracts expiring in the next two years (1-3 anyhow: possibly age based), should have a future profitibility calculation run. This would be based on the anticipated players demands and whether the team would normally choose to resign or release the player at that point.

At this point the player should have a reasonable value assigned to them for trading block purpouses.

Any trade involving large amounts of cash should take into consideration what the team could get in an average free agency market for the same amount of money. Spending 45 million on a prospect should be the equivelant decision of getting three 15 million dollar players from FA.

While this sounds like it is fairly complex, it is entirely possible that the majority of the subroutines exist in the program already. The payoff in return of trading AI strength would probably be justified in that case. Anyhow, it is a potential approach to the problem.

Cash for player trades should be evaluated as cash:salary ratio. This number could be used as a cutoff point or a percentage chance for a commissioner veto occouring. A cutoff point, while less desirable, is probably easier to implement as it would reduce the need to track similar trades. This could probably be implemented as a league setting option.

Anyhow, my thoughts on the matter.

[ohms_law]

1. What that team would be willing to pay if the player were a free agent, for a contract of the existing length (pro-rated for percentage of active season completed).
From what I understand of the AI and the "Tradezilla" system, this is actually considered. It's also the primary barrier to the AI considering a player worth anything at all.
The way I understand it, the AI generally considers the players ratings, and then compares them to his contract. A newly signed player who's contract is at or above what the AI would sign the player for is then basically worth nothing, if not worth a negative value, and therefor the AI simply refuses to deal for that player (you can throw in money on your side, to force the issue).

That effect was apparently lessened slightly over the 9.44 lettered patches, so it's not as immutable as it used to be. The AI is still, obviously, a bit immovable from it's position on many players though, in that it simply will not consider trading for a player that it thinks is over-priced. Of course, the primary problem with this is that it doesn't take into consideration need. If there are very few third basemen league wide, for example, the AI will simply not step up to be competitive about seeking out a third baseman. They all quickly become "too expensive" for the AI.

[Timofmars]

There is no artificial trade barrier. (http://forum.sportsmogul.com/showpost.php?p=630927&postcount=25)

If I've read that post correctly, that is just showing how the free agent you've signed has no value to other teams immediately after signing the player. That's the way it should be, since other teams would have just offered the player a higher salary if they wanted him when he was a free agent. They wouldn't want to pay his salary plus give you players or cash in a trade on the following day after you signed the free agent.

But the problem is that the barrier IS artifical. Immediately when the next season's free agency period begins, you're able to offer the player in a trade for large amounts of money if he's highly rated, even if he was just as highly rated when you originally signed him. The cash amounts being offered can be much more than his yearly salary, meaning you profit by signing the player no matter if you use him or not.


I agree with G.D. In fact, after Ohms' argument against using cash in trades, I modified my approach to be more representative of reality: no more than $1M may change hands unless one of the teams is paying a portion of the salary of one or more of the players being traded away.

Tim, it seems like your biggest complaint is that it's too easy for you to cheat. I have any easy fix. Feel free to call it "9.46." Ready? Just don't cheat.

Let me know how that works for you. :p


I don't want to have to decide where I should give a freebie to the AI because I think it is flawed, and where it is legitimate strategy. If the AI undervalues top young prospects, should I not buy those players from them? Should I offer extra cash and players for those trades?

If I'm able to trade a good player with a high salary to get an equal player with a low salary and only have to pay a small extra price, should I deny myself the trade because I think I get the better deal overall?

When I negotiate a player's contract and have to decide whether to offer a longer contract to get a lower cost per year or a shorter contract with a higher cost per year, I should take into account that he could be worth a reduced amount to the AI in a trade in a few years if I give him the longer contract, which could counter the amount I save on his salary now. Maybe I could get stuck with the player while I have a prospect that has improved to take over his position. Unfortunately, the AI makes the decision easy.

I shouldn't have to decide where the AI is being stupid or where they are doing what's best for their team. Am I making a good deal or am I taking advantage of an exploit?

If I made every deal into the most equal deal I could imagine, I'd probably never improve my team.

GreenDiamond, I can think of quite a few more considerations the AI could take into account. Ohm's observations of the scaracity of good players at a position is just one.

[ohms_law]

For what it's worth, I agree with both of you. There isn't an artificial barrier, per se, but the way that the AI works it creates an artificial barrier.

I agree also, and I think that pretty much everyone else agrees as well, that the AI needs more attention. I know that I have been, and will continue to ask for AI improvements. Several other's will as well, and Clay and Ian seem willing and able to make changes. Clay wasn't willing to make really big changes to 2007, during the season is all. With beta coming up, I'm assuming that there will be some rather large changes made, since we've collectively been harping on these sorts of issues for months now.

All I'm trying to say is that I'm certain that there will be changes made, as much as possible without completely breaking the system, sooner rather than later. Keep talking and pointing out problems, and I assure you that Clay and Ian are reading these things. Just don't be surprised when we offer work a-rounds, since we've been dealing with some of these things all season.

My primary suggestion all season long for Clay and Ian is to implement a "player agent AI". While it would be very inefficient looking from a planning perspective, I believe that if there were a separate AI that specifically looked at deals from a players perspective, it would solve many of these problems immediately. Players should be able to say "No" to teams though, which they really can't currently. Especially to the user. The AI seemingly works out deals amongst itself right now, with each team having a value for each player, and an amount that it is willing to spend on players. If the Player's had their own AI looking out for their interests specifically, the AI teams wouldn't always be able to get their first choice of player. Especially if the player AI were programmed to ensure that the player received as much playing time as possible (turning down large offers from the Yankees because they already have two second baseman and don't need a third, for example).

[jonnymo]

If the AI undervalues top young prospects, should I not buy those players from them? Should I offer extra cash and players for those trades?

If I'm able to trade a good player with a high salary to get an equal player with a low salary and only have to pay a small extra price, should I deny myself the trade because I think I get the better deal overall?

...Am I making a good deal or am I taking advantage of an exploit?
Yes; no to cash and yes to players; yes; the latter.


If I made every deal into the most equal deal I could imagine, I'd probably never improve my team.
Seems the game is suddenly more challenging, no? To paraphrase what Ohms wrote in an earlier thread, since the AI never asks for nor offers cash, either in trades with other AI teams or with the human team, why should we?


GreenDiamond, I can think of quite a few more considerations the AI could take into account. Ohm's observations of the scaracity of good players at a position is just one.
If we're not going to change each other's mind on the cash issue, I'm willing to agree to disagree. However, I will agree with you that Ohms makes an excellent point.

[Timofmars]

[FONT="Georgia"][COLOR="DarkRed"]Seems the game is suddenly more challenging, no?

But then it's not really a game. There's hardly any reason for me to select a team to play. I might as well just watch things unfold as a simulation, rather than an interactive game. The biggest part of the game for me should be the management of players, buying, selling, trading, managing economics of it all. But take that away and all thats left is low-level management such as play-by-play and lineup/pitcher selection. I'm not interested in that aspect of the game.

But to not buy young prospects because I think they are undervalued by the AI, to not sign the best free agents because they ask for little (or sign them for much more than they ask), to refuse to accept the additional cash to balance a trade, etc... It's just not fun. I can understand not using the player options in play negotiations, but the rest of what you suggest just isn't fun. This is like a game of tic-tac-toe. I can either play the best moves and never lose, or I can play bad moves on purpose to give a chance of losing. Neither way is fun to play. The reason tic-tac-toe isn't fun either way isn't because of me, but instead because the game of tic-tac-toe isn't challenging!

Anyway, making the player AI look out more for its own self-interests would be a very good start. It might possibly fix a lot of the problems. Maybe the high salaries of the top players would drive teams to value the cheap salaries of prospects.

[GreenDiamond]

That effect was apparently lessened slightly over the 9.44 lettered patches, so it's not as immutable as it used to be. The AI is still, obviously, a bit immovable from it's position on many players though, in that it simply will not consider trading for a player that it thinks is over-priced. Of course, the primary problem with this is that it doesn't take into consideration need. If there are very few third basemen league wide, for example, the AI will simply not step up to be competitive about seeking out a third baseman. They all quickly become "too expensive" for the AI.
I can see where this can become complicated to program, particularly in light of the possibility of moving players to alternate positions. The lineup manager still needs work in regard to this aspect, so that probably has to be considered first, as it is a necessary step in the player pool evaluation.

Unless the process is fairly quick, it should probably be done at the end of the season, the start of the season, and somewhere between all star break and approaching the trade deadline (when it normally considers trade offers for the final run).

1. Team should calculate alternate position ratings for any players meeting a basic defense rating threshhold. Players with significant contact/power ratings should also be included in this routine.

2. Each team should optimize its lineup.

3. The AI should consider positional adjustments based on this info set.

4. This should be done for all teams.

5. A median skill range (by position) should be determined for all starters.

6. A count of elite positional players should be made. This is necessary to determine if the position is achieving an average due to high skill players and low skill players, or whether the abilities are fairly uniform for the position.

7. A secondary score should be generated considering other potential major league players: bench players, high rated minor league players, and any player on the DL for a reasonably short period of time. (ex:for less than 25%of the remaining season length).

8. Using this information, the AI should be able to take stock of the relative availability of players.

For instance, it would know that there is a surplus of third base talent and that the average overall player rating is 88 at third base, league wide and that there were sixteen players that could perform at this level in the league that are playing other positions currently.

Using this information it can create a position by position profile of the leagues strength at any given position.

With an accurate league wide positional profile in place, the AI would be able to make effective salary evaluations based on supply and demand. It would be able to consider the degree to which a trade is likely to increase its lineup strength in general.

Depending on the speed of these calculations, it might be able to incorporate a two or three year projected value (based on typical skill improvement/decline) to get a much more accurate assessment of the league's overall picture.

This would be useful for longer term decisions such as moving players in the low minors to fit the teams future needs (based on that teams current profile, particularly positions staffed by aging, expensive or scarce players). It would also be useful for helping the AI recognize league wide positional gaps in talent that could drive a players value up or down.

The secondary players rating in the teams profile (bench, top minor league prospects, any player on DL) can also be used to determine whether it is worthwhile to make a long term contract or trade for a player with a long term contract. If the team has no players in its reserve system that would be able to step up at secondbase and there is a league wide shortage of secondbase talent then it would increase the player valuation of secondbasemen. It would further increase the value if their current secondbaseman is below league average in ability, in terms of trade prioritizaiton.

The deeper this profiling can be made, the stronger the AI trade ability will become. Whether implementation of a system like this is a long term thing, or something which can be gradually brought into the game, I think it should be considered in terms of on-going development.

Ultimately a superior trade/finance AI will make the game stand apart from its competitors and increase the challenge level of the game significantly.


Anyway, making the player AI look out more for its own self-interests would be a very good start. It might possibly fix a lot of the problems. Maybe the high salaries of the top players would drive teams to value the cheap salaries of prospects. Another angle on the same problem, that has merit and could be used in conjunction with the above system.
Player AI's would probably not consider positional profiles of potential roster shifts, this is more of a GM angle.

They would consider things like length of time already spent in the team's city, likelyhood of being a starter at their primary position, their overall happiness and inbuilt climate preferences. Ambitious players would also have to consider playoff chances and other personality factors as well.

Players who had a lot of time invested in a city, unless particularly greed motivated, would be more likely to negotiate at reasonable rates. This would create a mechanism that might assist in helping a team maintain a core set of players, as many players would forgoe a certain amount of increased cash in exchange for some lifestyle continuity. Changing cities means the player will have to uproot, sell their house, leave their friends behind and soforth. The longer a player is with a team, the deeper and more meaningful those connections will be to that player, or at least most players. What that lifestyle stability is worth will probably depend on the players personality profile.

Players held in the minors until a late age would probably demand large salaries when negotiating with their own teams in order to either gain a starting spot or be traded to someone who will. Less talented, younger players would be more flexible in this regard. This can be reflected in a cumulative "happiness" effect when the player feels that they are advancing too slowly or at an acceptable pace.

The player should remember their happiness rating with their last team when becoming free agents, and demand more to be resigned by that same team than it would with other teams. This could force something of an up or out mechanism which might benefit the league as a whole.

If both AI's are eventually implemented, it could create some interesting contrasts and would likely force the market to balance itself out better through supply and demand.

[ohms_law]

Exactly, well said. I'm really hoping for such changes/additions to be made in Mogul 2008, personally.

[LetsGoOakland]

To me, Baseball Mogul has been all about becoming the general manager of any team at any point in MLB history (look at how the description of BM is worded: http://www.sportsmogul.com/baseballcd/index.htm#What). Unfortunately, I think that this series is losing that focus.

I buy Baseball Mogul every year for the updates on the general managing side of baseball. I was quite disappointed last year when the main changes to the game were based on the play-by-play action. While this side did need an upgrade, I am further disappointed by the beta thus far. I am quite aware that 08 is in beta and more features are planned, I simply want to voice my opinion now that I want to see more updates to the general managing side of Baseball Mogul. I am curious if the plan is to try and spend development time on play-by-play.

What happened to new features for general managers? Hiring managers, 3-way trades, signing contracts before final year, more in-depth negotiations with free agents, bidding for free agents, Rule 5 draft, minor-league options, improved trading block, waiver wire, back/front loaded contracts, overall improved off-season, etc...

While I understand play-by-play is important to some people, I feel this game's strength is general managing a MLB team. I hope to see more effort, in the month(s) leading to release and over the next few years, on improving this strength.

Thoughts?

[ohms_law]

I tend to agree somewhat, but the play by play is a big competitive advantage. we'd all be feeling that something was lacking in Mogul vs. other games if not for the option to use play by play.

We are getting inflation, and a generally more accurate financial model (once all the bugs are worked out), so play by play isn't the sole focus. I do agree that work needs to be done aside from play by play, but those are the type of features that keep users, not the kind that brings fans into the game to begin with. Enhancing the financial aspects (beyond bug fixes) after release is perfectly fine by me. Remember, this is sports mogul we're talking about, not EA or Sega. The end of beta hardly means the end of work on new features or bug fixes (never mind the fact that beta just started).
:)

[boomboom]

I agree with this [somewhat], but you do have to realize in order to sell this game. They had to improve the Play-By-Play.

Maybe in 2k9, they can get some more general manager options going.

[ohms_law]

Heck, hopefully over this year we can get more.
:)

[dejota]

I agree with this [somewhat], but you do have to realize in order to sell this game. They had to improve the Play-By-Play.

