Broken Created Player Development?

[Adam9586]

I am about fed up with the game and ready to go back to 2k6... but figured I'd bring this up here before giveing up on this and future versions.

After reading a lot of the threads there seems to be a lot of issues with player development that cant be properly identified. Here are the things I've seen.

1) Players on the roster initially will develop just fine.
2) Players I draft NEVER improve.

I have not checked if its only the random players generated in 2007+, but I can say with 100% certainty that not a single player drafted in 10+ years simming have gotten good enough to move to A ball (56/x). I thought maybe I wasn't putting enough money into it at first, so started with the Yankees and simmed 10 seasons with 65mil going into the farm system. Because there has been some question about swaped rateings for medical, scouting, and also farm system I also set medical and scouting at 20mil. I then released every single person in the minors and simed till the June draft, drafted the 6 guys, put them in the minors and over the next 10 (well, the following 9) seasons I filled up the minors... yet not a single guy ever hit 56/x.

So the question is, what is going on? How do you develop players?

For Clay or Ohm's Law (I'm guessing you GOTTA be an electrical engineer...) here is the test to see if your seeing what I am.

1) start a new sim useing a 2007 team
2) release all players in your minors
3) set farm expenditures at some really high number and set farm system to A+
4) set scouting to A+ and put 20mil or so on scouting
5) draft players in the first draft (and optionally force trade a few other draftees onto your team so that youfill out the minors in the first year)
6) write down (or put into excel) the overall and peak for each player
7) sim 4 or 5 seasons (just re-sign every fa so that you dont have to take the time to manage the team) noteing the overall and peak for each player

Before you do this, ask yourself what percent of guys should go from the normal range of 40 to 55 overall that come out of the drafts up to at least 60, 65, or higher? I wasnt quite as detailed as all this, but ran it for 10 seasons and not a single guy hit 60 overall, not a single one hit 65... that is broken.

Just for info I know will be asked... I'm useing 10.31, the game came from a cd bought at amazon, i've played 2004 and 2006 for years, so am comfortable with the what and how's of playing... cant think of anything else offhand your going to want to know... ask and I'll answer. I'd really like for the game to work instead of haveing to use 2k6 again.

Adam

[Adam9586]

Going to do some sims of older seasons just drafting guys I recognize. The purpose of this is I wonder if one possible reason for what I'm seeing is that only players from the Leymons (or however you spell that) list, those that are programed into the game, improve and the others are being overlooked.

Adam

[Adam9586]

For this one I took the '94 Yanks. They get a LOT less income than the 2007 version, so some of the numbers are different. Farm 35 mil, scouting 7 mil, medical 7 mil. Scouting was 2nd in spending and the others were 1st... which does bring up a question. Is dollar amount spent on an aspect improtant or whether your 1st or 30th in spending? Will 35 mil in '94 produce the same results as 35 mil in 2007? If the number 2 spender is at 15 mil do you get develop players better by spending 100mil than you do by spending 16 mil?

Ok, I did what I said to do above with the exception of the season. The players I took in the first draft are below along with the initial overall/max, and the overall/max after the 1st season and after the 2nd season. They clearly get better. Because the players from the 2007 season did not get better this makes me think that the players haveing problems are the ones that are randomly generated.

Alfonso Soriano 54/90; 58/94; 65/94
Pat Burrell 55/84; 57/88; 61/91
Mike Colangelo 55/82; 55/82; 59/82
Jeff Weaver 48/83; 55/83; 62/90
Carl Pavano 50/79; 57/88; 62/93
Ryan Dreese 49/71; 57/80; 64/78

Adam

[HoustonGM]

I'm in 2012 in my current game, and there have been drafted players that have developed fine. However, it does seem as though less players develop than used to be in Mogul, although that's just my perception and may or may not be accurate.

I do think this is related to the farm system rating issue. I also don't think that the expenses are "swapped" at all. I think there's a fundamental issue in the expenses > ratings "conversion."

For what it's worth, of all 30 teams, there are only 8 players featured in a starting lineup or rotation that were drafted since the game started in 2007. That certainly indicates something odd.

However, players that existed in the game at the start are developing just fine, which leads to an interesting question. Is the issue with the farm system rating, or is it something within the created and drafted players? Given the multiple other issues with draftees, I wouldn't be surprised...And your last post would support that.

I'm renaming this thread and moving it.

[ohms_law]

Is dollar amount spent on an aspect improtant or whether your 1st or 30th in spending?
Your ranking is the only thing that matters.


Because the players from the 2007 season did not get better this makes me think that the players haveing problems are the ones that are randomly generated.
You've done half the work, so far. The next step is to do the exact same test but stating in 2007 with fictional players.


I do think this is related to the farm system rating issue. I also don't think that the expenses are "swapped" at all. I think there's a fundamental issue in the expenses > ratings "conversion."
Me too.


However, players that existed in the game at the start are developing just fine, which leads to an interesting question. Is the issue with the farm system rating, or is it something within the created and drafted players? Given the multiple other issues with draftees, I wouldn't be surprised...And your last post would support that.

I'm renaming this thread and moving it.
Good guess, and good call.

[Adam9586]

For this one I did as stated above, useing the 2007 Yanks (I keep useing them because they have the income to put really high numbers on all 3 farm/medical/scouting). Farm 45mil, scouting 20 mil, medical 20 mil... 1st in all 3.

First years drafted players followed by the first years offseason as well as the following 2 years. i was going to do just the second year but noticed nissen (the only pitcher in the group) had improved appreciably, so did the 2nd year where he again improved nicely.

Byron Cohen 50/82; 49/81; 50/84; 54/86
Ryan Blanton 52/74; 52/72; 51/69; 53/73
Enrizues Gustavo 49/74; 49/74; 51/84; 53/80
Noberto Molina 49/70;49/69; 49/68; 51/68
Dan Cannon 50/62; 49/59; 50/58; 50/58
Cory Nissen 56/83; 56/79; 61/80; 65/77

The findings: Cohen went from 50 to 54
Blanton went from 52 to 53
Gustavo went from 49 to 53
Molina went from 49 to 51
Cannon stayed at 50
and Nissen went from 56 to 65
Overall they improved a total of 20 points in 2 full years and the half year from June to October in the 1st year. For the group of players from '94

Soriano went from 54 to 65
Burrell went from 55 to 61
Colangelo went from 55 to 59
Weaver went from 48 to 62
Pavano went from 50 to 62
Dreese went from 49 to 64
Total increase of 62 points in 1 full season and the half a season from June to October.

The difference is better than 200% MORE improvement in ONLY 60% of the time!

The only inconsistency with this and my initial impression as well as the 10 year test I did that I unfortunately didnt record numbers for like above is that in this test Nissen did actually improve appreciably where NO player had before.

So, please, try out the test I have outlined above and let us know what kind of results you come up with.

Adam

[Adam9586]

Wow, didnt realize anyone else would be up at this time! Glad you guys have looked at this so fast... and the 2nd half has been completed.

I'm still asking you guys to try it out though, because some of the other posts I've seen indicate that some people are haveing problems and others are not. There is a possibility that some of the disks/downloads have a courrupted file which is causeing the issue... which means we need someone who has not seen the problem before to check it out useing the same test parameters to see what they find. I am leaning right now very heavily towards the issue being caused by randomly generated players. I suppose it could also be an issue with the teams letter rateing for farm system, but honestly I never look at that rateing, so I cant tell you how I fared in the above tests.

Adam

[HoustonGM]

I'm in total agreement that the issue is confined to the game's randomly generated players. It's not to say that such player's NEVER develop, but that they RARELY develop and don't develop in the same manner as the other players not generated by the game.

[BM_MAN]

I am seeing this with Fictional Drafted Players

[Rongar]

I am about fed up with the game and ready to go back to 2k6...

Is that better than 2k7? I, myself, like 2k7 very much, and you can download it for free, if you do so ere the Rockies sock it to Boston, or Boston rocks Colorado.;)

[Coach Owens]

Just set the "Hall of Fame Careers" option higher in the League Editor.

[HoustonGM]

Just set the "Hall of Fame Careers" option higher in the League Editor.
This has nothing to do with it. "Hall of Fame Careers" deals with the amount that player's WITHIN their peak start and peak end have their ratings fluctuate.

[Adam9586]

Just set the "Hall of Fame Careers" option higher in the League Editor.

Yea, I tried this along with setting 100% draft predictability and even increasing the draft talent 100% because my initial thoughts were the rookie league guys were not developing (maybe the AI didnt find enough games played so punished them or something)... however the test on the '94 team showed that the problem isnt in the rookie league. The hall of fame careers and draft predictability didnt change anything.

Adam

[BM_MAN]

Check the Players I have Drafted from this Game:

[Wizkid]

Just curious as to how long I should expect it to take prospects to develop. I have a few players who are 4-5 years along in the minors and their ratings really havent improved. I have progressed them along from rookie through AAA and still nothing. I have also used them as September callups to get them some major league experience - thinking that might help to improve their skills and rating...but with little or no result. Is there something I am doing wrong, should I have more patience with them? Any help would be appreciated! Thanks :)

[HoustonGM]

See Broken Created Player Development? (http://forum.sportsmogul.com/showthread.php?t=159886)

[NYDodgers]

It's all about their peak starts and and their "Potential" rating.

The only thing I noticed was that once the fictional players start coming in, their ratings are all drastically worse than real-life draftees. I used to be able to get a ~78 peak guy with the very last pick pre-2006 or so, but now I find it hard to find a +78 after the 2nd or 3rd round in the year 2011.

[HydroSqueegee]

After not playing for a few months, i recently updated to the final patch. I immediately noticed the change in the rookies. For the few that actually develop, it takes them an eternity. The rest just stagnate and i eventually release them when they show no sign of ever hitting 70. I never had the problem until that final patch with 08.

[Wizkid]

Thanks for the info! So I am guessing, from some of what I read, that it might just be easier to trade the prospects away for veterans - since it appears that my prospects wont ever really develop? And somewhere in the link it mentioned that perhaps its only the computer generated players whose ratings dont improve with time...does that mean that after enough seasons are played and most of the current group of players are retired, I will be left with a bunch of computer-created players who will never develop into real major league talent? I guess I would have to play a heck of alot of seasons for that to happen, but its not as far off as it might seem. Please tell me that isnt the case? LOL thanks again :)

[Alloutwar]

From what I have seen of your issue, yes, once the current group of actual, real-life MLB players fade off and retire, you will be left with the underdeveloping prospects entirely.

For those of us who prefer to play completely fictional leagues, with no real-life players, this makes the long-term game pointless for BM08. BM07 did not have this problem (which is why I still play that version today).

[Wizkid]

Wow...that is really a shame! I guess I better go dump my prospects before the trade deadline comes up LOL Is there a patch or anything that can correct the problem of computer-generated prospects not developing? :)

[ohms_law]

I would take this with a bit of a grain of salt before you make any real decisions. The testing on the effects have been limited. Who knows, maybe they do end up developing...

Anyway, there will be a patch. Just need to be patient, is all.

[HoustonGM]

It's not that they NEVER develop....It's that it seems that on the whole, a very high amount don't fully develop...and those that do take very long (ie. they're drafted when they're 16...don't really become major league ready until they're 28, or whatever).

[ohms_law]

I'm not so sure about this. I started a game in 2007 and simmed multiple seasons out to 2027. The distribution of player talent appears to be normal, and there are plenty of good players:

Player Name Team Age P Roster Exp. Overall Peak Scouting
McCarty, Matt ATL 31 SP Middle 5 93 93 9.8
Roosevelt, Richard BOS 35 SP Starter #1 10 95 95 9.8
Ruilas, Jorgé CHC 36 SP Starter #1 9 94 94 9.8
Lanning, Glenn CHC 37 SP Starter #2 8 93 93 9.8
Torres, Oscar CHC 35 SP Starter #3 8 93 93 9.8
Keiper, Paul CHC 30 SP (DL) 5 93 93 9.8
Trowbridge, Tony CHW 33 3B 3B 8 93 93 9.8
Pankey, Kevin CHW 30 SP Starter #2 5 93 93 9.8
Colbert, Kevin CIN 35 SP Starter #1 9 93 93 9.8
Lehane, Travis CLE 33 CF 2B 6 94 94 9.8
Thomas, Joe DET 35 SP Starter #1 9 95 95 9.8
Sergovia, Ramon FLA 34 SP Starter #1 10 93 93 9.8
Collarin, Lance HOU 34 SP Starter #1 6 93 93 9.8
Sasaki, Brandon KCR 33 2B 2B 8 94 94 9.8
Witham, Chris LAA 32 SP Starter #2 8 93 93 9.8
Gravestock, Corey LAD 34 SP Starter #1 10 95 95 9.8
Wilson, Clayton LAD 37 SP Starter #2 9 93 93 9.8
White, Bradley NYM 34 RP Closer 9 93 93 9.8
Sagum, Aaron NYY 33 2B 2B 6 93 93 9.8
Guardiola, Fausto NYY 35 LF LF 8 95 95 9.8
Pollett, Erik NYY 33 SP Starter #1 5 93 93 9.8
Lester, Glenn SDP 34 C C 8 93 93 9.8
Jimenez, Rafael SFG 31 RP Closer 4 93 93 9.8
McBee, Justin TBD 31 SP Starter #1 5 93 93 9.8
Sprayberry, Lewis TEX 33 SP Starter #1 8 94 94 9.8
Yoshino, Yoshi TEX 32 SP Starter #2 7 94 94 9.8
Campbell, Joel TOR 33 SP Starter #2 9 94 94 9.8
Thompson, Garret WSN 32 SP Starter #1 5 93 93 9.8
Hestany, Gauher WSN 36 SP Starter #2 6 93 93 9.8


What may be going on is that the 2007 rosters are messing up the distribution of talent somewhat. If that's the case, there's probably only a couple of lean years with the transition from historic veterans to all fictional players.

[ohms_law]

ah... OK. test, Take 2.

:)

[HoustonGM]

I said this in the other thread, but I think that rather than the players failing to ever develop, the case is probably that they take much longer to develop than they used to. Peak Starts might be being listed too late

In addition, even if peak starts are okay, I think that the development of the players may be a bit skewed. It's been my perception that a lot of times players spend 7-8 years in the minor leagues, never really going anywhere, and then they suddenly shoot up in rating. Perhaps the development needs to be tweaked so that, on average, the progressions are smoother.

I think the amount of major league caliber early 20's players tends to be unrealistically low. And looking at your list, which I'm assuming is sorted by "scouting" rating, there's no players under the age of 30, let alone early to mid 20s. That's completely unrealistic. In real life, there's PLENTY of stars in their 20s. In my rosters (based on 2007 stats up through about August 22), looking at the top 30 players by scouting score, 19 players are in their 20's.

This may also be partly due to the unrealistic distribution of draft talent among high school and college players, as has been talked about a lot in this thread (http://forum.sportsmogul.com/showthread.php?t=152891). To quickly sum that up, the game-generated draftees with high peaks are nearly exclusively high school players, and the top college players have low to just slightly above average peaks, while also often starting at the same low current overall rating as the high schoolers.