Maybe in 2k9, they can get some more general manager options going.

So what? Us loyal users for multiple years should just deal with it? I'm completely with the original poster here. Very little has changed as far as managing the game since '06. The majority of the suggestions on this very forum have echoed similar opinions for two years now (ever since I started lurking).

I'm not complaing, I love the game. But Play-by-Play is for the playoffs and WS exclusively for me...except for the occasional franchise that I play daily. But often times going into play-by-play drives me mad. A) Runs charged to a pitcher are hardly ever accurate B) My players play about 20-30 percent better in play-by-play mode and C) At the end of a season spent mostly in Play-by-Play I'll have 2-3 50HR guys (I use the Astros or a bottom feeder) and at least one or two mediocre pitchers who pitch like Cy-Young candidates.

Maybe that's the stuff that's being fixed but I still feel simulating games gives me the most realistic results. I'd prefer to do a three-way deal, get realistic arbitration figures from my players and not have to get rid of budding stars their first year in arbitration because the game doesn't understand how the process works. Give me a more accurate Minor system where I can only start 5 guys regularly and could customize my line-ups and rotations. How about a prospect value being placed more on his expected potential than current talent? Club control rules?

This isn't MVP baseball or The Show...that's why I freaking play the game. If I want an accurate in-game game then I'll go get '07 in a week...I think that if the goal is to ultimately get more users the best approach isn't trying to take customers from an established niche it's to create your own. And so far BBM has done an excellent job at that...why abandon it now. No other game is as compelling to me year in and year out as BBM but if we continue upgrading the Play-by-Play without progressing the management side of things the consol games are going to catch up. Franchise and Dynasty modes improve every year on consols and the differences that make BBM so great are become fewer and fewer.

Please don't abandon what makes this game so great to tweak a cool feature, IMO.

Sorry for the novel. But I love me some baseball mogul and have been very dissapointed with the lack of updates the last few years management-wise.

[Swampdog]

Ok, I started to second what LetsGoOakland said, but then I read the post by dejota...and I am not intelligent enough to add anything of value to his sentiments, except, well said......VERY well said.

[WHAK0985]

I'm gonna have to agree with pretty much everything said in this post, even Ohms' wistful hope of adding the said managerial features over the coming year. While it is VERY early to comment on the BETA and BM08, I have to admit that there was a bit of disappointment on my part upon opening up the Beta for the first time and discovering that it was very much the same game as BM07 other than some cosmetic changes to the display and the PBP. I do hope that we don't have to wait until 09 to see the changes the majority of us desired when discussing the beta and 08 all those months ago.
But that encyclopedia feature sure is going to be ridiculously awesome though!

[FRENCHREDSOX]

Agreed with the original poster but I think that when the BETA basis is set that added features (like better management/40 man rosters etc) will be added.THE 9.61 patch is a vast improvement on 9.6 - I mean you can actually PLAY lol but lets give it time to get the basis stable & see the direction ...just my 2 cents

[ohms_law]

The primary goal behind Mogul 2008, at this point anyway, is obviously to deal more with infrastructure issues that have cropped up with 2007. There's been a lot of work done on the database, and structural things like inflation and the encyclopedia. Developing those types of features fall into the "heavy lifting" category, it's tough grunt work that isn't really noticeable on the surface unless you know where to look. The good news is that with that work (basically) accomplished, it lays the foundation for some great leaps to be made in the things that we all really want, that some of you have mentioned here.

In short, don't be too disappointed. The game itself has become much better now, and the feature additions are only a matter of time.

[boomboom]

2k7 was the leap years ahead of 2k3 (which was leapyears ahead of 2k)

from the front page.



NEW Financial Model - Automatically adjusts revenue levels and salaries to the correct level for each era, from 1901 to 2007. And new options give you complete control over your universe: disable free agency, change the inflation rate, or adjust the luxury tax.

This would be an advancement to the GM part of the game right?

[mflournoy]

I fully agree with LetsGoOakland, dejota, and others here.

I never use play-by-play. Not because it sucks (which it does, really), but because I'd play a "real" baseball game if I wanted to do that. I really can't imagine any improvement to play-by-play mode that will make me want to use it.

I play Baseball Mogul for the general management features. I've lurked here for a while, and I've noticed that 40-man roster functionality, the Rule 5 draft, true-to-life disabled list functionality, minor league options, and a non-screwy arbitration process are desired by everyone. The moderators here respond to such requests by saying, "Yeah, we've heard that before, it's a commonly requested feature." Their attitude conveys a message of, "Geez you are dumb, we know you want that, but just shut up and stop annoying us." Talk about a backwards message.

In my estimation, the quickest way to stop these complaints is to fix the problem, not alienate the users. Does anybody really think that play-by-play mode was the element holding this game back? Instead of doing one thing (general managing) really well, Baseball Mogul looks to be lackluster in everything. Rather than catering to a niche market, it caters to no one.

Now, for some specifics:


This would be an advancement to the GM part of the game right?

No, not really. It is an enhancement of the environmental options, not the general management options. And it's nothing that you couldn't "fake" by manipulating the money anyway, if you wanted to.


In short, don't be too disappointed. The game itself has become much better now, and the feature additions are only a matter of time.

I am a programmer, so I can truly appreciate intensive framework improvements. But the bottom line: customers don't care about improvements that they can't discern. New features are only a matter of time? Then when that time comes, I'll buy the game. As it is, I won't buy BM08.

[ohms_law]

"Yeah, we've heard that before, it's a commonly requested feature." Their attitude conveys a message of, "Geez you are dumb, we know you want that, but just shut up and stop annoying us." Talk about a backwards message.
I'm sorry you feel that way. There's really nothing I can say that will change your perception on this issue, but I will state for the record that that's not what I've said anywhere to anyone. There are old posts that cover many topics in extreme depth here, but often when someone new comes along they end up either reopening discussion in a new thread (making it obvious that they haven't read what's been said before), or they impose themselves in the middle of an ongoing conversation repeating something that has been dealt with before. There's nothing wrong with that many times, but occasionally someone will say something that is so obviously misinformed that it begs a correction.


customers don't care about improvements that they can't discern.
Sure, but you know as well as I do that you can't get the improvements that people really notice without the framework development. It's easy to say "I'm a programmer", and it's easy to say "just fix the problems", but neither you nor I are the ones doing the work.

Anyway, you're not buying the game apparently anymore, so what do you care? This reply is nothing but trolling for an argument, is all.

[mflournoy]

Sure, but you know as well as I do that you can't get the improvements that people really notice without the framework development.

I can appreciate that, of course. I didn't intend to convey that I don't think framework overhauls are necessary, only that they are not very marketable in and of themselves.


Anyway, you're not buying the game apparently anymore, so what do you care? This reply is nothing but trolling for an argument, is all.

That's a very cavalier attitude to take. I clearly said that I will buy the game when it features the improvements that interest me. In the meantime, I'll keep using BM07. Why wouldn't I care about the game anymore?

[dejota]

from the front page.


This would be an advancement to the GM part of the game right?

To an extent, yes. But those aren't the features that I'm looking for. At least not in that light. Some negative impacts of inflation in BBM07 was that it'd impact arbitration. Despite the current market value for guys like Willis or Cabrera neither were close enough to FA to make too much in arbitration. Miguel would've never accepted 7.Xmillion in Baseball mogul...And from what I can tell it has no idea about club control rules.

Or why is it that Andy Pettitte is so willing to take a paycut despite having success over the past two/three years? Why is Carlos Lee demanding a bigger paycheck that Soriano was when I was doing my roster edits to keep up with MLB? Does contract demands not take into account how many skills the player has?

Pitcher's aren't overvalued like they are in real life leading to the Yankees dominating pitching rotation after three or four years. Tomo Ohka wants 5-6 million in FA while Zito and Schmidt are satisfied with 7-8.

These inconsistancy's are why I think the inflation programming is inaccurate. I'm no programmer and really don't know which variables inflation is based or contract demands but if those issues were adressed I think that's where you'd find the improvements that everybody would notice.

Just my 2cents.

[ohms_law]

That's a very cavalier attitude to take. I clearly said that I will buy the game when it features the improvements that interest me. In the meantime, I'll keep using BM07. Why wouldn't I care about the game anymore?
OK well, That's not the way that your post came across is all. Like I said before, the tone was more like you were looking to pick a fight; trolling.


Pitcher's aren't overvalued like they are in real life leading to the Yankees dominating pitching rotation after three or four years. Tomo Ohka wants 5-6 million in FA while Zito and Schmidt are satisfied with 7-8.
Curious, what level do you play at? Every game that I play, pitchers make substantially more than most other players do.

Also, you can't really draw direct parallels to real life contracts, even with "inflation" in BM08. The fact is, it's a game. The value of a dollar in game will never really be what it is in real life. There are other things going on in real life that simply don't happen in Mogul, as well.
That's not to say that the AI couldn't be better. It should be better. I'm fairly certain that it will be better as well. Still, I think that there's a bit of unrealistic expectations occurring here.

[mflournoy]

OK well, That's not the way that your post came across is all. Like I said before, the tone was more like you were looking to pick a fight; trolling.

I'm sorry. I didn't mean to come across that way.

I enjoy BM07 greatly, and I'd been looking forward to BM08 for some time. Just like everyone else, I have my own laundry list of features that I want to see included. But from what I can tell, none of mine have been implemented, so I'm disappointed. In fairness, I have not tried the beta for myself, and I realize it is just a beta, but I can't imagine brand new features being introduced before the official launch.

When Baseball Mogul includes expanded roster management, I will enthusiastically buy, but until then, I think I'll pass.

[ohms_law]

but I can't imagine brand new features being introduced before the official launch.
They were added just before, immediately after, and all during the year this last year. Same thing happened with BM06 as well. Sports Mogul is not EA or Microsoft. As far as I can tell, there's never any feature freezes.

Anyway, I understand what you're saying now. It's just a perception problem, really. Their a month away from launch still, for one thing. It's a very early beta. And, like I said above, an SM launch isn't even close to the end of the road.

[mflournoy]

I see. I admit, I am a relative newbie to Baseball Mogul (of no more than four or five months), so I'm not too familiar with the release process.

I will keep monitoring the progress, but it strikes me that there doesn't seem to be a very clear feature list. Is there any specific page I can refer to for a list of new features with each release?

[ohms_law]

Not really. Clay is pretty tight lipped about it, although actually I don't think that he's really keeping anything from us. It's probably more like, he's not really sure himself exactly what is coming. I can tell that there's things that he wants to do, but he normally doesn't commit to anything until it's basically done. You can't really blame him too much, since it's just him and Ian that work on the game.

I guess that right now, the big thing consuming his time is the encyclopedia feature. It really is a big deal too, since now we'll get all the stats that our hearts could desire, from the entirety of the game that's been simulated. But then, I'm a stat head so things like that impress me easily.
:)

The play by play feature is new as of 2007, so it's not really surprising that there's been a lot of time spent on it. It's kind of going through a version 1 - version 2 transition right now, if you could version anything in Baseball Mogul.

As a programmer, I think you'll understand what I mean when I say that, overall, Clay is basically an adherent to the Perpetual Beta (http://en.wikipedia.org/wiki/Perpetual_beta) development model. He's not an ideologue about it by any means, but that's generally the way that he works. It's not that uncommon that people that are newer to the community here simply don't understand or recognize the concept, though. I suspect that for many folks, it's their first encounter with the development model.
*shrug*

[mflournoy]

The encyclopedia feature is one that interests me. I had forgotten about it. (Easy for me to do, not having played the beta...) On my mental list of desired additions is something that amounts to an auto-generated Baseball Reference for each simulated universe. So if the encyclopedia performs well here, then that's a definite plus.

So no list of future additions - understandable, but what about list of past additions? I mean, one page where new features and updates are bullet-pointed for each past version. That would be a lot easier to keep up with than sifting through umpteen threads on a message board.

[djbenz1578]

I'd love to see the 40 man roster, waiver wire, rule-5 draft, etc., but I think that might be too advanced for the casual gamer/baseball fan. If they do want to implement those things into the game, maybe an option to turn off or not have to do those things.

[NYDodgers]

I know it's a been promoted a lot but things like waiver wire and 40-man rosters aren't absolutely necessary. I just don't think they would add too much enjoyment to the game.

I agree with both sides of the argument of how we need new changes to GM mode and also how Clay's added some amazing features months into the release before so we shouldn't reach any conclusions just yet.

However, the one and only thing I've wanted for BM08 is a new free agent system. It's unrealistic being able to sign any player you want and winning the world series every single year. I'm currently on a 36-year world series winning streak with the Reds(starting from 1901, I won it first in 1904, then from 1906 on till now).

There's obviously other factors besides the free agency system that factored into my streak(such as the CPU undervaluing prospects), but I think it's one of the most prominent factors. I've been playing mogul for a long time now but I think even mediocre players with little knowledge of BM can win almost every world series with a team like the yankees, since they can just sign the best people at each position.

[dejota]

Maybe I'm overvaluing such things as the waiver wire and rule 5. But I think all these things I'm asking for go hand-in-hand.

Players don't just start asking for more money or less money one day. They compare themselves to similar players in reference to revenue and their own personalities. Until a player can "think" and a manager has all the tools at his disposal a truly accurate mogul is impossible. Maybe I'm asking for too much but when two years pass and very little has changed on that side of the game questions are beckoned.

Gary Sheffield (for example) isn't just going to play anywhere. And he's not just going to sit by if he's traded to a mid/low market team. That's part of his personality. And that should be reflected in as many aspects of the game as possible, IMO.

Or if you've got a young gun like Willis tearing up the league before he's supposed to. He's going to expect the big bucks but, like baseball, there should be stop-gaps in place preventing that. Nothing is more frustrating than your top-prospect tearing up the league in his rookie season when you were just planning on testing the waters with him then in arbitration he requests millions. At that point and for the next two or three years his pay should never break 5 million dollars.

How about building into the players personality a system that rates their opinions of teams. I.E. A prospect from Houston would start out his career with certain prefernces. If he's a primadonna type then the NY and LA teams plus Houston and the team that drafts him should be his top teir selection. Then if a player has a good experience with a team (basically success and adequate playing time etc.) he'd be more inclined to resign and/or take a hometown discount (if he's not a primadonna). If he's not happy (losing seasons, inadequate playing time, personality conflicts) there's no way in **** he'd resign and that team could have a negative value as far as that player is concerned. I admit I'm no programming expert but I have experience and a personality model would be as simple as assigning a +1, -1 system.