[HoustonGM]

I'm going to merge this with the thread in suggestions so that we keep everything together.

[HoustonGM]

I'm also looking at the experience of those players you posted, and the earliest any of them started was age 24, and that was just one player. Everybody else started at age 25 or above...

[ohms_law]

True. I'm following 6 2007 draftees right now. Results to come.

Good call on the thread merge, IMO.

[ohms_law]

OK, Here's what I did. I started a game and put all teams under AI control. Simmed to the ammy draft, and chose one player from each round. Opened their scouting cards, and left them open, them simmed multiple seasons out to 2013.

Chris Williams
20 year old, 1st round (#1st overall) drafee in 2007
11564

Hideo Kumagai
19 Year old, 2nd round draftee (#32nd overall)
11565

Bryan Lodge
21 year old, 3rd round draftee (#63rd overall)
11566
This one is interesting. I'd certainly classify Lodge as having "failed to develop". The question is obviously, why?
His Peak start, Peak end is listed as 29, 31.
Potential = 63 (!), Longevity = 127;

Potential of 63 seems extremely low, but then he was a third round draft pick... This does seem a bit extreme, however.

Dan Mentzer
21 year old, 4th round draftee
11567
Draft predictability seems to be at work for Dan. Lucky him!
:)

Andy Carroll
20 year old 5th round draftee
11570
Looks like Andy is receiving a boost from Draft Predictability as well.

I think that all of this may be a perception issue. Generally speaking, there seems to be enough players who are developing normally, to me.

Again though, the problem may largely be the transition from historical players/custom rosters to fantasy players. It may be a good idea to look at the rosters with an eye towards balancing them for play into the future. There are likely too many historic players that fall off too quickly (have too low Longevity) to maintain the majors at the same talent level while the draftees develop.

Draft predictability appears to be a large issue with this as well. It looks to me at though Mentzer and Carroll are getting help, but Lodge was just killed. Now, weather or not that's draft predictability or there actually is something wrong with the development model I can't be absolutely certain. This does put doubt on the idea that there's something wrong with the development model though, at least to me.

[HoustonGM]

Something does seem weird though that with your other post, none of the top-rated players were below the age of 30...

And there is still certainly the problems brought up in the draft thread regarding college/high school players, but that's probably unrelated I suppose.

Just to be sure although I'm pretty sure I know the answer, all default settings right?

[ohms_law]

Just to be sure although I'm pretty sure I know the answer, all default settings right?
Yup.


Something does seem weird though that with your other post, none of the top-rated players were below the age of 30...
Most likely that's just an artifact. Statistical variation.
I can run a few sims though and check it out. You never know until you try.
:)

[Adam9586]

I just ran one from 2007 to 2014. Here are some of the results.

Ranking all current players by scouting the youngest player in the top 25 is jarrod Saltalmacchia at 28. Player 26 is only 25 years old, he is the first player in the list that was drafted from 2007 on. When I limit it to players 25 and younger I get only 8 players with overall of 90 or higher, of those only 4 are drafted. There are 43 players 80 and up with 31 of those drafted.

So, after 7 years of simming we have an average of just one player per team who is overall 80 or above who was drafted and developed. Maybe in 10 years things will start to average out again, but even if it does player careers dont last 20 years. Having 7 years where you get virtually no improvement is just not conducive to running a team. Even if things get better later, we have a serious gap where SOMETHING is not working the way it is supposed to be. I go back to my original title for the thread... the game is broken. So how do you get the free download for 2k7? If I want to play only in the past the game works fine, but I cant play a team going foreward... seems to me the older version that actually works is the only way to do that.

Adam

[ohms_law]

Even if things get better later, we have a serious gap where SOMETHING is not working the way it is supposed to be. I go back to my original title for the thread... the game is broken
While I tend to agree with the argument, I disagree with the conclusion. If there's something wrong, it's more likely that it's the rosters rather than the game itself, at this point. At least, that's my opinion, and I could easily be wrong about that.

So that we know, Adam9586 are you using the default rosters or a custom roster at all?

[HoustonGM]

I'm going to run a little test. I'll have the results tomorrow, but here's the preliminary information.

I'm going to run a sim from 2007 through 2014, using the default rosters, and check on the first round picks of the 2007 through 2013 drafts and see how they match up with the below real life data. If all is working well, the 2013 draft should appear similar to the 2006 draft, the 2012 draft to the 2005 draft, the 2011 draft to the 2004 draft, and so on, with 2007 corresponding to 2000.

Listed below are the number of top 30 picks that have made their major league debut in real life from each year this millenium. Players that did not sign are not included.

I then break it down further - how many years after the player was drafted did he debut. A player drafted in the 2004 draft will get a "0" if he debuted that year, a "1" if he debuted in 2005, a "2" if he debuted in 2006, and so on.

So, here's how I have it set up:

Year - # of draftees that made their major league debut
0 - # of draftees that debuted that year
1 - # of draftees that debuted the following year
2 - # of draftees that debuted two years after they were drafted
and so on.

2006: 4
0 - 1
1 - 3

2005: 13
0 - 3
1 - 3
2 - 7

2004: 11
0 - 0
1 - 1
2 - 8
3 - 2

2003: 19
0 - 3
1 - 1
2 - 6
3 - 7
4 - 3

2002: 21
0 - 0
1 - 1
2 - 8
3 - 7
4 - 5

2001: 17
0 - 0
1 - 2
2 - 4
3 - 7
4 - 4

2000: 15
0 - 0
1 - 1
2 - 3
3 - 4
4 - 3
5 - 1
6 - 2
7 - 1

The average number of draftees that debut in their drafted year is a little less than one. In real life, of the 7 seasons, only 2 seasons had players debut the same that they were drafted. Each of those seasons had 3 such players.

We should expect to see about 2 players debut each year that were drafted in the previous year, 5 players debut each year that were drafted 2 years prior, 4-5 players debut each year that were drafted 3 years earlier, and 4 players debut each year that were drafted 4 years earlier.

Now, this won't conclude whether or not the correct number of "stars" are developing at a realistic pace, but it will tell us whether or not the players are developing at enough of a pace to get them into the majors.

Now, there are two problems. First, in Mogul, draftees often debut earlier than they should because of injuries at the major league level and no adequate replacement in the system. This does not happen in real life. Secondly, Mogul forces each team to call up 9 hitters and 6 pitchers in September. For teams with thin minor league systems, this might result in draftees making an appearance (thus, debuting) earlier than they should.

These problems might result in the numbers being similar to in real life, when in actuality, had these problems not be occuring in Mogul, the numbers would be low. It could also result in the numbers being slightly higher than the real life numbers. If that happens, it's safe to conclude that Mogul is accurate to real life according to the parameters of this study, since the extra debutees can be attributed to the two problems above. If the Mogul draftees are substantially higher, that's another issue. So, I'm going to run the sim over night and hope to gather the results sometime tomorrow afternoon around 5:30/6:00 EST and will post them immediately, along with the conclusions I draw from it.

[HoustonGM]

Ohms, so that I'm on the safe side in case I may have inadvertantly changed my 2007Rosters.csv file, do you mind posting yours?

[Adam9586]

Ok, ran that above sim through 2026 just to get results like Ohms law (after switching to all computer controlled teams in 2014). I "find players" for all active players and got these numbers of players at each level for overall. The number after the comma is from a sim I then ran starting in 1980 running through 2000, giveing the number of players at each overall level.


90 - 9, 26
91 - 3, 21
92 - 10, 24
93 - 14, 37
94 - 9, 17
95 - 2, 6
96-100 - 0, 1

As you can see, there are a LOT more players in the 80 to '00 list than in the 07+ list. There are 132 players 90+ from the '80 to '00 compared to only 47 from the '07+ list. This STRONGLY questions the statement that it looks like normal player development later on, after a transition period. I'm not sure about the rest of you, but 47 or roughly 1.5 players per team overall 90 or better does NOT seem normal. The 132 from the pre-fictional players seems MUCH more realistic. Once again, I can only come to the conclusion that fictional players are somehow broken while the real players are not.

Adam

[HoustonGM]

Are the ratings viewed in Commisioner Mode?

Interesting results. And that you used the historical rosters, this indicates that it's not an issue with a custom roster, nor an issue with players that have Lahman data. It IS a fictional player issue.

[Adam9586]

While I tend to agree with the argument, I disagree with the conclusion. If there's something wrong, it's more likely that it's the rosters rather than the game itself, at this point. At least, that's my opinion, and I could easily be wrong about that.

So that we know, Adam9586 are you using the default rosters or a custom roster at all?

Um, the last part first... useing the default roster that came with the cd from amazon (I point out where it came from because I still am not certain its not a product of corrupted files on some games and not others since some have reported not seeing any problems). I'm also useing 10.31.

Now to the first part. I admit it has been AGES since I did any programming (and even then it was the old languages like C++ or Turbo pascal... no clue how to use java or html), so I could be missing something in my thoughts of how the game must be working/written. What I see wrong is a failure to properly develop only the fictional players. Now if programming still works the way it used to somewhere in the program there is a call function that says something to the effect of at <date> run <improvements>. The function improvements would read something to the effect of chance to improve = <team farm spending rank> * someconstant; if chance to improve > 1 then overall = overall + rand(1,20). Now assuming the programming works in some sense like that (ignoreing things such as happiness, being in the right minor league level, etc) then there would be no difference in how fictional and historical players develop. What this means is that it is possible that when the improvement function is called it is possible there is a different one being called for fictional and historical. If that is not the case then one of the variables called in that function or one of the constants used is different. For the life of me I just cant figure out how the problem could be related to the roster. IF i came up with similar numbers in the above test for number of players at each level of overall then I could see the problem being somehow associated with the roster. But at the moment I just cant see it. And as you stated above, I could also be totaly off base here... wouldnt be the first time.

Adam

After rereading this, before posting it I noticed that line saying ignoreing things such as happiness and such. It occurs to me that perhaps the game treats fictional and historic players differently in one of those areas. Things like this is why I got away from programming =)

[Adam9586]

Are the ratings viewed in Commisioner Mode?

Interesting results. And that you used the historical rosters, this indicates that it's not an issue with a custom roster, nor an issue with players that have Lahman data. It IS a fictional player issue.

Um, no, not in commisioner mode. As I understand it that would only mean your not getting the variation from the scouting +-? However, not being in commisioner mode both times would effectively cancel out any inaccuracy since both would be viewed the same way, right?

Adam

[ohms_law]

Default 2007Rosters.csv:

[HoustonGM]

Um, no, not in commisioner mode. As I understand it that would only mean your not getting the variation from the scouting +-? However, not being in commisioner mode both times would effectively cancel out any inaccuracy since both would be viewed the same way, right?

Adam
Not necessarily. The +/- is effectively applied randomly. When reporting ratings for studies that pertain to the accuracy of ratings, it's always best to look at them in Commisioner Mode as that gives the true rating. At any rate, the results of your test show drastic differences between the created players and the real players, so I'm positive that looking at it in Commisioner Mode won't change the results enough to matter.

[HoustonGM]

I just realized, until a new patch fixes the bug that results in the Amateur Draft not occuring when simming multiple seasons, I'm not going to be able to have the results. If I do the test on 10.31, the bug that places players in the wrong minor league level exists and will throw off the results.

Actually, I can just manually sim season-by-season and make sure to do the Amateur Draft...perhaps I'll do that tonight if I have time.

[Adam9586]

Can someone explain how the happiness adjustments under league editor works? I started thinking that the problem might be when the fictional players are created they do not have their happiness initialized so that the computer reads it as 0. I am thinking this is a possibility because when doing the multiple year sims it seems as if a player who starts to develop continues to develop. Its as if that first year he gets more than a couple points improvement changes something and after that he always gets improvements. If I'm guessing right on how the program must be written it would make a lot of sense and fit that pattern if the fictional players were not initialized for happiness.

To test that I started a sim and set the two happiness/personality effects to -100%. Nothing changed in development. So I am wondering if this setting works like some of the others in the game where -100% does not mean totally remove that parameter, but instead reduce it to some much smaller level. If that is the case then the non-initialized happiness would still be 0, and would then be multiplied by that adjustment factor which would still result in 0... thus each player would effectively always be 0 happiness unless something else effected it at some later time. If I'm guessing right about the program there are (rare) events which can give +x to happiness instead of multiplying by some percent. When one of those events happen the player now has a valid number for happiness and will then start to progress normally.

So the question is, does the happiness adjustments remove that part from the game or just make a big difference in how much it affects the game?

Adam

[HoustonGM]

So the question is, does the happiness adjustments remove that part from the game or just make a big difference in how much it affects the game?
Setting it to -100% does remove it from the game. The reason you saw no difference is because happiness has no effect on player development. It slightly (at the default level) affects the in-game performance of players and it affects how he acts in contract negotiations. It'd be alarming if you DID see a difference in player development with the different settings.

[HoustonGM]

I'm running the test I outlined above now. I also figured that since I'm going to be individually looking at each individual draftee, that I can make an educated guess as to whether or not he was called up due to the one of the two problems I outlined above - inadequate replacements for injury, or September call-up. With each year, next to the number of draftees who have debuted, I will list my educated guesses for how many of them were likely the result of either problem in this format:

2007: 16 (3 injury, 3 Sept.)

This would mean that of the 16 draftees that have debuted in the majors, 3 of them were likely only called up because of inadequate injury replacements, and 3 debuted in September as a result of teams being required to call up 15 players. That would mean that the "true" number of callups, the number best used to compare to the real life numbers, would be 10, not 16.

And since I actually just noticed that simming multiple seasons worked fine, we're going to have two different sets of simulated data, which can only help. It'll all be up at some point tonight.

[ohms_law]

Sounds good. Looking forward to what you find!

[HoustonGM]

Okay, well, since the two issues I raised above turned out to be much more common than I thought, it took a lot of time to compile one of the sim’s results, and since those results are so telling, I’m not going to bother doing the second one as well as similar results will definitely come. First, just as a reminder, the real life numbers – the total, as well as a breakdown of which year the debutees debuted.