A player's drafted by Tampa Bay so they get a +1. Well he's from Cali so there's a -1 for whatever reason. For every winning season that they contribute to regularly is a +X amount and for every person he get's along with on the team is a +.25...or whatever and so-on. The value's arent important and could be tweaked but the structure doesn't seem that complicated. I'm by no means familiar with anything as complicated as BBM and haven't the slightest clue how you'd impliment and/or intigrate something like that but it's not as complicated as say inflation models.

From this formula you get a basic, numeric value of how a player feels towards a city/players/etc. Apply basic stats (to figure out means and standard deviations and ranges and how players opinions compare to one another) and from there you could add a whole new aspect to the game that would greating improve the free agency market, IMO. Teir the player's opinions of the season and then you can even improve the rumor mill. If the Cubbies are looking for a second baseman and a certain 2nd baseman had the Cubs in their top teir they'd probably be talking.

Or if a two teams are offering a contract to a player, one in his top tier and on in his, say, third, it's not unreasonable for player to take a minor discount to go to his prefered team (again, unless personality (love of money, etc.) conflicts with that). And now that I'm double parenthe-sizing (not even a word) I'm going to walk away hoping I made some sort of sense.

I mean no presumptions and don't mean to come across unappreciative of BBM or its staff. But I figured it's easy to point out problems and much harder to suggest a solution.

[NYDodgers]

I whole-heartedly agree with the basis of what dej just proposed. The new "player personality" feature for 2007 was a step toward the right direction, and I think it's an area that has huge potential for BM08. It's hard to create and force new features on a game every year, but I think the players' personalities part of the game is a great area to really put some improvements on. It was touted as a significant improvement for BM07 but there was very little need to pay any attention to it(charisma was somewhat useful since he'd be a fan favorite)

[Red Sox Fan]

I noticed in beta 9.62 that the draft is still being held in January...also noticed not a lot in terms of Game Features has changed for 2008....when will we see some of the game changes we all voted on?

[ohms_law]

I agree completely, regarding Free Agency and the players lack of self interest as discussed. I just don't think that a waver wire and 40 man rosters are the magic bullet that will solve those problems. I also agree that adding them will make the game more difficult for the user's, and not really add that much to the experience. What really needs to happen is what I've been advocating for the last several months, which is player specific AI.

If Players had their own AI in the same fashion that managers each have their own AI, things would be instantly better. The players need to be able to look out for themselves, and simply refuse to negotiate with teams where they will not receive playing time, or teams that meat other criteria. That alone will go a long way in resolving the problems with "super teams" that tend to develop in Mogul.


I noticed in beta 9.62 that the draft is still being held in January...also noticed not a lot in terms of Game Features has changed for 2008....when will we see some of the game changes we all voted on?

It's still too early to see most of the brand new stuff, really. Clay mentioned that he's trying to finish the encyclopedia first, and since that's the largest work sized feature to be added, it'll go quicker once that's complete.

[NYDodgers]

First, this would address two concerns:

1. 99% of the time, there is no reason to sign anyone before free agency. There was a thread about this before, but I release every player every time at the end of the season, then deal with them on free agency. Why? Because it's a lot cheaper. There was one case when a player asked for 28 million before free agency, then asked for 18 on the market. I do not edit any revenue settings.

2. You can sign any free agent you want, regardless of who you are. Even beginners can have streaks of winning the world series with a big market team.

My first proposed solution is to have free agents base their salary on either the formula already in the game, OR the maximum amount of money any one team is willing to pay him -- whichever is higher at the time. For example, let's say the game says that Player A wants 10 million for 5 years. Right now, I can haggle and sign him for around 8.5 million with, let's say, the Brewers. However, if the Dodgers are willing to pay up to 12 million for him to fill a gap, he should most definitely go with them unless there's very, very compelling reason to come to Milwaukee.

Second, there should be NO bargaining on free agency. You should either accept his contract or offer him MORE money. This happens every single season in the game: Theres a 95/95 player on free agency and he's asking for 20 milion, and the rumor mill says "...is close to signing a 5 year contract with the New York Yankees." Then the human GM comes along and offers him 16 million and the player takes it. This makes no sense. The only time bargaining with a free agent might work is if there are no other teams interested in him and it's Feb-March. But for the many players that DO have teams negotiating with them, the player should not be allowed to bargain with him, and can only offer a larger contract.

This would also fix the problem of the human GM never having to sign anyone before free agency, since he can just release a player and sign him back for cheaper there. If there was a big demand for, let's say, 2nd basemen, in the league, then the human GM should sign him after the season and not test the market. However, if there was an abundance of talent in that position and the human GM feels that there would not be as much competition, then he can take a risk and simply release him and try to sign him for cheaper at free agency. This would add another fun aspect to the game.

I think the best part about my suggestions is that they would be easy to implement. You are not bringing something completely new like 40-man rosters or waiver wires, and would also not give any more burden or responsilbiites to the gamer. Clay doesn't need to figure out some way to implement anything new, he would only be modifying a few things.

This would essentially be a step toward free-agent bidding, and it would look great on the advertisements to say that bidding has been implemented into the game for BM08. Also, if there are any glaring flaws or a clarification is needed please let me know, but feedback in general would also be appreciated.

[ohms_law]

The two big problems with Free Agency are that the user gets first shot at all free agents, and that the players do not look out for their own interests at all.

You'll notice that there's no counter bidding system in the game. You enter a bid, and the player either accepts or not. And you can enter bids for as long as you like. There's a point where his asking price will start to increase, but even then you could simply "accept player demands" and the player is yours. Since you, as a human user, are able to act before the AI does, the end effect is that you can sign whomever you like without having to worry about underbidding or anything. So, one important aspect that needs to be added is a bid system, so that there's no instant signings (players wouldn't sign for a few days after getting an offer, so that other teams could bid on them).

The other problem is that players simply do not care what sort of team they are going to. The Yankees can generally sign whomever they wish as a free agent, simply because they have all the money. The players don't look at the team offering them contracts at all, so they just accept the largest AI bid on the table. This is the primary reason for "super teams" and "super losers", because the large market teams end up with the top 3 first basemen in the game all being on the same team.

So, what I've been advocating for months now is for Clay to develop an AI for each player. The player AI would decide whether or not to deal with a team at all, depending on how much playing time he will receive. This should be fairly easy to do, if the AI could generate "what if" auto-sorted lineups (which is what the AI teams use anyway), inserting himself into the lineup and seeing where he would end up. If it's a veteran player with MLB service time, he wouldn't be willing to talk to a team where he would end up in AAA. Older veterans might be willing to accept a bench role, but players in their 20's and early 30's probably wouldn't unless that that's all they could find. Prospects would simply want the highest position they could find. Things like that. It could be tweaked as time goes buy, but the important thing is just to get it into the game.

[Clay Dreslough]

Maybe I'm overvaluing such things as the waiver wire and rule 5. But I think all these things I'm asking for go hand-in-hand.

Players don't just start asking for more money or less money one day. They compare themselves to similar players in reference to revenue and their own personalities. Until a player can "think" and a manager has all the tools at his disposal a truly accurate mogul is impossible. Maybe I'm asking for too much but when two years pass and very little has changed on that side of the game questions are beckoned.

Gary Sheffield (for example) isn't just going to play anywhere. And he's not just going to sit by if he's traded to a mid/low market team. That's part of his personality. And that should be reflected in as many aspects of the game as possible, IMO.

Or if you've got a young gun like Willis tearing up the league before he's supposed to. He's going to expect the big bucks but, like baseball, there should be stop-gaps in place preventing that. Nothing is more frustrating than your top-prospect tearing up the league in his rookie season when you were just planning on testing the waters with him then in arbitration he requests millions. At that point and for the next two or three years his pay should never break 5 million dollars.

How about building into the players personality a system that rates their opinions of teams. I.E. A prospect from Houston would start out his career with certain prefernces. If he's a primadonna type then the NY and LA teams plus Houston and the team that drafts him should be his top teir selection. Then if a player has a good experience with a team (basically success and adequate playing time etc.) he'd be more inclined to resign and/or take a hometown discount (if he's not a primadonna). If he's not happy (losing seasons, inadequate playing time, personality conflicts) there's no way in **** he'd resign and that team could have a negative value as far as that player is concerned. I admit I'm no programming expert but I have experience and a personality model would be as simple as assigning a +1, -1 system.

A player's drafted by Tampa Bay so they get a +1. Well he's from Cali so there's a -1 for whatever reason. For every winning season that they contribute to regularly is a +X amount and for every person he get's along with on the team is a +.25...or whatever and so-on. The value's arent important and could be tweaked but the structure doesn't seem that complicated. I'm by no means familiar with anything as complicated as BBM and haven't the slightest clue how you'd impliment and/or intigrate something like that but it's not as complicated as say inflation models.

From this formula you get a basic, numeric value of how a player feels towards a city/players/etc. Apply basic stats (to figure out means and standard deviations and ranges and how players opinions compare to one another) and from there you could add a whole new aspect to the game that would greating improve the free agency market, IMO. Teir the player's opinions of the season and then you can even improve the rumor mill. If the Cubbies are looking for a second baseman and a certain 2nd baseman had the Cubs in their top teir they'd probably be talking.

Or if a two teams are offering a contract to a player, one in his top tier and on in his, say, third, it's not unreasonable for player to take a minor discount to go to his prefered team (again, unless personality (love of money, etc.) conflicts with that). And now that I'm double parenthe-sizing (not even a word) I'm going to walk away hoping I made some sort of sense.

I mean no presumptions and don't mean to come across unappreciative of BBM or its staff. But I figured it's easy to point out problems and much harder to suggest a solution.Most of what you mention here is already factored in. Players value where they grow up, teammates they have had, a team's ability to win, prospective playing time with a team, etc.

It's possible that the weight assigned to these factors isn't enough for it to be obvious when you play. The biggest problem is probably the _GM_ AI, not the player AI. That is, the other teams aren't holding onto valuable players or outbidding you (yes there is a bidding process that goes on at the START of the off-season to set each player's asking price).

Although you hear a lot about hometown discounts and where players "want" to play, the truth is that the vast majority go for the money.

Johnny Damon said he wanted to stay in Boston, but the Yankees offered a couple million more per year. I actually think Mogul does a good job of simulating Damon, and I tweaked the code a year ago with his situation in mind. He basically decided "I'll stay in Boston for $12M, or go to the Yankees for $14M" (I forgot the exact numbers). In other words, there WAS a hometown discount, but the Yankees bid enough to eliminate the hometown discount. Same with Pedro.

Correct me if I'm wrong, but I can't think of any prominent example where a player didn't jump at the biggest offer. Did A-Rod really want to play in Texas? Of course not. But he took the money.

My baseball life is very Boston-centric, but Bronson Arroyo is the only guy I can think of who gave a significant hometown discount. And he got burned for doing it.

And although sports-radio loves to talk about guys performing well in the contract year, or performing horribly when they are unhappy, there's not much evidence this happens. Manny Ramirez is supposedly depressed about being stuck in Boston, and the front office keeps trying to trade him. But he keeps showing up every day on Opening Day and posting Hall-Of-Fame offensive numbers year after year.

Sheffield might be a prima donna, but I'm guessing if the Pirates offered him $20 million he'd hop a plane to Pittsburgh -- and he'd probably put up the same on-base percentage that he would in New York.

[Clay Dreslough]

ANYWAY, I didn't really address the main point of this thread. And it deserves addressing.

The funny thing is that I was worried I was doing TOO LITTLE to improve Play-By-Play in this version. I was hoping to add more detailed PBP text, not just the one-liners you currently see.

But there were three areas we knew we had to work on:

1) Stat tracking. This was the #1 issue. Mogul's sim engine is great. You can play 100 years with a few clicks. But then you lose most of the data. It took a significant amount of work to get Mogul to keep all this data inside the saved game file, without

1a) I'm adding the 'Encyclopedia' as '1a' since it's related to 1. What's the point of playing for 50 seasons if you can't go back and look at what happened...

2) Historical finances. Maybe it didn't bug everyone, but it was just silly to have Babe Ruth asking for $27 million. IMHO, getting this stuff working right really changes the feel of historical sims.

3) Improving the interface. This may not seem like a high priority to most people, but when you opened BB2K7 and clicked 'Maximize', it looked like a kid's game with big buttons and giant fonts compared to any other sports game out there.

Plus the fact that Wal-Mart won't carry it if we don't have nice screen shots. :)

When you think about the stuff we added between 9.00 and 9.45, it's not like we had a whole year to work on the above list. It's actually a big achievement IMHO if we get the above into BB2K8 and get the bugs worked out before release.

The PBP changes are actually pretty small by comparison. We changed the stadium from a 'diagram' to an actual photo, and added the ball. But that was easy since the game already tracked the ball coordinates.

Clay

One more post. I'm really glad you guys care enough to pour out your thoughts in this thread so I'm doing the same.

I do have to disagree with the folks that say that if they want to play pitch-by-pitch, they'll go buy MLB 2007. I worked on Tony LaRussa and Microsoft Baseball and MLB: Slugfest. And I spent many hours tearing apart MVP and MLB and All-Star Baseball as part of my "competetive research". And none of the console games come close to being realistic, especially in terms of the batting and pitching experience.

As just one example, it's pretty well agreed that the average MLB pitcher can throw his curve ball for a STRIKE about 60% of the time. But if you pick up MVP or MLB, almost any pitcher can paint the corners with his breaking stuff. That's just silly, and it affects the whole game.

IMHO the best batter-pitcher mode I ever saw in an "arcade" baseball game was Tony La Russa III. Of course, that's because I programmed it. But basically I put in a lot of unpaid overtime because I wanted it to be good. And I didn't know a lot about baseball that I now know.

By contrast, when the folks at EA and Visual Concepts get together to build a baseball game, they don't decide to put their best programmers on the batter-pitcher mode. In some cases, they don't even have a serious baseball fan on staff. Just look at Triple Play where they used "BOB" as the abbreviation for "Bases On Balls".

So, I don't see Baseball Mogul as just a GM simulation. I see it as a baseball simulation for people that want:

1) realism
2) a game where strategy, not reflexes, decides the outcome

Because play-by-play has more areas that can be exploited, I understand that it might not feel as realistic as the sim. If you are consistently beating the computer in pitch-by-pitch, I recommend you crank up the 'Play-By-Play' difficulty level. If you are STILL kicking butt on Mogul level, let me know how and I'll tweak it.

As for the GM side, with the risk of oversimplification, I'll lump this thread into two complaints:

1) You want to see better implementation of real MLB rules. 40-man rosters, Rule 5 drafts, etc.

2) You want the game to be more challenging -- and you either think that #1 will help this and/or that better AI is needed.

In response to each:

1) I really thought I'd have time for some of these features this year, but the three things I listed in the previous post took longer than planned (I also hired the wrong team to work on Football Mogul, but that's a different discussion).

These 'technical' things are actually relatively minor features by themselves. The real issue is writing smart AI for each of the new rules. If I add a Rule 5 draft with bad AI, it won't make the game more challengine. It will just be another area that a good GM can outsmart the computer.

2) I am constantly working on the AI. I actually was in the middle of improving the award AI before I jumped into this discussion. Admittedly, it doesn't make the game any more challenging, but it is part of the overall 'AI' picture. Big changes, like the database overhaul, tend to happen BETWEEN versions (like for BB2K8 version 10.00 due out in March). Smaller changes, like AI, tend to occur in patches, like v10.10, v10.20 etc.

Anyway, I think the main concern is making the game seem more challenging, like the real-life job of a GM. This is always going to be tough since it relies on making the computer behave as intelligently as real-life players and GMs. And it's exacerbated by the lack of parity in baseball, which means that if you win a World Series in Baseball Mogul, you now have a team with lots of money that everyone wants to play for, and so it's easier to win another.

Some of the examples in threads like these are really useful to improving the AI (or deciding to add a feature like 40-man rosters). So please keep up the discussions, even going so far as to use screen shots. If I see that Zito isn't asking for enough money, it's good if I can see the whole picture. For example, maybe he's got great ratings but he only went 12-11. Also mentioning the difficulty level would help, and giving your thoughts on exactly how you think the game SHOULD be behaving.

Thanks again!

Clay

[ohms_law]

I agree with your take on everything addressing the original post, so I don't really have anything to add on the subject.

I'm going to talk about the player AI some though, since I've been after it all year (along with several other people, in one respect or another).

The problem doesn't really have anything to do with valuation. Baseball Mogul's GM AI, and the players AI in regards to valuation are excellent, as you've pointed out. The missing aspect, mainly for the players, but also somewhat for the GM's, has to do with forethought and planning.

You're correct, most players in real life will take the highest offer and not think twice about it. However, currently in Baseball Mogul teams with room in their budget will be bidders on the free agent market regardless of their need for players at that position. Sure, there's the Yankees example where they picked up A-Rod when Texas offered him (although, I'd point out that that was a trade, not a free agent signing. That makes a bit of a difference). However, they were able to move him (it could very well have been Jeter who was moved as well) to third, so they did have an at least marginal need that they filled. Currently in Mogul however, The large market teams absolutely dominate simply because they can be bidders for anyone and everyone.

The White Sox will have a 29 year old center fielder at peak in the 90's, and a 26 year old guy who's 78/99, and yet they'll still bid on some other center fielder rated in the 90's who hits the market. Since they have the cash to outbid everyone, they end up with someone on the roster, someone on the bench, and several aging guys in AAA. Then, since the large market teams suck up all the talent in this way, the Pittsburgh's and Florida's end up with nothing but scrubs.

My idea for a solution is that the GM's, but mostly the players, should be looking at the effect of their signing with a team. If the player will end up in AAA on the team making the offer, then they shouldn't be interested at all. If the player will replace the current player on the major league roster, but the two players are about equal (overall, every player is obviously different in details), why should the GM be bidding? What does the team gain, aside from payroll (which is actually a loss for the team...)?

In terms of bidding, we as users have an unbelievable power to overcome the AI on the free agent market, since we receive instant signings. The AI can't compete with a user's team, since they have no ability to offer counter bids during the user's bidding process. Simply changing the system so that players will consider a bid for at least a day before accepting would help significantly.


Because play-by-play has more areas that can be exploited, I understand that it might not feel as realistic as the sim. If you are consistently beating the computer in pitch-by-pitch, I recommend you crank up the 'Play-By-Play' difficulty level. If you are STILL kicking butt on Mogul level, let me know how and I'll tweak it.

Be sure to read that post by GD someone, here in suggestions... actually I'll find it and post a link.

It's GNR945, and here's the link:
http://forum.sportsmogul.com/showthread.php?t=134766

[NYDodgers]

The AI for each player sounds good, but I doubt it'll be easy to implement. There are many great suggestions on the forums, it's just that they would be hard to actually code into the game. I'm just being realistic and suggesting a modification of the original game without too many changes, while still having a strong, positive effect.

Even the individual AI thing wouldn't fix the problem where there's no point in signing players after the season and before the free agency. And it would also make it harder for a big market team to sign anyone since the "maximum a team is willing to pay" would spike up top players' salaries.

[ohms_law]

Here we go with this old bug again.
*sigh*

AI teams are constantly signing and releasing certain players off of the free agent market. I suspected as much with the "sim speed slow during the offseason" post, here's one example as proof:
7272

Can't find any examples in 10.03 yet, but I'm still looking.

There also seems to be a problem with one year contracts now. I don't really have a handle on it yet, but there's a bunch of journeymen who seem to be getting hired by the AI teams every year around March:
7291

October mercenaries seem to be popular as well. Isn't there a September signing moratorium? (Of course, Toronto wasn't a playoff contender, so this probably isn't really a bug):
7292

[dolfanar]

Well, signing a borderline/journeyman for a year isn't so unusual as long as the team doesn't do the sign & release tango as in versions past. That is, it's fine as long as it is at a position where the team has a need.

[nematode]

When computer teams are in debt their players are awarded multiyear deals by arbitration, but when a human player is in debt they have to release all players.
Can the Mogul folks explain this.
Its simply not fair to have different rules for the computer and the human players.

[bigpapi]

I know there was a thread about this, but I couldn't find it. I just simmed a couple of seasons in the early-mid 1900's (1930 and 1940) and each time there were about 10 total free agents when the season ended. I guess looking back then, there weren't any free agents anyway, so I'm not sure if this is working as intended, but I felt the need to bring it up.

[Wassit3]

good catch, maybe they were simply released by their teams maybe???

[xie]

I don't think this really needs any introduction:

7621

7622

Payroll is over $300M, team is bankrupt. This was during a "sim multiple seasons" run.

[Red Sox Fan 734]

360 million dollar payroll.....typical yankees:p

[Cowabunga]

This is going on in my game, too. Not just with the Yankees. Other teams, too. What's with that?

[ohms_law]

The problem is, right now, all of the AI teams are re-signing everyone regardless of their financial situation. There's a new setting in League Editor for Contract Renewals, and I think that that's probably messed up slightly right now.

[Cowabunga]

Is there anyway to counteract that by tweaking the contract renewals dropdown?

[ohms_law]

I'm not sure yet. Play around with it, and we'll find out...
:)

[Wassit3]

The problem is, right now, all of the AI teams are re-signing everyone regardless of their financial situation. There's a new setting in League Editor for Contract Renewals, and I think that that's probably messed up slightly right now.



On the transactions screen, press and hold the mouse button over any of the three buttons at the top. That will give you menues for each, so that you can select what you want to see.

mine seem to resign very few guys....

[ohms_law]

In what years?

[Wassit3]

In what years?

i'm playing in the 2020 to 2025 time frame...I always start with 2007 season

btw forgot to add I have the contract renews set to 40%, I'm about to go to 70% and see if that helps...once I saw a 23 year old pitcher rated 94 and only asking for 2 million per year for 6 years and no one wanted him, I signed him just before the trade deadline....

[mystupidmouth]

I simmed from 1901-50 and have been playing since 1950, and have noticed the same thing. It is actually kind of neat because I feel its more accurate, because like in the 50s I have to rely entirely on the draft and trades to improve my team. Whether that is the intended effect or not I don't know. I figure by the time I hit the 70s I'll fiddle with the renew contract feature and try to get free agency really going.

[ohms_law]

It's strange. At 0%, every team seems to resign everyone. with a negative value, it seems to create a lot of the sign and release behavior from the AI. Now, according to Wassit above, if you make it a positive value no one is signed at all.

Something is wrong with the contract renewals option, for sure.

[MFazio23]

Yeah, same thing happened to me with Atlanta.

EDIT: And the Angles pitching staff. I checked after they won the World Series:
http://img92.imageshack.us/img92/5866/laapitchingstaffgr2.jpg

[Cowabunga]

Yeah, Fazio. In my Historic Seasons, I counted about 55 players in my minor league system. It's QUITE crazy. You have to constantly release all of the dead weight and even that doesn't work most of the time. They just stockpile right back up.

[cody8200]

I played through 1982-86 today and have found there was only a handfuls of free agents during this time. Is this supposed to occur or is this a bug, I'm not sure. The handful of free agents there are are always horrible. Most are just washed up players. It ends up with me the players I just released being the only free agents available. Most of these guys are under a 70 overall but the computers just gobble them all up. Some teams will sign 5 pitchers under a 70 overall. Something seems wrong.

[JustinB0716]

The two big problems with Free Agency are that the user gets first shot at all free agents, and that the players do not look out for their own interests at all.

You'll notice that there's no counter bidding system in the game. You enter a bid, and the player either accepts or not. And you can enter bids for as long as you like. There's a point where his asking price will start to increase, but even then you could simply "accept player demands" and the player is yours. Since you, as a human user, are able to act before the AI does, the end effect is that you can sign whomever you like without having to worry about underbidding or anything. So, one important aspect that needs to be added is a bid system, so that there's no instant signings (players wouldn't sign for a few days after getting an offer, so that other teams could bid on them).

The other problem is that players simply do not care what sort of team they are going to. The Yankees can generally sign whomever they wish as a free agent, simply because they have all the money. The players don't look at the team offering them contracts at all, so they just accept the largest AI bid on the table. This is the primary reason for "super teams" and "super losers", because the large market teams end up with the top 3 first basemen in the game all being on the same team.

So, what I've been advocating for months now is for Clay to develop an AI for each player. The player AI would decide whether or not to deal with a team at all, depending on how much playing time he will receive. This should be fairly easy to do, if the AI could generate "what if" auto-sorted lineups (which is what the AI teams use anyway), inserting himself into the lineup and seeing where he would end up. If it's a veteran player with MLB service time, he wouldn't be willing to talk to a team where he would end up in AAA. Older veterans might be willing to accept a bench role, but players in their 20's and early 30's probably wouldn't unless that that's all they could find. Prospects would simply want the highest position they could find. Things like that. It could be tweaked as time goes buy, but the important thing is just to get it into the game.



That really would be a nice addition. It would be nice if there were some players who would not sign with a small-market/rebuilding team no matter what they offered, because the player had a high ambition rating and whatnot, or just wants to play for a contending team. My other pet peeve is a player who signs with my team and then it says how he "hates [CITY] and wants to be traded".

It would be nice if he "let me know" before I signed him. :p

[prairieball]

I think the simplest way would be for the free agent player to wait on making a decision (this is used in other games). The player waits several days for offers to come in and then decides whether to sign. Based on the stats of each team and player personality, players will have target prices...i.e. it may only take $14mil to sign with a playoff contender like the Yankees, while the Royals might have to ante up $16 mil.

At the end of the day, the free agent could state which offer it liked most (this should be a fairly straight-forward calculation based on the comparision of offers and by how much they exceed the player's demands for the given team). But instead of accepting, the player waits for counter offers in a certain time window (I would say 2-3 days). If the player doesn't get a better offer, then the player signs.

[dolfanar]

In terms of Player AI, a simple way to do it would be to compare the player against all other players at his position. If he is a top 30 (or however number of teams in the league) then he should NOT negotiate with any team with a equivalent or better rated player at that position. If he isn't a legit starter then all bets should be off. ditto for starting pitchers (Top 150) and relievers (top 180).DH's would be a tough variable to account for but not impossible.

[BTerran]

I definitely agree that there needs to be a bidding system. I think the issue will be less what a Player AI/Player wants and more about how high an AI team will bid on a player. Ideally the market will decide how much players get. If there's a lot of teams that have a lot of free payroll room but there's not a lot of talent (ala the last few FA markets) then players should be bid up to ridiculous amounts. If it's the opposite, then they'll be substantial bargain bin picks (which isn't far from what it's like now, if I wait until Feb. then what's left is very cheap).

But there does need to be something done so that we don't get an unchallenged pick of the litter. Even if, as a stopgap measure, when we make a final proposal to a player, any team can instantly outbid us and if the player likes their bid better, then we're offered the choice to let him go or outbid the AI Team, and then so forth until one of us gives up. That idea isn't my preference, I like the idea of day/week long fights over multiple players at the same time, but that would at least give the CPU teams a better chance.

[Yamato]

Yeah, same here too on the total resign by AI, noticed some FA,but not many, on V10.12, but only played one season of 10.12

[dolfanar]

It's strange. At 0%, every team seems to resign everyone. with a negative value, it seems to create a lot of the sign and release behavior from the AI. Now, according to Wassit above, if you make it a positive value no one is signed at all.

Something is wrong with the contract renewals option, for sure.


umm... why is that strange?

Positive Value Equal MORE likely to resign.
Negative means LESS likely.

What's the mystery?

Now the baseline may be off or not (0%), but that's another matter.

[HoustonGM]

The baseline of 0% is why it's strange. At 0%, everybody's resigned, which is strange. :p

[ohms_law]

Here's the other aspect to the Trade AI weaknesses issue. The AI is decidedly strange about who it decides to sign from free agency.

Here's an example. I started a game in 2007 and simulated one week with all computer controlled teams. On the very first day, NY goes and signs a Free Agent:
7620

According to commissioner mode, Mabeus is a 70/70 overall 28 year old pitcher. Giving the AI the benefit of the doubt, I set myself in control of NY and sure enough they have +/-7 scouting and think that Mabeus is actually an 80/80 pitcher, just hitting his peak. NY obviously has salary room, and you can never have enough pitching, so this sounds good right? However, look at their roster and specifically where Mabeus ends up at:
7619

So, how exactly is he helping the team again?
To me, both of these issues are really almost the same thing. The AI sees a player who is available to it as being good enough to qualify for some threshold, and so it goes ahead and acquires him. The trouble is, It doesn't seem to be evaluating what it's going to do with that player or players once it does acquire them. As is evident in the screen shot of the Yankees lineup above, they obviously have some talent available for when someone in the active bullpen is injured so really, even if Mulder was an 80/80 pitcher, their actually hurting themselves by acquiring him.