2006: 4
0 - 1
1 - 3

2005: 13
0 - 3
1 - 3
2 - 7

2004: 11
0 - 0
1 - 1
2 - 8
3 - 2

2003: 19
0 - 3
1 - 1
2 - 5
3 - 7
4 - 3

2002: 21
0 - 0
1 - 1
2 - 8
3 - 7
4 - 5

2001: 17
0 - 0
1 - 2
2 - 4
3 - 7
4 - 4

2000: 15
0 - 0
1 - 1
2 - 3
3 - 4
4 - 3
5 - 1
6 - 2
7 – 1

And now, the results of the simulation:

2013: 5 (1 Injury, 4 Sept)
0 - 4 (1 i, 3 s)
1 - 1 (1 s)

2012: 8 (2 Injury, 4 Sept)
0 - 2 (1 s)
1 - 2 (2 s)
2 - 4 (2 i, 1 s)

2011: 13 (3 Inj, 6 Sept)
0 - 3 (2 i, 1 s)
1 - 2 (2 s)
2 - 3 (2 s)
3 - 5 (1 i, 1 s)

2010: 16 (2 Injury, 10 Sept)
0 - 2 (2 s)
1 - 2 (2 s)
2 - 3 (3 s)
3 - 5 (2 s)
4 - 4 (2 i, 1 s)

2009: 20 (3 Injury, 12 Sept)
0 - 4 (3 i, 1 s)
1 - 4 (4 s)
2 - 2 (2 s)
3 - 5 (4 s)
4 - 1 (1 s)
5 - 6

2008: 26 (5 Injury, 7 Sept)
0 - 3 (3 i)
1 - 2 (2 s)
2 - 4 (2 i)
3 - 3 (1 s)
4 - 7 (2 s)
5 - 4 (1 s)
6 - 4 (1 s)

2007: 27 (8 Injury, 10 Sept)
0 - 4 (2 i, 1 s)
1 - 9 (2 i, 6 s)
2 - 6 (2 i, 3 s)
3 - 2 (2 s)
4 - 2
5 - 1
6 - 2
7 – 1

I was really lenient, as well, on who I designated as their reason for their call up as September call up or injury replacement. Meaning, a lot of times, I gave the player the benefit of the doubt and did NOT count his debut reason as Injury or Sept. I counted a few guys as normal that, say, were of age to debut, but still clearly debuted because of one of those factors.

Okay, now, below are a much of different conclusions, as well as other things I noticed while looking through the players.

1) Plain and simple – the creation of drafted players is screwed up, period, whether or not there’s an issue with development. Some of the issues with draftees are mentioned extensively in this thread (http://forum.sportsmogul.com/showthread.php?t=152891). I reiterate some of those below, as well as mention others I may have noticed

2) Way too many drafted players make it to the major leagues. The overwhelming reason for this is the issue with September call ups. Teams should not be forced to bring up 15 players. They should be allowed to bring up UP TO 15 players, and as a check, it should be limited to players that are in AAA, or a lower level but have a 75+ rating. Overall, 115 drafted players made it to the major leagues. This is comparable to the 100 over the same amount of time in real life. HOWEVER, 77 were due to being an injury replacement at an otherwise barren position or making a couple appearances in September. That means just 38 players debuted thanks to their own skills.

3) There are way too many catchers drafted in the first round.
---3a) Catchers made up a large part of the injury-related call ups, as well as a good amount of the September call-ups. This is because the game only starts with 2-3 catchers for each team, and the AI always tries to have at least 2 catchers on the major league roster. When one catcher gets injured, a common occurrence and rightly so, since there’s no other catcher in the system besides recently drafted ones and the back up, one of the new guys gets called up out of necessity. I am going to attempt to alleviate this issue myself for my BM2k9 rosters by adding an extra 3-4 catchers per team.
---3b) Of the drafted catchers, way too many are 16 or 17 years old. In real life, college catchers are much more common than high school catchers. Combine this with 3a and a lot of catchers debut in their teens.

4) There are way too many 16 and 17 year old draftees. In addition, college players are at a significant rating disadvantage compared to high school players. I’ve went over this more in the above linked thread.

5) Players come up in September, exclusively, for too many consecutive years. September call ups in real life consist of veteran players that aren’t that good, minor league veterans, and the occasional near-MLB ready top prospect. In real life, not many players have streaks of years where they play in 1-10 games and that’s it. In Mogul, it’s pretty common to see players that appeared in 5 different years from ages 19-23, never appeared in more than a handful of games, and then finally start to have a career. This is entirely unrealistic.

6) Players stay in Rookie-ball for way too long. The most extreme case I saw was a player needed six and a half years in Rookie-ball.

7) For pitchers, movement factors into the overall rating of a pitcher WAY too much compared to control. A pitcher with a high movement is going to have a decent enough rating to make him look major-league ready, even with an abdominally low control. In 2012, the guy that appeared that season was a reliever with a 97 movement and a 63 control. He had an overall rating in the high 70s and stayed exclusively in the majors. In 3 years, he pitched 225 innings. He had a career ERA of 7.10. He struck out 83 batters while walking 181, and his movement didn’t help him – he gave up 266 hits. Another example was a 19 year old with a good movement and a low control debuting and throwing 22.2 innings with a 7.54 ERA, walking 15 while allowing 27 hits. Control needs to factor in more, or movement less.

8) In real life, most of the players that debuted went on to continue playing at least semi-regularly. Some got injured, a few just didn’t go beyond a couple cups of coffee, but generally, they went on to be a big leaguer. In Mogul, the drafted players often debut exclusively for the above reasons – injury or September call up – and that’s it.

9) Unlike Adam9586’s test ( http://forum.sportsmogul.com/showpost.php?p=903034&postcount=37 ), I got a distribution of ratings above 90 that were comparable to his test with real life players.

90 – 39
91 – 13
92 - 21
93 – 31
94 – 13
95 – 11
96 – 6
97 – 1
98 – 2

Of the 137 “stars”, only 19 were drafted. The rest existed at the time the game started. I did a quick judge of the real life sample, and by my count, 19 players would likely translate into a 90+ rating in Mogul, which is pretty eerie. However, in Mogul, of the 19 drafted “stars”, only 9 of them were first round draft picks, suggesting that Mogul’s top picks don’t translate into stars as often as real life top picks do.

However, this goes back to my point about catchers being taken in the first round too often – they’re not projected to be stars. If the first round of the draft was distributed better across the positions, it might better match real-life.

10) It does seem as though recently drafted players don’t make the majors and stick early in their 20’s as often as real life. This is probably due to the college issue.

The only conclusion I can come to, and one that I’ve actually stated before, is that the creation of draftees, as well as the draft AI, needs to be overhauled. I think there might be slight problems with the development of players, but I can’t conclusively say that, as players did develop.

[ohms_law]

2) Way too many drafted players make it to the major leagues. The overwhelming reason for this is the issue with September call ups. Teams should not be forced to bring up 15 players. They should be allowed to bring up UP TO 15 players, and as a check, it should be limited to players that are in AAA, or a lower level but have a 75+ rating. Overall, 115 drafted players made it to the major leagues. This is comparable to the 100 over the same amount of time in real life. HOWEVER, 77 were due to being an injury replacement at an otherwise barren position or making a couple appearances in September. That means just 38 players debuted thanks to their own skills.

...

5) Players come up in September, exclusively, for too many consecutive years. September call ups in real life consist of veteran players that aren’t that good, minor league veterans, and the occasional near-MLB ready top prospect. In real life, not many players have streaks of years where they play in 1-10 games and that’s it. In Mogul, it’s pretty common to see players that appeared in 5 different years from ages 19-23, never appeared in more than a handful of games, and then finally start to have a career. This is entirely unrealistic.

...

8) In real life, most of the players that debuted went on to continue playing at least semi-regularly. Some got injured, a few just didn’t go beyond a couple cups of coffee, but generally, they went on to be a big leaguer. In Mogul, the drafted players often debut exclusively for the above reasons – injury or September call up – and that’s it.
The only thing about this is that I don't see what difference it makes. As long as the players are not actually moved to the regular roster it doesn't look like their service time increases. So, the only effect that I see from having too many September callups is that more players receive extra playing time.
The short version: It doesn't hurt anything, and actually the increased playing time probably helps development.

Point #5 & 8 are obviously related, so I'll address it together with this.
I see what you're saying here, and I can't really disagree... I just end up back at the point which I made above, though. It doesn't really hurt anything, and actually the increased playing time probably helps development.

---3b) Of the drafted catchers, way too many are 16 or 17 years old. In real life, college catchers are much more common than high school catchers. Combine this with 3a and a lot of catchers debut in their teens.

4) There are way too many 16 and 17 year old draftees. In addition, college players are at a significant rating disadvantage compared to high school players. I’ve went over this more in the above linked thread.
All true, and I agree... for the modern era. Let's please not forget about history, though.


6) Players stay in Rookie-ball for way too long. The most extreme case I saw was a player needed six and a half years in Rookie-ball.
True. 10.33 seems to have helped this, slightly. I think we've at least gotten Clay on the correct track with this subject, as long as the work continues.


7) For pitchers, movement factors into the overall rating of a pitcher WAY too much compared to control. A pitcher with a high movement is going to have a decent enough rating to make him look major-league ready, even with an abdominally low control. In 2012, the guy that appeared that season was a reliever with a 97 movement and a 63 control. He had an overall rating in the high 70s and stayed exclusively in the majors. In 3 years, he pitched 225 innings. He had a career ERA of 7.10. He struck out 83 batters while walking 181, and his movement didn’t help him – he gave up 266 hits. Another example was a 19 year old with a good movement and a low control debuting and throwing 22.2 innings with a 7.54 ERA, walking 15 while allowing 27 hits. Control needs to factor in more, or movement less.
One thing that I've noticed is that the above is extremely dependant on the role which the pitcher plays. For starters the above is not as true as it is for relievers.
Anyway, this should be it's own topic IMO...


However, this goes back to my point about catchers being taken in the first round too often – they’re not projected to be stars. If the first round of the draft was distributed better across the positions, it might better match real-life.
This is definately true, but I come back to the weaknesses in the fielding position/defensive ratings system currently in Mogul. That needs to be addressed first before this can be addressed, in my opinion.


9) Unlike Adam9586’s test, I got a distribution of ratings above 90 that were comparable to his test with real life players.

90 – 39
91 – 13
92 - 21
93 – 31
94 – 13
95 – 11
96 – 6
97 – 1
98 – 2

Of the 137 “stars”, only 19 were drafted. The rest existed at the time the game started. I did a quick judge of the real life sample, and by my count, 19 players would likely translate into a 90+ rating in Mogul, which is pretty eerie. However, in Mogul, of the 19 drafted “stars”, only 9 of them were first round draft picks, suggesting that Mogul’s top picks don’t translate into stars as often as real life top picks do.
That's what I noticed as well. With that being the case, I've basically dismissed the possibility that the problem is as originally stated.

My conclusions are that there are several issues which still need to be addressed regarding draftees and the minors (most of which were brought up here). The idea that fictional draftees simply don't develop correctly due to the fact that their fictional just doesn't hold up though.

[Adam9586]

9) Unlike Adam9586’s test ( http://forum.sportsmogul.com/showpost.php?p=903034&postcount=37 ), I got a distribution of ratings above 90 that were comparable to his test with real life players.

90 – 39
91 – 13
92 - 21
93 – 31
94 – 13
95 – 11
96 – 6
97 – 1
98 – 2

Of the 137 “stars”, only 19 were drafted. The rest existed at the time the game started. I did a quick judge of the real life sample, and by my count, 19 players would likely translate into a 90+ rating in Mogul, which is pretty eerie. However, in Mogul, of the 19 drafted “stars”, only 9 of them were first round draft picks, suggesting that Mogul’s top picks don’t translate into stars as often as real life top picks do.



I think there might be slight problems with the development of players, but I can’t conclusively say that, as players did develop.

Only parts of the original post were quoted.

Houston did you save that game? If so if there a chance you could make all teams computer controled and sim it another 12 or 13 years? The reason I ask this is you got a distribution similar to mine for the players from the 80's, however most of the players were not fictional. I'd like to see if you get results similar to what I did for all drafted players or if your getting results similar to the historical players. I'd like to see either some validation or evidence I'm wrong useing the same test.

Adam

[HoustonGM]

The only thing about this is that I don't see what difference it makes. As long as the players are not actually moved to the regular roster it doesn't look like their service time increases. So, the only effect that I see from having too many September callups is that more players receive extra playing time.
The short version: It doesn't hurt anything, and actually the increased playing time probably helps development.
It's just unrealistic. I hate looking at a scouting report, seeing a player play in 1-9 games each year from ages 20 to 24, and then getting a full time job at 25. At any rate, 15 mandatory September call ups is unrealistic, regardless of whether or not it's resulting in such players. I'm a stickler for realism.


All true, and I agree... for the modern era. Let's please not forget about history, though.
Yes. Unless I specify otherwise, I'm always talking about the modern era. I don't know much of anything about historical draftee ages.


One thing that I've noticed is that the above is extremely dependant on the role which the pitcher plays. For starters the above is not as true as it is for relievers.
Anyway, this should be it's own topic IMO...
Yep, good point. I should've mentioned that. It's relievers that get hit by the more than starters. And I will make a topic for it.


This is definately true, but I come back to the weaknesses in the fielding position/defensive ratings system currently in Mogul. That needs to be addressed first before this can be addressed, in my opinion.
True. Those weaknesses, if that includes the AI for playing players outside their "default" position, should also help the number of players that are prematurely called up as injury replacements.


My conclusions are that there are several issues which still need to be addressed regarding draftees and the minors (most of which were brought up here). The idea that fictional draftees simply don't develop correctly due to the fact that their fictional just doesn't hold up though.
Sounds about right.


Houston did you save that game? If so if there a chance you could make all teams computer controled and sim it another 12 or 13 years? The reason I ask this is you got a distribution similar to mine for the players from the 80's, however most of the players were not fictional. I'd like to see if you get results similar to what I did for all drafted players or if your getting results similar to the historical players. I'd like to see either some validation or evidence I'm wrong useing the same test.
Fair enough. I do have the game saved, so I'll sim ahead to 2028 and post the updated results. Also, for clarification's sake, the sim from 2007 through 2014 was ran with all teams computer-controlled anyway. I'll get back with the results.

[HoustonGM]

Well, apparantly, simming multiple seasons on 10.33 results in the amateur draft being skipped, so, the test I just ran is a moot point as theres no players, period. >_<

I'm going to use my other test from above that i didnt post results from, since I inadvertantly saved after simming up to 2028 with this test run, and run that until 2028. I did the above test on 10.33 since it fixed the problem with draftee placement in the minors, which I didn't want skewing the results, but for a ratings snapshot, 10.31 shouldn't matter, and since that's what you ran your tests on, it's probably better to be consistent anyway. I'm going to run the sim while I sleep and post the results tomorow afternoon.

[Coach Owens]

I would not want to be your computer tonight. Make sure you back the file up.

[HoustonGM]

I would not want to be your computer tonight. Make sure you back the file up.
Running a 13 year sim is not bad or hard for the computer at all. I just ran the one that I found out that bug while doing multiple other things on the computer.

[HoustonGM]

Another slight issue..it seems as though pitchers that develop really quickly and do well in the majors quickly are pretty rare. In 2015, there's only a couple pitchers rated above 90 that were drafted. In real life, there's many instances of guys developing quickly, despite not necessarily having a high "overall" rating on draft day - Tim Lincecum, Yovani Gallardo, Francisco Liriano, etc. These types of players seem to be really rare in Mogul.

I think this has to do with the player development model, as it seems rare for players to rise quickly at a young age...(And I'm cross-posting this to the drafted player thread)

[Coach Owens]

Running a 13 year sim is not bad or hard for the computer at all. I just ran the one that I found out that bug while doing multiple other things on the computer.