All of the above is from versions dating back to 9.13, at least. It is basically all still occurring in 10.13.

After reading all 7 threads that this thread used to be, cleaning them all up, and re-reading them all again, I decided to merge all of this together into this one thread. It blurs the focus slightly, but everything in this thread should be considered when thinking about the Free agency system. To recap:


To maintain some competative balance between the AI and user controlled team(s), there really should be a Free Agency bidding system (no instant signings)
GM AI needs to consider value of player against the value of players currently on the roster.
Whatever is causing the AI sign & release issue really needs to be fixed. If it is a money dumping strategy, then the teams should really be dumping money into Farms/Scouting/Medical instead of the FA market.


I'll also toss in that there also should be a separate system that turns off free agency in general prior to the mid 70's, and turns it on at the appropriate time. That's decidedly secondary to all of the above, however.

[dolfanar]

The "5-Year Rule" would also help this tremendously. Players with 5 years MLB experience cannot be sent down to the minors without their consent. In BBM until 40-Man rosters, proper DLs and waivers are implemented this should be an ABSOLUTE. It should be impossible to send a 5 year Major Leaguer to the minors, and if you need the space for another player you should have to release hima nd buy him out.

Advantages:
1) The minor leagues would be reserved for... Minor Leaguers! Imagine that?
2) Stockpiling All-Stars in AAA (and AAA (And A... yes I've seen it!))) would be impossible... even for the Yankees.
3) Assuming teams like the Yankees have to pay for buy-outs, if they want to upgrade at a position they already have covered... it is possible but EXTREMELY expensive... which will drain their resources.

Is this completely realistic? No... but I don't equate BBM with realism anyway. Too many factors are missing, overlooked and glossed over right now. This is a stop-gap that would tide us over until REAL MLB roster rules could be instituted (Which BTW don't involve just Waivers, 40 man rosters and DLs... let's not forget that you CAN NOT just send minor leaguers up and down willy nilly in real MLB... there are rules for that too).

As far as player AI. Again quick and dirty:

In terms of Player AI, a simple way to do it would be to compare the player against all other players at his position. If he is a top 30 (or however number of teams in the league) then he should NOT negotiate with any team with a equivalent or better rated player at that position. If he isn't a legit starter then all bets should be off. ditto for starting pitchers (Top 150) and relievers (top 180).DH's would be a tough variable to account for but not impossible.

And ofcourse before you can do ANY of this you need to:
http://forum.sportsmogul.com/showthread.php?t=136308

[moneyp]

Great thread. This is really the heart of the AI issue. It's gotten to the point where I feel the need to equalize the cities before each dynasty. Yes, the New Yorks and Los Angelii of the world should have some economic advantage, but the difference in the game is utterly ridiculous. I simmed from 1901 to 2007 a few times, and the New York Yankees/Giants/Mets won about 80% of the pennants they were contending for. That's way too large an advantage. The other thing was to look at the Hall of Famers and how many of them wound up playing for the New York teams. Or conspicuously absent Hall of Famers who, you find, were buried for a good portion of their career on one of the New York teams (in one sim, Christy Mathewson inexplicably went from perennial 20-game winner to middle relief the rest of his career).

Another thing that bugs me. Players sticking with the position that they're playing depending on the year you start the sim. Start in 1901 and Honus Wagner winds up a Hall of Fame right fielder rather than a shortstop. It doesn't affect rookies, but it's still annoying.

[HoustonGM]

The "5-Year Rule" would also help this tremendously. Players with 5 years MLB experience cannot be sent down to the minors without their consent. In BBM until 40-Man rosters, proper DLs and waivers are implemented this should be an ABSOLUTE. It should be impossible to send a 5 year Major Leaguer to the minors, and if you need the space for another player you should have to release hima nd buy him out.

Advantages:
1) The minor leagues would be reserved for... Minor Leaguers! Imagine that?
2) Stockpiling All-Stars in AAA (and AAA (And A... yes I've seen it!))) would be impossible... even for the Yankees.
3) Assuming teams like the Yankees have to pay for buy-outs, if they want to upgrade at a position they already have covered... it is possible but EXTREMELY expensive... which will drain their resources.

Is this completely realistic? No... but I don't equate BBM with realism anyway. Too many factors are missing, overlooked and glossed over right now. This is a stop-gap that would tide us over until REAL MLB roster rules could be instituted (Which BTW don't involve just Waivers, 40 man rosters and DLs... let's not forget that you CAN NOT just send minor leaguers up and down willy nilly in real MLB... there are rules for that too).

As far as player AI. Again quick and dirty:


And ofcourse before you can do ANY of this you need to:
http://forum.sportsmogul.com/showthread.php?t=136308

I agree, but...what about like a 35 year old guy thats washed up, but oculd serve as insurance? That's why if the 5 year rule was to be implemented, the players would need to have an AI. A 27 year old in his prime would refuse to be demoted. A guy whos best years are behind him would accept...

[dolfanar]

I agree, but...what about like a 35 year old guy thats washed up, but oculd serve as insurance? That's why if the 5 year rule was to be implemented, the players would need to have an AI. A 27 year old in his prime would refuse to be demoted. A guy whos best years are behind him would accept...

And once you have realistic roster rules in place you can start to make those distinctions, until then it's best not worry about the exceptions.

Put it this way, I would rather be forced to release a 35 YO journeyman and have him retire then to have Frank Thomas end up in AAA while int he prime of his career because he was dumb enough to sign with the Yankees.

[LetsGoYanks]

I don't know if anyone mentioned this but it seems free agents, or at least guys you're resigning, make contract demands based on your available cash.

I was playing around with the fantasy draft and made my super team that cost some 500million in payroll. In order to upkeep that, I edited my cash to the max. Well after season 1 I was down to 700mill in cash and guys were asking for 36mill a year. One arbitration case went to 41million! The next year I was in debt at the end of the season. So before Free Agency began, I edited my cash so I could resign players. That's when things got weird. Guys like Pujols and Santana were happy to take contracts for around 1million/yr for 7 years.

I played a few seasons and started to give short 2 year contracts as guys got older. And sure enough depending on whether I had cash or was in debt (or just low on cash) their demands would go from 30+ mill to 1 mill.

I'm not sure if that's a bug or not, but it was certainly odd.

[ohms_law]

I would really rather not see any "quick and dirty" solutions implemented with this. That sort of thing ends up doing more harm than good, in my opinion. We're much better off being patient and waiting for Clay and Ian to be able to implement a good system that takes all of this stuff into account. The patchwork fixes are part of what got us here to begin with, really.


I don't know if anyone mentioned this but it seems free agents, or at least guys you're resigning, make contract demands based on your available cash.
etc...

That's not a bug, it's working as intended.

[dolfanar]

Sorry Ohms. It's version 10.XX I've been playing these games since version -1 (Tony LArussa 3)

It may be version 20 before any comprehensive changes are made. Clay has not exactly stated that realistic and comprehensive roster rules are at the top of his list. We will NOT see 20 new roster rules implemented at once. It just AIN'T gonna happen.

The 5 year rule is NOT a patch work solution. It is a REAL MLB rule, that should be implemented, and would cut down on this GAME CRUSHING AI DEFICIENCY. What HGM is talking about is an EXCEPTION to the 5 year rule, which could always be added in later but is hardly a big deal. A washed up 35 YO in the BBM context probably SHOULD retire because there isn't full Minor LEague rosters anyway and sending him down to AAA means that a true prospect (who may be almost MLB ready himself) will probably be shuffled back to AA.

The farce of it is the talk of this being an AI issue.. it ISN'T, it's a question of implementing the same rules that MLB implemented to stop the very same behavior would be a problem in the real Majors if those rules were not in place. There are many BIG issues here, but the biggest is the signing of All Stars by big market teams and sending them to AAA or AA. The 5 year rule fixes that, so that should go in first. Once that is done we can see what the next biggest problem is (My guess is we will see the Yankees buy out a player every time a new superstar comes on the market, and play musical chairs... but atleast they won't be sending Sammy Sosa to AAA) and so on.

But, as I stated above, until we get a sepearate DL, almost NONE of these changes can be implemented.

[ohms_law]

First, I don't care if it takes until version 20 to get it correct, as long as it does get corrected.

Second, we don't need roster rule changes at all. I totally disagree that this has anything to do with roster rules, and this is purely an AI issue. Who cares about the DL, Rule 5 draft, 25/40 man rosters, etc... What difference do those make? The AI would still be signing players based on it's flawed decision making abilities. The fact is, that the current AI simply doesn't consider anything beyond the player that it's currently looking at. It's a design flaw, and all this talk about roster rules and whatnot is simply clouding the issue. If you work around the issue, adding things to "fix it" without actually touching the design flaw itself, then the flaw is still there.

No roster rules, no additional roster positions/areas/designations, nothing at all should be touched with the rosters, period. Not until the AI is otherwise behaving sensibly.

[dolfanar]

The AI is behaving EXTREMELY realistically based on the rules that it has to play by.

If the Yankees could buy every superstar in the league and guarantee a pennant each year... they would. The *reason* these rules are in place, are to prevent the very thing we see in BBM in the first place.

You *CAN'T* have realistic roster AI without reealistic roster rules. It's ridiculous. It's like programming a fishing game with realisitc Fish AI where the players can fish with Nukes... what's the point?

As for waiting another 10 versions... that's YOU. I for one will NOT wait that long.

[ohms_law]

The AI is behaving EXTREMELY realistically based on the rules that it has to play by.

How in the world do you figure that that's true? I understand your point about high revenue teams normally being able to purchase whoever they wish, but that's not the issue. Well, it is, but in a round about fashion...

Look, just get past the roster rule thing and actually watch what's occurring. The New York, Los Angeles, and Chicago teams are able to win constantly already because they can grab whoever they want off the free agent market, there's nothing wrong with that. The problem is, the AI uses a scatter-shot approach to these signings, simply deciding for each individual player that their good enough to meet some qualification to the AI. That is a fundamental design flaw, since it often doesn't actually help the team in any meaningful way. I'm not talking about roster rules, I'm referring to the fact that in any one game you can only have 9 players on the field at any point in time.

I'm perfectly fine with signings that actually improve the active roster. If the Yankees, Angels, Red Sox, etc... can sign a player off the open market and he would play every day (even if it's on the bench), that's a good thing. Even if their being signed in order to platoon with a different player, that would be fine as well. If signing a player ends up sending another player to AAA, that would actually be a good thing since it means that the player was signed in order to improve the team. Players do get old, their skills degrade, and they need to be replaced, after all. The problem is, the high revenue teams are signing guys to major league salaries in order to play in the minors! After being signed, they often never see a day of major league service time again, and end up retiring after a couple of years.

Even if there were no roster rules in real life, I just couldn't see Steinbrenner or anyone else signing players to major league contracts just to play them in AAA. It's completely pointless. Not only that but it ends up hurting their own prospects, as has been pointed out in the past. Roster rules would help the competitiveness of some teams, but their not going to help an already flawed AI decision making process. If anything, that would make the situation worse at this point.

The bigger issue with the lack of any kind of targeted approach to free agent signings is the small market team behavior. All of the AI's seem to operate in basically the same manner. They look at individual players, and if they meet qualifications (skill vs. cost) they bid on them until they get the player or the player does not meet the qualifications. The only thing that seems to make a difference to those qualifications is the teams revenue (and cash, to a small extent). Small market teams need to operate much differently, however. They should be bidding on much fewer players, but be willing to pay more for those particular players that meet their targeted needs. The way that it is now, they just end up with the left overs that no one else wants. With the system that is currently in place, it's no wonder that there are issues with super-winners and super-losers.

[dolfanar]

How is it pointless? If *I* sign the guy, he can't play against me. If I have THAT much money to burn that is.

You don't need an AI fix to stop this behavior as it is would be impossible to do under MLB rules. That;s like coding AI for the game not to use 11 starters on the field for each game. You don't NEED to code an AI to not use 4 OF... because it just doesn't happen, period.

The strictly AI approach has been tinkered for the better part of 10 years now... and it doesn't work. In fact it gets worse and worse the more complexity you add in with financials and minor leagues and whatnot.

What's WORSE is that if you try to code in a bunch of bogus AI operating in a vaccum (WHich has NOT worked for 10 years)... you will have to RE-DO that AI once you actually try to incorporate actual MLB roster rules.

[HoustonGM]

The thing is, the PLAYER ai is flawed. While it obviously makes sense for big market teams to gobble up players, the simple fact is that the players arent going to sign unless they're getting into a situation thats favorable for them -AAA for superstar...no playerwould sign that.

[LetsGoYanks]

it seems like something that could be fixed if teams were limited to a 25man roster. As in some players wouldn't accept an assignment to the minor leagues..

[dolfanar]

The thing is, the PLAYER ai is flawed. While it obviously makes sense for big market teams to gobble up players, the simple fact is that the players arent going to sign unless they're getting into a situation thats favorable for them -AAA for superstar...no playerwould sign that.

IRL? I would. Sure Georgey boy, I'll take that 25 million. There's my signature.

What, you want me to go to AAA? Nope. Play me, pay me or trade me Georgey!

The signing isn't the problem (Though the team ISN'T getting much of a return on its investmant). The player AI isn't the problem at all when you really think of it. He just wants to get paid baby! It's the lack of the proper roster rules that is the biggest issue.

Understand that up until I played OOTP this year I was totally on board with the "let's fix the AI" band-wagon, the thing is "AI" for WHAT? AI for a loosey goosey bastardization of MLB rules?

Fixing AI without implementing the real rules is pointless. I can deal with them being implemented one at a time, but lets thinka bout this.

I have been the biggest proponent of fixing the AI for Minor LEague releases for older Minor leaguers. Guess what? Minor League free agency deals with this to a great extent. The rest of it would be fixed by having full minor league rosters for some older minor league vets to sit on the bench or platoon with the youngsters.

I hate when teams send guy's to the DL for 1 day injuries. Real DL rules fix this.

I hate the way teams call up and send down players through auto-sort EVERY DAY. Minor league options deal with this.

I could go on, and on, and on.

I mean, ****, the choice beween OOTP and BBM used to be a slam dunk, but this year OOTP has made HUGE strides forward while BBM has actually managed to slide back (Financials are a mess, roster movement is a mess, and even the interface, the BEST PART of BBM is uglier and less inviting than ever).