Trust me, on mine it would. ;) My computer goes crazy if left on all night.

[HoustonGM]

Trust me, on mine it would. ;) My computer goes crazy if left on all night.
My computer has been left on for months now, with maybe the odd restart for a new update or something every once in a while. I've currently been signed on to AIM for over 2 days, but that's because of an AIM restart, not a computer restart. I've been signed on to AIM for 12 straight days once..:p

Anyway, the results.

This is viewed in Commisioner Mode, and looking at the differnce between Commish and non-Commish, that might have skewed your results, Adam. I forgot that when not in Commish, the game rarely, if ever, shows you an overall rating above 95, when such players do exist. Here, I'll post the Commish results first, and then the non-Commish ones so you can compare.

90+ Rating Breakdown, Commisioner Mode:
90 - 28
91 - 17
92 - 20
93 - 31
94 - 13
95 - 15
96 - 8
97 - 5

90+ Rating Breakdown, Non-Commisioner Mode:
90 - 24
91 - 22
92 - 25
93 - 34
94 - 22
95 - 3
96+ - 0


Again, the only conclusion I can come to is that players, in the long run, develop fine, but the issue is with young players not developing quickly like they often do in real life, as I said in my previous post in this thread, which we're talking about more in this thread (http://forum.sportsmogul.com/showthread.php?t=152891).

[FrankNoChance]

I created a league (starting in 2007) with all fictional players. I've kept my farm system spending at #1 throughout all seasons. For the 2007 season, it seemed that farm players were developing ok. By 2009, I noticed that most of my farm system players weren't improving. For example, at the beginning of 2009, I had two pitchers in AA ball with ratings of 67/93 and 67/94. I'm now at the beginning of May 2010 and neither player has gained a single point -- not in over a year of being #1 in spending on the farm system. Both pitchers were getting appropriate playing time.

The same seems to be true of most of the other players in my farm system. All of the guys that were in my R league a year ago, are still there now -- with exactly the same ratings.

I have seen improvements in some of my farm players, but not many -- and something profound seems to have happened between 2007 (or 2008?) and 2009 because farm system development on my team virtually stopped. Again, it's not true of all players. My sense is that my guys in AAA are improving ok, but not the others. Certainly, my major league players are improving, regardless of their age.

What this suggests to me is that there isn't a problem with the development of fictional players, per se -- at least not ALL fictional players. Could there be a problem with development of DRAFTED players or players created by the game after the first season? Another thought I had was that maybe there could be a problem with R, A, and AA ball not helping players improve as much as they should. If players develop too slowly at those levels, then you might see what some people are reporting -- drafted players that take an unrealistically long time to make it to the majors.

It seems to me that they way to test this is to get detailed stats on all of a team's farm system players and keep track of the ratings for the whole farm system for that team across, say, 10 or 20 seasons. Noting which players develop and which ones don't might reveal something important. I'll try to do this test this weekend and post the results. It should be interesting to see year-by-year ratings for all of the R, A, AA, and AAA players for 10 or 20 seasons. By correlating the changes in ratings with variables like player age, farm level, etc, we might track the problem down.

[ohms_law]

Just for the sake of clarity, since it's somewhat confused in your post above: FrankNoChance, are you currently using Baseball Mogul 2008, or Baseball Mogul 2007? If you click Help -> About, what version number is Baseball Mogul currently at?

ps.: Welcome to the Baseball Mogul community!
:)

[Coach Owens]

Hm. I had a guy named Rob Clay, the #13 pick of the 2007 draft. He started out around a 55, and three years later was a 63.

[FrankNoChance]

Just for the sake of clarity, since it's somewhat confused in your post above: FrankNoChance, are you currently using Baseball Mogul 2008, or Baseball Mogul 2007? If you click Help -> About, what version number is Baseball Mogul currently at?

ps.: Welcome to the Baseball Mogul community!
:)

Thanks for the welcome. I'm playing version 10.3.1.0 of Baseball Mogul 2008. Fantastic game, BTW. I've waited years for a game like this.

I created a league with all fictional players. The first thing I did was take the Atlanta Braves and spent $30,000,000 on the farm system to make sure we were #1 all year. Then I wrote down some numbers for some top farm prospects:

All ratings were checked using Commish mode.

Aboytes 66/95
Russell 67/88
Taylor 66/94
Sazameda 61/78
Cahill 74/93

Then I simmed a whole year and rechecked these guys to see their improvement. The same guys a year later:

Aboytes 73/97
Russell 75/90
Taylor 71/90
Salzameda 65/77
Cahill 81/94


I realize that this is only a few players, but it confirms what I've seen in other leagues I've created -- the ones that are in the game when you first create the leage (whether fictional or not) improve fairly rapidly.

Next, I wrote down the names and stats of top draftees for the Braves for the 2008 and 2012 seasons and tracked them for a year (same spending on farm system):

Waldman 50/88
Foley 49/95
Saxton 50/92
Davis 63/88
Brewer 46/80
Frey 47/74

Here are the same draftees a year later:
Waldman 56/89
Foley 50/94
Saxton 50/89
Davis 65/96
Brewer 50/83
Frey 48/72

It's a small sample, but something odd seems to be happening. Shouldn't all farm system players improve at least a little in a year if you are spending $30,000,000 on the farm system? All of the farm players in the game at league's creation improved by at least 4 points and most improved more than that.

One thing I noticed was that the potential ratings seemed much higher for players that were in the game when the league was created than for players that were drafted. When you click on "potential" in the player editor, it says that this variable affects the rate of improvement towards the peak. So, I decided to compare the potential ratings for draftees to those for (fictional) players in the game at the start of the league. To do this, I wrote down the potential ratings for every first round draft choice in the league for the 2012 season. Here they are:

117 128 100 71 169 107 63 105 77 111 93 100 167 100 74 93 66 83 143 136 125 55 91 125 67 83 135 125 200 143

So, there was only one player with a potential rating of 200 or more and only 3 with potential ratings above 150. Next, I wrote down the potential ratings for 30 prospects that were in the game at the league's creation. I was careful to make sure that the actual and peak ratings were comparable by choosing some of the top prospects in several organizations.

Potential ratings for 30 guys in the game at league's creation:
186 100 183 177 113 119 226 100 299 85 175 168 242 195 171 62 141 151 190 119 293 342(!) 212 185 375(!) 125 117 118 174 120 288

Note that there are only two players on this list with potential ratings under 100. In contrast, there are 19 players over 150 and 8 over 200.

So, I think that one reason why people are seeing a sudden slowdown in the improvement in their farm system is that the draftees have *MUCH* lower potential ratings than farm system players that are in the game when you create the league.

The other thing I noticed is that most of the farm system guys in the game at the beginning of the league were in their 20's. Very few were under 19. In contrast, the draftees were quite young. Is the rate of improvement affected by how CLOSE a player is to his peak age? That is, if a player is four years away from his peak, does he improve faster than a player with the same actual/peak ratings that is, say, 8 years away from his peak? If so, then that might be another factor.

I'll try to find time to do more testing this weekend but, based on what I'm seeing, it looks to me like the problem may be with the potential ratings of draftees -- or at least with the way the game handles them. Specifically, it looks like there aren't enough really high potential ratings for drafted players. In real major league baseball, there are some young stars that rapidly shoot through the minors and make the majors at a young age. That seems to be too rare (or nonexistent?) for draftees in MB2008.