[rj064cliq]

I really dont think that the game can incorporate much more realism other than the basic roster rules. Unless they allow you to start drawing up your own contracts how are they going to put in minor league agreements and trade clauses and all the other ****. the best bet would be to the 25/40 rule, real DL rules, and the 5 man rule. Then change the AI to where a player at a starters level will not sign if he doesnt have a chance to start. its already done on console games like NCAA Football, playing time matters to recruits. as for big market teams, look at the hoopla surrounding Dice-K. The reason he got so much from the red sox,almost twice the 2nd highest bid, was because they were afraid the yankees would get them. Big market teams will continue to buy players until there is a salary cap, if not to make the team better then just to make sure the other team doesnt have them.

[ohms_law]

Look, there are no 25/40 man rosters, and there likely won't be for the foreseeable future. According to what Clay say a while back, adding those restrictions to the rosters themselves would require an almost complete rewrite of the GM AI's, which isn't a feasible solution, let alone any sort of "quick and dirty" fix. All this talk about 25/40 man rosters is nothing more than a red herring, and it simply distracts from these discussions, with the end result being that we see less attention payed to the free agency issues that do exist. If 25/40 man rosters, disabled lists, rule 5 draft, etc... rules are added, that would be great but until they are talking about them doesn't do us any good.

This:

The thing is, the PLAYER ai is flawed. While it obviously makes sense for big market teams to gobble up players, the simple fact is that the players arent going to sign unless they're getting into a situation thats favorable for them -AAA for superstar...no playerwould sign that.
Is exactly the sort of thing that I've been advocating for several months now. The reason that I'm a proponent of adding such a system is that it works within the current AI to improve it, without requiring a massive rewrite of what does work.


What, you want me to go to AAA? Nope. Play me, pay me or trade me Georgey!
And therefore you wouldn't accept the contract.


The signing isn't the problem (Though the team ISN'T getting much of a return on its investmant). The player AI isn't the problem at all when you really think of it. He just wants to get paid baby! It's the lack of the proper roster rules that is the biggest issue.
You're right, the signing isn't the problem from the teams perspective, as I said above. Obviously, if a team like NY were in a position where they could, and were allowed to sign up the vast majority of free agents in order to keep them from Boston, Chicago, LA, etc... they probably would (and occasionally do) do that when they could get away with it. They wouldn't do it unless the player would significantly hurt them by being on a different roster however, so the frequency that they feel the need to attempt to keep a player away from competitors needs to be reduced regardless of anything else that's done.
The largest regulator of any such behavior though should be, by far, the players themselves. There is no way that Barry Bonds would accept a contract that allowed a team to send him to AAA right now. JD Drew probably wouldn't have gone to NY this off season for any amount of money in the world, since he would then be competing with Bobby Abreu and/or Melky Cabrera (although, his agent would certainly attempt to use any offer from New York against competing offers...).
Get past being stuck on the real life roster restrictions for a second and consider the logical behavior of both teams and players if the CBA allowed teams to send players up and down at will (as is true in Mogul). All players would obviously be willing to hear offers from any team in the league, and several of the rich teams would be willing to bid up the real quality players in order to keep them from competing teams. In the end though, the players would not necessarily take the absolute highest offer on the table unless there was a reasonable chance that the new guy would be the primary player on the teams roster. This is the key component that is missing from the free agency system, and it has nothing at all to do with roster rules.

[dolfanar]

And therefore you wouldn't accept the contract.

That's a ludicrous statement. If I had the 5 years in, I'd be nuts NOT to, since I'm protected against being "stock-piled" in the minors.

[ohms_law]

*sigh*
There is no such protection in Mogul, and there likely will not be for the foreseeable future. Are you really that stuck on roster rules that you're completely unable to imagine this?

[HoustonGM]

That's a ludicrous statement. If I had the 5 years in, I'd be nuts NOT to, since I'm protected against being "stock-piled" in the minors.
What if the teams puts him on the bench? He doesn't need to give his permission, and yet...if he signs the contract, he still won't be playing, and he won't like that. If a team has 3 star outfielders, a free agent star outfielder is not going to sign with that team.

[Clay Dreslough]

Look, there are no 25/40 man rosters, and there likely won't be for the foreseeable future. According to what Clay say a while back, adding those restrictions to the rosters themselves would require an almost complete rewrite of the GM AI's, which isn't a feasible solution, let alone any sort of "quick and dirty" fix."Rewrite" might be too drastic, but it's definitely not quick and dirty. As a "game" where you enjoy the challenge of building a team, Mogul's #1 AI priority is that CPU interactions with human players are reasonable, and not easily exploitable. The #2 priority is that CPU teams behave correctly in their dealings with players, and with other CPU teams.

In other words, many players don't care much about what the other teams are doing. So I'd hate to break the Trading Block (or create a hang bug) in an attempt to fix this. Which means we probably won't see any roster rule changes until next year, but I will tackle the AI a fair bit this year -- and if there are roster rules that seem low-risk, they may get added sooner (the 5/10 rule is on my list, but no promises -- things like the 40-man roster and Rule 5 draft will require significant changes).

This problem will take BOTH AI and roster-rule improvments to fix properly. AI changes are needed to keep teams from benching superstars. But the roster rules will prevent teams from burying them in the minors. And the GM AI nees as much or more work than the player AI. If I'm Nick Swisher and the Red Sox want to pay me $6M to be their 4th outfielder, while the A's are offering $5M to be a starter, I might take the Red Sox deal.

I've tweaked this some in 10.15 and I found a bug that was causing teams to bid on more than 1 free agent at the same position -- that might help. I'm testing right now by running 50 years but you guys can test yourself when I post the unofficial patch.

Clay

[ohms_law]

:cool:

Rest assured that we'll certainly beat up the system for you! (and keep arguing about it, almost regardless of what you do...)
:p

Seriously though primary the issues seem to fall into the "easily exploitable" category, at least the way that I look at them. Superficially, the AI does do a good job, but once you learn it's weaknesses it's extremely easy to get around (which really, is the bane of any AI programming...).

I understand where you're coming from with the Nick Swisher example(s) that you've been giving, but the problem is that Mogul doesn't have all of the roster rules in place. Even if it did, as I mentioned above, the impression that I've always received from stories, articles, and other analysis of the free agent system indicates to me that while agents and the players they represent are decidedly not shy about receiving and using the highest bid offers from any team coming, in the end the most important thing to them is usually playing time. From the players agent's point of view, this is understandable because they want to maximize the players potential as much as possible in order to make the most money off of him during the sort duration that they can. From the players point of view, well, most players just want to play. Obviously, getting the most money that they can is important to them, but if their going to end up sitting on the bench or in AAA I don't think that many would be willing to accept that. Both the player and the agent would feel that their investment is being destroyed.

Within the current system that Mogul utilizes, with out roster rules, I tend to think that the GM AI/auto sort does a good job with respect to benching or demoting "superstars". For the most part, when that happens in mogul it's due to the player being on the decline and a younger player being on the rise. From the teams perspective, demoting or benching the "superstar" player is usually an excellent idea since it furthers the kids' development and in the end earns the team more runs or run prevention and therefore more wins. Tread easy with that aspect of the AI, is my recommendation. The only reason that such things don't occur in reality is because for the most part teams can't do it.

From my personal perspective, getting the "most correct" behavior out of the AI now (as I've been incessantly talking about for the last two days here), before any roster additions are implemented, is absolutely vital regardless of what comes next. The roster rules limit teams and players behavior in real life, but they have come about due to the fact that without the rules some of the things we've been discussing here are smart strategies (depending on point of view). It's a good thing that the AI wants to do such things, in my mind, but would be limited by some rule that is in place in real life. Once the rules are implemented, it puts even greater pressure on the AI to actually do something about the problem once it is no longer limited by whatever the rule is. Basically, my view is that if the AI works well giving the current conditions that it's operating under, that should make adding more or different conditions to it later on easier. Make sense?

Just for the heck of it (and because I'm avoiding doing something else), here's what I envision the decision matrix should look like:

Team GM

Break current players down by position
Do I have a legitimate starter at this position? defined by:

find the league average (median?) of of all starting players ratings at this position league wide
Compare to the current player(s) on the roster at that position
Is my player better or worse?

Better: I'm not interested in players at that position. Significantly discount all players at this position.
Average: I'm only interested in players that are significantly above the average at this position.
Somewhat Worse: I'm interested in at least average players at this position.
Significantly worse: I'm interested in just about any player better than my current player at this position.


Find player that meets or exceeds my criteria on the Free agent market
Is the player worth perusing?


how much revenue have I been receiving?
what is my current payroll, and therefore how much can I add to it?
How old is he?
What salary is he asking for?
What duration of contract is he looking for? (less revenue should look for shorter contracts)

Yes: Make offer.
No: do not make offer, go to next player.



Currently, that should really be it for the team. the AI right now seems limited to most of what is in steps 3 & 4, which is where part of the problem comes from (avoiding the analysis in steps 1 & 2.

Player

Hear offer.
If offer is less than the previous (3, 5, 10?) offers that I've received, disregard it.
Do I want to play for this amount, or should I hold out for more?

How many offers have I received?
< (3, 5, 10 offers?) hold out for more.
> (3, 5, 10 offers?) evaluate offers.
Would I play full time on this team?

No:

Would I play on the bench?

yes: discount offer by (20, 25, 50%?) (these could be personality dependent)
No: discount offer by (75, 90%?)


Yes:

"Home" team? Increase offer slightly (10%?) Personality dependent.
Otherwise, consider offer at given value


Accept highest offer



And that would be about it for the players. Right now, I'm not sure that the AI really does any of that except for the step 3 "accept highest offer" portion. Even then, if it's a user's team making a bid then I've personally been able to underbid contracts occasionally...

It's a bit rough around the edges, but it's basically what I envision as being the most "correct" system. The main thing is, with the implementation above, that it wouldn't specifically worry about any rules that don't actually (currently) exist. Each portion of the AI would work towards it's own self interest. If 25/40 man rosters were to be implemented at some later date, that could then be dropped into the process at the time. Modularity is the key to the whole thing, really.

The other side of this decision tree (deciding whether or not to resign a player) needs some attention as well, but that can probably wait until later on. Generally, it suffers from some of the same problems. It is a different decision process though (or it should be), and teams should generally be somewhat more willing to want to keep a current player, while some players should want to be released and others should be more willing to accept their current team's offer. That should mostly depend on the other players going to the FA market, though.

Anyway, back to doing something productive for me...

[dolfanar]

What if the teams puts him on the bench? He doesn't need to give his permission, and yet...if he signs the contract, he still won't be playing, and he won't like that. If a team has 3 star outfielders, a free agent star outfielder is not going to sign with that team.

I agree with you 100%, BUT.. what is the bigger issue? Star players on the bench (which is totally possible, if unlikely in RL) or Star Players in AAA and AA? You are worrying about the rash on the victims left hand while he bleeds to death from his severed right hand. Sure the rash needs to be treated, but it's not the biggest problem.




In other words, many players don't care much about what the other teams are doing. So I'd hate to break the Trading Block (or create a hang bug) in an attempt to fix this. Which means we probably won't see any roster rule changes until next year, but I will tackle the AI a fair bit this year -- and if there are roster rules that seem low-risk, they may get added sooner (the 5/10 rule is on my list, but no promises -- things like the 40-man roster and Rule 5 draft will require significant changes).

This problem will take BOTH AI and roster-rule improvments to fix properly. AI changes are needed to keep teams from benching superstars. But the roster rules will prevent teams from burying them in the minors. And the GM AI nees as much or more work than the player AI. If I'm Nick Swisher and the Red Sox want to pay me $6M to be their 4th outfielder, while the A's are offering $5M to be a starter, I might take the Red Sox deal.

See THIS is what I'm talking about. I know FULL well that wholesale changes where 90 differant rules/changes are made at once isn't your style. The 5 year rule (THe 10 partof that would be nice too ofcourse) would totally plug the most obvious hole in the system right now. A seperate DL (No special rules additions... just a place to put players on the DL that isn't the minors) is another minor change that will be a welcome addition too. Finally Minor LEague Free Agency is another high impact/low risk feature to include. Just those 3 would be a huge improvement.

[HoustonGM]

I agree with you 100%, BUT.. what is the bigger issue? Star players on the bench (which is totally possible, if unlikely in RL) or Star Players in AAA and AA? You are worrying about the rash on the victims left hand while he bleeds to death from his severed right hand. Sure the rash needs to be treated, but it's not the biggest problem.
While it's possible, I can't think of any STAR free agent that signed with a team knowing he'll be a reserve.

I'd rather see both issues be taken care of. With one, goes the other, it seems. If a star player gets an offer from a team that already has a starter at his position, he won't sign. That prevents him from going to AAA or the bench.


See THIS is what I'm talking about. I know FULL well that wholesale changes where 90 differant rules/changes are made at once isn't your style. The 5 year rule (THe 10 partof that would be nice too ofcourse) would totally plug the most obvious hole in the system right now. A seperate DL (No special rules additions... just a place to put players on the DL that isn't the minors) is another minor change that will be a welcome addition too. Finally Minor LEague Free Agency is another high impact/low risk feature to include. Just those 3 would be a huge improvement.
Minor league free agency is a must. 6 years in the minors, a player should be allowed to file for free agency. I'm sick of seeing 32 year oldplayers rated 80 that have bene with an organization for 12 years and never got a taste of the majors...

[LetsGoYanks]

Minor League free agency would be cool, it'd add another dimension to building a dynasty. Either bring up the young guys and get rid of older stars or trade away to replenish the farm.

I don't think it'd be that hard to make, after X years in minors, player can file for free agency. But it'd take some AI work to stop teams from losing all their prospects.

[dolfanar]

I'd rather see both issues be taken care of.

Do you think I would object to that myself? :p

I just know that asking for an all-in-one solution generally doesn't happen with BBM. If it does I would be pleasantly surprised, but just from a testing stand point, it is much easier to get the absolute rule fix in, and then worry about the more fuzzy logic issues.

BTW, are you seing those 32 year old 83/83 releievers on top teams that I'm seing while bottom teams spend 2 million for washed up 28 year old 73/73 guys?

[HoustonGM]

Basically...I tihnk both issues HAVE to be solved at once, really, like I expalined right below the part you quoted...

[dolfanar]

Basically...I tihnk both issues HAVE to be solved at once, really, like I expalined right below the part you quoted...

They're too related issues, but sending 5 year bvets to the minors is NOT strictly a Free Agency AI issue. Your not looking at this from the true perspective of strictly a general roster management issue. You can fix the roster management issue (Don't send 5 year vets to minors, and if you don't have room release/buy out said player) without touching Free Agent AI.

The result ofcourse will be teams like the Yankees potentially signing a guy and immediately releasing someone else (MAybe even sign and release), but atleast they won't be hogging ALL of the FAs, and some competitive balance will be injected. Is it ideal? No. Remember I was the one who suggested this:



In terms of Player AI, a simple way to do it would be to compare the player against all other players at his position. If he is a top 30 (or however number of teams in the league) then he should NOT negotiate with any team with a equivalent or better rated player at that position. If he isn't a legit starter then all bets should be off. ditto for starting pitchers (Top 150) and relievers (top 180).DH's would be a tough variable to account for but not impossible.


Which doesn't require huge computations like giving EVERY player in the league his own individual AI and/or bargaining logic (WHich would be ideal... but not exactly processing time friendly).

Though to do this you do still need this:

http://forum.sportsmogul.com...d.php?t=136308

which is really a critical factor. It's so simple but, but as long as you need to send people to the minors to put them on the DL, you can't really do much more than Minor League Free Agency.

[ohms_law]

Minor league free agency won't work, since you don't need minor leaguer's at all. Sure, the AI can be trained to use them, but why would most human controlled teams do so? It'd just be a waste of money, usually.

Arbitration should take care of this aspect of the problems 90% of the time anyway. If the AI were generally more aggressive with arbitration then there wouldn't be as much of a problem.

Anyway, 10.16 seems to be an improvement overall. It's still suffering from (apparently) not really planning out it's roster at all, though. Here's one example of a situation:

San Diego has the following players at second base:


Team Player Name P Roster Overall Peak Scout Salary Age Con Pow Spd Eye Arm Rng Fld End Con Pow Mov Years MLB Service
SDP Giles, Marcus 2B 2B 82 82 8.8 $5,800,000 29 78 74 75 79 85 84 77 -- -- -- -- 4 6 years, 166 days
SDP Walker, Todd 2B (AAA) 80 80 8.5 $2,400,000 34 82 69 67 87 79 68 73 -- -- -- -- 3 10 years, 169 days
SDP Ryan, Jimmy 2B (AAA) 63 69 6.4 $30,800 21 67 69 74 66 79 55 76 -- -- -- -- Arb '10 None

November 5th, they let Marcus Giles go to free agency, and sign Todd Walker (in arbitration, I think) to his $2.4 million contract. Then, on November 5th after free agency opens up, they end up resigning Marcus Giles. I'm speculating here that the AI beleaves that it doesn't have enough second baseman. They put Walker back in AAA and are playing Giles full time again for 2008. Their roster after the first day of free agency:


Team Player Name P Roster Overall Peak Scout Salary Age Con Pow Spd Eye Arm Rng Fld End Con Pow Mov Years MLB Service
SDP Bard, Josh C C 86 86 9.3 $390,000 29 86 83 58 78 84 80 71 -- -- -- -- 3 4 years, 61 days
SDP Gonzalez, Adrian 1B 1B 86 86 9.3 $390,000 25 92 88 52 75 77 75 86 -- -- -- -- 2 2 years, 100 days
SDP Giles, Marcus 2B 2B 82 82 8.8 $5,800,000 29 78 74 75 79 85 84 77 -- -- -- -- 4 6 years, 166 days
SDP Hill, Bobby 3B 3B 74 74 7.6 $660,000 29 74 70 62 83 65 55 80 -- -- -- -- 3 3 years, 130 days
SDP Greene, Khalil SS SS 83 83 9.0 $390,000 28 73 84 66 74 79 75 86 -- -- -- -- 3 4 years, 70 days
SDP Alou, Moises LF LF 83 83 9.0 $7,300,000 41 90 92 57 77 61 46 67 -- -- -- -- 2 15 years, 113 days
SDP Cruz, Jose CF CF 77 77 8.1 $2,200,000 33 69 76 77 90 85 79 82 -- -- -- -- 3 11 years, 0 days
SDP McAnulty, Paul RF RF 80 85 8.9 $390,000 26 81 82 72 80 75 75 80 -- -- -- -- 1 1 year, 102 days
SDP Bowen, Rob C Bench 76 81 8.3 $390,000 26 69 76 56 76 71 81 69 -- -- -- -- 1 2 years, 122 days
SDP Blum, Geoff SS Bench 79 79 8.4 $1,950,000 34 75 72 64 68 83 85 73 -- -- -- -- 3 8 years, 150 days
SDP Giles, Brian RF Bench 73 73 7.5 $10,000,000 36 74 71 67 95 76 63 72 -- -- -- -- 2 12 years, 106 days
SDP Cust, Jack LF Bench 74 74 7.6 $500,000 28 68 81 51 95 62 61 70 -- -- -- -- 3 2 years, 10 days
SDP Kouzmanoff, Kevin 3B Bench 73 83 8.2 $480,000 26 80 77 65 70 64 55 73 -- -- -- -- 2 166 days
SDP Sledge, Terrmel LF Bench 72 72 7.3 $390,000 30 73 82 79 80 70 70 80 -- -- -- -- 2 3 years, 64 days
SDP Peavy, Jake SP Starter #1 93 94 9.8 $4,583,000 26 -- -- -- -- -- -- -- 93 87 97 71 2 6 years, 0 days
SDP Young, Chris SP Starter #2 93 93 9.8 $550,000 28 -- -- -- -- -- -- -- 83 78 94 88 1 3 years, 88 days
SDP Hensley, Clay SP Starter #3 85 85 9.3 $390,000 28 -- -- -- -- -- -- -- 82 79 81 76 1 3 years, 0 days
SDP Thompson, Mike SP Starter #4 75 75 7.8 $390,000 27 -- -- -- -- -- -- -- 79 84 66 80 Arb '09 1 year, 22 days
SDP Estes, Shawn SP Starter #5 76 76 7.9 $2,500,000 34 -- -- -- -- -- -- -- 84 79 72 75 2 13 years, 0 days
SDP Linebrink, Scott RP Closer 84 84 9.1 $1,557,000 31 -- -- -- -- -- -- -- 22 93 94 85 1 7 years, 132 days
SDP Meredith, Cla RP Setup 88 91 9.6 $390,000 24 -- -- -- -- -- -- -- 18 100 81 100 1 2 years, 100 days
SDP Ring, Royce RP Short 77 86 8.7 $390,000 26 -- -- -- -- -- -- -- 24 71 89 78 1 1 year, 135 days
SDP Brown, Andrew RP Short 77 82 8.4 $390,000 26 -- -- -- -- -- -- -- 33 65 84 81 Arb '09 1 year, 19 days
SDP Bell, Heath RP Middle 75 75 7.8 $390,000 30 -- -- -- -- -- -- -- 26 90 98 47 1 3 years, 107 days
SDP Rakers, Aaron RP Long 75 75 7.8 $390,000 30 -- -- -- -- -- -- -- 25 87 97 58 1 2 years, 33 days
SDP Allen, Luke RF (AAA) 69 69 6.9 $327,000 29 71 74 64 70 74 67 75 -- -- -- -- 1 90 days
SDP Tessier, Dale C (AAA) 64 82 7.5 $30,800 18 67 60 67 71 83 61 72 -- -- -- -- Arb '10 None
SDP Myrow, Brian 1B (AAA) 71 71 7.2 $510,000 31 69 77 74 79 72 65 74 -- -- -- -- 2 1 year, 17 days
SDP Walker, Todd 2B (AAA) 80 80 8.5 $2,400,000 34 82 69 67 87 79 68 73 -- -- -- -- 3 10 years, 169 days
SDP Ryan, Jimmy 2B (AAA) 63 69 6.4 $30,800 21 67 69 74 66 79 55 76 -- -- -- -- Arb '10 None
SDP Ranish, Cory CF (AAA) 65 87 7.9 $30,800 18 62 67 84 69 64 65 69 -- -- -- -- Arb '10 None
SDP Deago, Roger SP (AAA) 75 75 7.8 $327,000 30 -- -- -- -- -- -- -- 63 78 79 70 1 61 days
SDP Cassidy, Scott RP (AAA) 73 73 7.5 $327,000 32 -- -- -- -- -- -- -- 24 74 97 55 2 3 years, 1 day
SDP Anderson, Jason RP (AAA) 74 74 7.6 $327,000 28 -- -- -- -- -- -- -- 23 73 79 67 1 2 years, 68 days
SDP Adams, Mike RP (AAA) 72 72 7.3 $327,000 29 -- -- -- -- -- -- -- 22 78 84 61 1 2 years, 98 days
SDP Strickland, Scott RP (AAA) 71 71 7.2 $327,000 31 -- -- -- -- -- -- -- 22 79 82 50 5 5 years, 120 days
SDP Kent, Steve RP (AAA) 70 70 7.0 $390,000 29 -- -- -- -- -- -- -- 30 62 69 67 Arb '09 1 year, 0 days
SDP Brooks, Frank SP (AAA) 69 69 6.9 $327,000 29 -- -- -- -- -- -- -- 52 72 89 60 1 1 year, 86 days
SDP Johnston, Mike RP (AAA) 65 65 6.3 $327,000 28 -- -- -- -- -- -- -- 27 71 85 50 1 1 year, 120 days
SDP Stauffer, Tim SP (AA) 70 70 7.0 $327,000 25 -- -- -- -- -- -- -- 81 82 73 57 1 1 year, 119 days
SDP Hampson, Justin SP (AA) 71 77 7.6 $20,500 27 -- -- -- -- -- -- -- 56 77 82 58 Arb '10 60 days
SDP Sanchez, Felix RP (AA) 70 70 7.0 $390,000 26 -- -- -- -- -- -- -- 30 75 80 58 1 60 days
SDP Boone, Rob SP (R) 62 95 8.3 $12,300 17 -- -- -- -- -- -- -- 86 95 71 47 Arb '10 3 days
SDP Willard, Kevin RP (R) 56 82 6.9 $12,300 20 -- -- -- -- -- -- -- 22 61 73 48 Arb '10 None
SDP Sinclair, Simon RP (R) 55 84 6.9 $12,300 21 -- -- -- -- -- -- -- 19 72 74 47 Arb '10 None
SDP Cameron, Mike CF (DL) 84 84 9.1 $6,300,000 34 72 80 91 76 83 81 72 -- -- -- -- 2 12 years, 34 days
SDP Hoffman, Trevor RP (DL) 88 88 9.5 $6,750,000 40 -- -- -- -- -- -- -- 18 96 89 98 1 14 years, 123 days


The initial decision to release Giles and resign Walker was very good in my opinion, but then it was all undone when they resigned Giles. If the AI is worried about having enough players on it's roster (which is legitimate, overall the rosters are smaller which isn't a bad thing at all) they should be looking at either longer term players (lower level prospects), or players at positions where they really need help, such as third and center in this example (although, Jose Cruz is really much better than his Overall would indicate).

Mike Lamb (81/81) is still available after the first day of free agency, and is asking for less than Giles was ($4.5 Mil)...

Then SD goes and signs Tony Gaffanino (35YO, 2B, 82/82) to a $3.5 Million contract on November 23rd, benching Giles...
:confused:
...Now I'm suspicious that the AI is deciding what it wants/needs on the first day of free agency, and it's not adjusting that afterwards.

[HoustonGM]

Minor league free agency won't work, since you don't need minor leaguer's at all. Sure, the AI can be trained to use them, but why would most human controlled teams do so? It'd just be a waste of money, usually.
It's not that we want minor leaguers, it's that we might be able to use one tema's minor leaguer on our major league team. An 83-rated reliever is capable of performing in a major league bullpen, but he sits in a team's AAA team for 12 years and by that time, is on the decline.


Arbitration should take care of this aspect of the problems 90% of the time anyway. If the AI were generally more aggressive with arbitration then there wouldn't be as much of a problem.
No, because minor leaguers never go to arbitration. You need to have 3 years of MLB service time to go to arbitration....

See what I'm saying?

[ohms_law]

No, because minor leaguers never go to arbitration. You need to have 3 years of MLB service time to go to arbitration....

That doesn't appear to be true any more, as of 10.16... I'm still evaluating it though. And, that was my point. With an AI willing to negotiate most of it's arbitration cases, the problem mostly goes away.

[HoustonGM]

That doesn't appear to be true any more, as of 10.16... I'm still evaluating it though. And, that was my point. With an AI willing to negotiate most of it's arbitration cases, the problem mostly goes away.
Hm...O_o That doesn't make much sense anywya if people with no MLB service time go to arbitration....

[ohms_law]

Well, most of them have at least some MLB service time. I've been noticing that they tend to end up with at least 100 days by the time they come up for arbitration. A lot depends on the team that their on though.

Maybe I'm misunderstanding what you guys are saying though. How exactly do you define Minor League Free Agency for Mogul?

[HoustonGM]

It could be, any player that has yet to lose the "rookie" status (ie. not enough at bats/innings, or service time) can become a free agent after 6 years

Also, there's this problem. Say a player is primarily a bench player that shuttles from the minors to the majors. He accumulates 3 years of service time, goes into arbitration. Next year comes, he's above 3 years of service time, but not yet 4, arbitration. Next year, hes at 4 years of service time, arbitration...Next year, 4.5, arbitration. Next year, 5..arbitration...Nobody goes to arbitration that much.

Any player that has been with an organization for 6 years should be allowed to opt for free agency, like in real life.

[Shogey13]

I agree...

I am actually guilty of hoarding all of the AAA talent, too. None of my rookies go up for free agency or arbitration, so I have a huge stock of guys down in the minors who just hang around down there because they haven't developed enough and I throw a lot of cash around for stud players.

I guess I could wheel and deal a bunch of them out for big names, but I find it difficult to justify trades to some of these teams.

Come to think of it, I haven't noticed any teams backing up the truck when they are out of the race and trading for prospects. Is there anything in the AI that would lead a team to unload expensive players for prospects toward the trade deadline?

-Beau

[ohms_law]

Come to think of it, I haven't noticed any teams backing up the truck when they are out of the race and trading for prospects. Is there anything in the AI that would lead a team to unload expensive players for prospects toward the trade deadline?

No, I don't think that there is. That's a good point in my opinion, too.

[dolfanar]

What about teams in debt? I know they won't actually resign players, but do they ever try to dump players?

[ohms_law]

...I don't think so. The trade AI seems somewhat unwilling to make any deals right now. One of us should bring that up in the trade AI weaknesses thread.

[snapper]

I started a new game as the NY Yankees, 1989, at Manager level.

Everything seems to play fine except, no one goes Free Agent.

The only guys available in the FA period are real scrubs, or young undrafted players.

Any ideas?

Thanks

[snapper]

I'm playing as the Yankees starting in 1989, and virtually no one reaches Free Agency. Only a few scrubs.

Is this a known issues, or am I doing something wrong?

Many Thanks.

[ohms_law]

It's a known issue. See above.

[TylorTyler]

I noticed this too.

[jcbarr]

Yeah the problem is that all the teams are making a ton of money and they can afford to pay all the players what they want.

Keep playing, and when all the ticket prices and the salaries go up there will start to be some good free agents.

Watch out though cause it will shoot up pretty quickly when you get to around the 2000 season, at least it did for me.

[bunskins]

Since it has been nearly a week since somebody bumped this up I will. I really hope this is being worked on. These issues, especially nobody in free agency, make playing more than one year in the game pointless. All I need is some reassurance this is being handled. Hey, but on a positive note I can't wait to play once this is corrected.

[Clay Dreslough]

Hey guys,

Sorry I haven't been on the boards much. Fighting a sinus infection that's really wiped me out. I've made some significant improvements to the trade AI, and I'm now tackling this issue with every player getting re-signed before about the year 2000.

Hoping to have an unofficial patch up tonight or tomorrow.

Clay

[ohms_law]

:cool:

One word for ya: Benadryl.
;)

[jcbarr]

Hmm...I downloaded 10.19 and everything looked good for a while. I simmed three years and then I got to year 2022 and all **** broke loose. My best pitcher asked for 900k after winning 5 straight Cy Young awards.

Just about every player that was a free agent that year was resigned by his current team and every single player in the free agent market is asking for exactly the same contract.

Every player that I resigned asked for way too little money.

[ohms_law]

Roll-over bug...

[jcbarr]

Yeah, I guess I didn't read his post entirely and just assumed that this had been worked on in this fix.

Only thing is that In 2022 things actually regressed for me. It had gotten better with the lack of free agency and things, but then when 2022 hit, it went back.

Anyone else seen the game move in cycles like this?

[ohms_law]

What's happening is that the inflation index and the players salaries are probably getting out of control. Check the inflation index before and after the problem, and I'm willing to bet that it's going from being incredibly high to being incredibly low. That, and the salaries themselves, are roll-over bugs. What happens is that the value in the variable becomes too large to actually store in the amount of memory dedicated to the variable. So, the value rolls over to 0 and starts increasing again from that point.

[jcbarr]

Okay sounds good to me. As long as I know what I'm up against it doesn't bother me.

[ohms_law]

You know, this is still a good thread (especially pages 4 & 5). Most of these issues still exist, although they've changed slightly.

I still say the #1 issue is simply team awareness for the AI, especially after gaining access to the actual code. Players still accept deals to teams that are stacked at the position that the player plays, which is a huge problem. The team AI still trades away it's starters, sometimes for (good) players that they simply don't need and can't use. And then of course theres still the issue of quantity for quality trades.

@ Clay: I have a one word recommendation for you; recursion. The tools are all there, the AI simply needs to use all of them. The AI needs to be able to take the player out of it's lineup (it could just lock him in the minors...) and use Auto-Sort in order to ensure that it has a decent backup for the player(s) that it's dealing away. Adding some ability to... ghost? the proposed players that it'll receive would be extremely helpful as well.

Free Agent players need to look at the roster of teams making proposals to them, too. If there are 2 comparable players already there playing that players position, then the player should jack up it's asking price significantly (ie.: +200%).

The AI should start discounting the total valuation of a deal when more than 3 players are added to one side of a deal. The sum of value sent against the sum of value received is just way to simplistic. Even simply capping the total players that the AI will consider or offer at three per deal, maximum, would help.

Also, I wanted to bring up the idea of changing valuation based on team type again. Small market teams should value players differently than large market teams. Mid market teams can continue to use player valuation exactly as it currently is, I think. Small market teams should add "artificial" value to younger players though (players who are still arbitration eligable), while large market teams should add "artificial" value to older players. The opposite should be slightly true as well. So, a Twins or Marlins AI would place an ~33% premium on a Hanley Ramirez while he's Arbitration eligable (thereby pretty much ensuring that they'll resign all of their Arb players), while discounting anyone who's not arbitration eligable by ~10%. The New York, Chicago, and LA type teams could do the exact opposite.

[HoustonGM]

The AI should start discounting the total valuation of a deal when more than 3 players are added to one side of a deal. The sum of value sent against the sum of value received is just way to simplistic. Even simply capping the total players that the AI will consider or offer at three per deal, maximum, would help.
All great points, and I agree, except with this. There shouldn't be such a hard maximum, as trades with 4+ players on either side are relatively common in real life...

The AI just really needs to look for quality, though, rather than quantity, and have some concept of "top prospects" and what not.

[gosensgo101]

I've always thought that the AI could be impoved by leaps and bounds if the AI could do two things:

1. Evaluate their needs
2. Value quality over quantity.

If the AI could do these things, it would be great, because these are the two biggest things that human players do but the AI doesnt.

[Clay Dreslough]

The AI just really needs to look for quality, though, rather than quantity, and have some concept of "top prospects" and what not.The AI definitely is programmed to value quality over quantity. The only way to pry a superstar player away from a team is to offer another superstar, or a future superstar.

The valuation is exponential, meaning that one 95-rated player is worth *many* 85-rated players. But it also has diminishing returns built in, as suggested above, so you shouldn't be able to trade several average players for a star.

That said, I have been able to make trades like that on occasion, so I do need to find the bugs and fix them. But the suggestions about quality over quantity are already part of the basic design.

Clay

[Clay Dreslough]

I've always thought that the AI could be impoved by leaps and bounds if the AI could do two things:

1. Evaluate their needs
2. Value quality over quantity.Hmm. Those are the exact steps the AI follows. The first step before any trade valuation is a function called 'Evaluate_Team_Needs()'. So, we're on the same page. I just need to improve the way it works. I might be assigning too much "quality" to young players, so that it's possible to trade a few young players for a star, even if the prospects aren't projected to turn out very well.

[Clay Dreslough]

Also, I wanted to bring up the idea of changing valuation based on team type again.The AI changes valuation based on team needs, but also according to "team type" in at least two ways that I can think of:

1) Payroll budget. Unless I'm misunderstanding, I think this is the exact same as the "team type" factor you are discussing. A team with a high payroll budget will be more willing to trade for or sign expensive players.

2) Competitiveness. A team that is in the playoff race will be more willing to sign expensive veterans, and/or trade prospects to do so.

I think you need both of those together to be realistic. For example, the Marlins are currently considering trading for Manny Ramirez. If you just looked at "team type", then the Marlins would be the last team in the league that would trade for an aging $20M hitter.

[Clay Dreslough]

I'm sorry if I sound like I'm in denial, saying "thanks for the suggestions, but the AI already *does* that." This is a great thread for helping to find problems in the AI and get them fixed. But I also don't think it's a matter of basic design. As lifelong baseball fans we all have some good instincts about what teams need to consider when making trades, and those concepts are already in the code. Either the weighting is off (e.g. teams putting more emphasis on "competitiveness" than on "team type"), or there are outright bugs (like overflow errors with too many players in a trade).

I think the best way to proceed would be to find some mutually agreed upon bugs in the AI. For example, if you can start a game as the Red Sox using HoustonGM's rosters, then trade three crappy bench players for Hanley Ramirez on opening day, that's a bug. If this is easy for me to reproduce -- either by starting a new game, or loading a .mog file that you post -- then I can step through the code in the debugger and find out what the exact problem is. If we can create a set of these "bad trades" that everyone agrees upon, that's the best way to fix the AI.

Clay

[ohms_law]

But I also don't think it's a matter of basic design.
I don't either, which is why I haven't replied to this thread in a long time... and I seriously considered not replying to it again last night precisely because I now know that it's not a basic design issue. It's just... all the stuff that I've said in this thread in the past is still basically true!

So, lets see...

I think the best way to proceed would be to find some mutually agreed upon bugs in the AI.
I agree 100%. If I see 50 new posts in Bug reports over the next couple of days I'll be one happy camper.
:)


The AI definitely is programmed to value quality over quantity. The only way to pry a superstar player away from a team is to offer another superstar, or a future superstar.
I think that the highlighted point is the primary issue, currently. AI teams seem to be perfectly willing to both mortgage their future and give up the present based on the value of one or two players in a deal. Every experienced player knows that they can exploit the AI by using a 90's rated player (either current overall or especially young peak value players).
Quite often the question is "why would the AI make that deal?" The answer is obvious, that the AI sees the value of the players received. The problem is that it doesn't really see the impact of the move on it's current situation. As several of us have been saying here, the AI really doesn't ever consider it's overall team.


The AI changes valuation based on team needs
I know, but... something is obviously just plain off about how this is currently being done.
Maybe, and this is purely speculation at this point, but maybe the Payroll Budget usage is causing problems. I've been meaning to send an email to you for days about one possible issue in particular, which I just sent.
By the way, this:

2) Competitiveness. A team that is in the playoff race will be more willing to sign expensive veterans, and/or trade prospects to do so. Should probably be turned off, or adjusted in some why, so that it's only primarily used from about mid-May until the end of July.


The first step before any trade valuation is a function called 'Evaluate_Team_Needs()'.
yea... Evaluate_Team_Needs() is exactly what I've been harping about that needs so much work though.

3b. UNLESS rebuilding, don't trade a franchise player:
and
3c. When winning NOW, don't trade the best player at any position:
should probably be step 1, for one thing.

This:
3a. When rebuilding, trade veterans for prospects (and vice versa when winning now):
is basically what I was going on about above. This (and the similar functionality sprinkled throughout the AI code) should be a function by itself in my opinion, and all player valuation should be seen through the lens of that function. Trades, signings, re-signings... everything that the AI does, it ought to use it's overall team strategy to adjust the valuation that it settles on. Small, mid, and large market teams fundamentally think differently, and this is a primary example of how they do so. Which brings up the last point to respond to.


For example, the Marlins are currently considering trading for Manny Ramirez. If you just looked at "team type", then the Marlins would be the last team in the league that would trade for an aging $20M hitter.
I hear what you're saying, and I certainly don't disagree. The main point though is that the Marlins are obviously in a "Win now" frame of mind. The other point that I would make is that, until yesterday afternoon at least, no one considered the Marlins to be in the running for Ramirez, let alone the one team with a real shot at landing him. They were "the last team in the league that would trade for an aging $20M hitter.", until the rumors started and people thought about it. It's late July and the Marlins have a shot, so their changing their valuation of players. Their in "Win Now" mode based on their position within their division, so payroll is much less of a concern for them at the moment. In game terms, they've moved the evaluation of their Payroll Budget to the absolute bottom of the list. This stuff is all there already, it's just how it's currently used (or not used) that is creating problems, I think.

[beerchaser]

Not that this is all that relevant to the AI issues, but isn't one of the conditions of a Manny Ramirez trade that Boston eats the remaining money owed to him this year? So the real life Marlins haven't really abandoned their budgetary thinking; they will pay the price (if it happens) with Hermida and other players instead of cash.

[ohms_law]

Yea, that's definitely a factor as well.

[JeepGuy63]

I was wondering if it was just that play and something happened, so i loaded up another mogul run I did with the Red Sox 1951-1967. Same thing there. The Yankees suddenly went on a tear in the offseason signing and releasing marginal (uninjured) pitchers.. The Doodgers and Giants had a few episodes of doing it, but the yankees kept right at it after they started. From free agency in October right up to opening day, they sit there signing and releasing.
I ran into the exact same thing with the Yankees in my Seattle Pilots dynasty, running from 1969 to 1976. They were constantly signing and dropping people the same day for no reason at all. Never figured out why since the rosters all the way down the line were packed with people.

There were a few other teams that did this as well but not to the extent that New York did.

[beerchaser]

In the game I'm currently playing, Boston is the team signing up all the marginal players, even though their roster is already packed. The Yankees let Jason Giambi (rated 70/71) go to FA, and Boston quickly signed him for a couple years @ $9MM/year even though they still had Youkilis, Ortiz, et al.

[HoustonGM]

As for "bugs", non-contending teams should not be trading young star players. They should be trading their veteran players, and free agents to be. However, in Mogul, they end up trading their young very good players. For example, as was either mentioned in this thread or the trade AI thread, the Marlins would often deal away Dan Uggla, Mike Jacobs, Josh Willingham, Jeremy Hermida, AND Hanley Ramirez, and other players. The Rangers would always trade away Josh Hamilton and Ian Kinsler. The Orioles trade away Nick Markakis, etc. Those are the types of players young non-contending teams should NOT be trading. They should be KEEPING them, and trading older guys for prospects, and build for the future.

That's one thing I think is partially to blame for the competitive balance issues. Bad teams trade away their young good players, and they do that every time they have young good players but aren't contending, and thus, they get stuck in a neverending cycle of trading away their young players before the team has a chance to get good.



Yeah, this thread: http://sportsmogul.com/vbulletin/showthread.php?t=174110
Look at all the nonsensical trades people posted in that thread. That is what leads to the belief that the AI doesnt value quality over quantity and has no concept of team awareness. Teams are swapping half their starting lineup for half another teams starting lineup!

[ohms_law]

Yup. Thanks for digging that up HGM.

[filihok]

This WAS an AI FREE AGENCY thread so I'll post here.

I started a new season using BM 09 11.19, HoustonGM's rosters and CatKnight's competitive balance ideas. Equalized cities, ReCalc Salaries, etc...

Two weeks in I noticed a couple of things that seemed odd to me

1) Houston Street signed and extension to become the highest paid player in baseball. 3YR $14,000,000 surpasing ARod's Re-Calced $13 million and change contract.

2) Dodgers pitcher Derrick Lowe signed a three year extension for $3,500,000.

These two signing seem a little out of wack. First, with each other. Second, with reality.

[ohms_law]

Never mind "reality" though. The only reality within Mogul is the players in Mogul itself.
What were there ratings, what teams were involved, and what were there stats like?

[filihok]

It was two weeks into the 2008 season using HoustonGM's rosters.

from memory:

DLowe was 2-0 4.50 14IP
Street had 5 or 6 IP without a run.