[boomboom]

Hey guys, something with BMO that might translate in BM2k8, is that the talent level in the current rosters are really high, I would suggest simming 20 years into the future, with out multiple season mode, then show your results...because the game likes balance, if the game has too much talent, it will correct itself.

[kring]

Player development for draft picks definitely has a problem in Mogul 08. The distribution is heavily weighed to players who don't develop and are useless no matter how long you wait, regardless of the peak rating. It's not realistic. It's fine for there to be lots of disappointing players but it needs to be balanced by some who do as well as or better than expected. The ages and positions of those with high peak (>90) also seems to be off, not that it matters since they won't develop anyway. It seems that the top peak ratings are all 16-18 year old pitchers, who won't develop, followed by position players with much lower peak ratings, who also will not develop. In my opinion High heat baseball 2002 has a much more realistic player development simulation. Is Baseball Mogul ever going to catch up to where HH was in '02? Maybe by 2020?

[HoustonGM]

Hey guys, something with BMO that might translate in BM2k8, is that the talent level in the current rosters are really high, I would suggest simming 20 years into the future, with out multiple season mode, then show your results...because the game likes balance, if the game has too much talent, it will correct itself.
There's been a couple results posted by myself, ohms, and Adam.


The ages and positions of those with high peak (>90) also seems to be off, not that it matters since they won't develop anyway. It seems that the top peak ratings are all 16-18 year old pitchers, who won't develop, followed by position players with much lower peak ratings, who also will not develop. In my opinion High heat baseball 2002 has a much more realistic player development simulation. Is Baseball Mogul ever going to catch up to where HH was in '02? Maybe by 2020?
Yes. This has been brought up multiple times in thsi thread, and a link to the thread that originaly talks about it.

http://forum.sportsmogul.com/showthread.php?t=152891

[FrankNoChance]

Hey guys, something with BMO that might translate in BM2k8, is that the talent level in the current rosters are really high, I would suggest simming 20 years into the future, with out multiple season mode, then show your results...because the game likes balance, if the game has too much talent, it will correct itself.

Just to follow up. I simmed all the way to the 2027 season today. Potential ratings for all of the 2027 1st round draft picks:

90 111 189 51 143 111 160 95 113 83 100 76 111 76 79 71 77 125 83 71 125 100 113 67 116 200 200 97 71 83

Comparing that with the numbers in my previous post, it seems that the potential ratings are much higher for players that are in the game when you create the league (even leagues with all fictional players) than they are for rookies that the game creates for the draft in later seasons. That's why player development seems to slow down so much. There are still some players reaching very high levels in the 2027 season, but not as many, of course.

I think that this could be improved by altering the distribution of potential ratings for draftees, so that there would be more in the 200's and 300's. This would lead to a more realistic situation where an occasional draftee shoots through the minors and is playing in the majors at a young age.

The current situation is not ideal, but it's not close to what I would call "broken". Here's a couple of screenshots showing the rosters for the team that won the world series that year. As you can see, there are still some players with high ratings. It just seems to take most players an unrealistically long time to get through the minors.

Maybe that could be improved by bumping up the talent level for draftees and/or bumping up the percentage of hall of fame careers?

[FrankNoChance]

By the way, those screenshots were not taken at the end of the season, so those HR totals are *NOT* for a full season. :)

[HoustonGM]

Yeah, the potential ratings do seem very low as a whole...

[pirate66]

I have run BM8 far into the future, having started in 1901 with all of the default settings and simply allowed the game to go where it went. I have one human controlled team and the rest are computer controlled. When I upgraded to 10.31 I was already around the year 2025 if my memory is correct. So, I have had nothing but fictional players for quite some time. I can tell you that the new player talent and progression significantly changed when I upgraded.

I am currently in the year 2020. Having read this thread I checked and found that I have 56 players in the league rated higher than 90 (not commish mode), 19 of those are pitchers. There are no players rated higher than 93.

As for the draft pool (my scouting is +/-1) there are always a ton of pitchers with 90+ potential. There are only a handful of fielders per draft with 80+ potential. There are really not many with 70+ potential. By the third round I am almost always drafting players with potential in the 60's.

I have never really understood the progression, but here are my observations:

Pitchers are the easiest to figure out. I won't draft a pitcher with a potential less than 90. I usually get two or three per draft. Over the next few years their overall will increase and their potential will either stay the same or decline. I know that, depending on the pitcher, if you don't debut them soon enough their potential will begin to fall rapidly. If a pitcher is still in the minors after 5 years his potential may have fallen to as low as the high 70's.

Fielders are another story. You never know which way they are going to go. I've drafted guys that have a potential of over 80 and they never get any better. Plenty of draftees never develop at all. After several years they are no better than the day I drafted them so I just dump them. Some of them fall like rocks. I drafted a 16 year old SS with a peak of 74, but 4 years later his max is 59. But I have drafted guys with ratings like 48/66 and they have turned out to be 90+ power hitters.

This is how I figured that out: at some point in the future, or when I installed 10.31, I am not sure which, the game dynamics change somewhat. Batting averages and ERA's went up. The league batting average is about 280-300 yearly and the ERA is about 5.00 to 5.25. There are usually 30-40 players with batting averages over 325. But the home runs, and power in general, are down.

It is not an oddity to see the leader board showing that the #5 guy in both leagues hit 15-20 homeruns. There are usually 2 or 3 power hitters in the league. But these guys are belting out 80 - 110 homers in some years. I could never find them in the draft but I always thought that was because my draft picks were just too low and they were already taken. Wrong. After I picked more than a few as free agents over the years I began to notice something. As I was looking at their vitals and transactions these guys were all drafted deep in the draft with draft day potentials in the 60's. It seemed that they would end up being put in MLB and then blossom overnight.

In point, the best power hitter in the league right now was drafted when he was 18 by the White Sox as the 90th pick. He was a 57/71 rating. It stayed there until he was 21 years old when his rating rocketed up to 58/88 overnight. He hit the majors at 24 as a 71/94. At 26 he was 88/95. He is now 35 and still at 93/93. He hit 82 home runs last year, more than twice as many as the next guy. He is one of only 2 guys in the top ten list that had peaks of over 70 when they were drafted. This is not unusual.

My point is that I don't really see much predictability in the draft other than in pitchers. But fictional players definitely do progress.

[pirate66]

I apologize, I meant I am currently in the year 2120!

[CatKnight]

I saw this thread ... groaned mightily... then tried a few tests.

Version 10.31:

2007 Draftee # 1: 50/92, 48/88, 48/86
2007 # 2: 54/79, 49/84, 53/83
2007 # 3: 53/70, 48/65, 49/62

2008 # 1: 74/97, 66/91
2008 # 2: 58/97, 47/86
2008 # 3: 64/87, 54/72

I thought I'd backtrack, so made a new folder/copy with 10.17, the one I downloaded.

2007 # 1: 72/86, 75/90, 73/84
2007 # 2: 56/95, 60/95, 64/98
2007 # 3: 61/88, 64/87, 64/83

2008 # 1: 69/93, 67/90
2008 # 2: 61/95, 57/85
2008 # 3: 62/95, 57/91

In both cases I made sure my team was A+ in all expenses and paying to stay there.

GRRRRR! What the devil is happening?

[ohms_law]

That looks like draft predictability, to me. That's the whole intent behind it.

[HoustonGM]

What exactly are each of the sets of ratings? Commish mode/not commish mode?

[CatKnight]

No...that looks like players not developing to me.

Houston: For the 2007 players, that's their commish mode ratings on draft day, on draft day 2008 (one year later), and draft day 2009 (two years later.)

For the 2008 players, it's draft day, then draft day 2009 (one year later.)

[ohms_law]

try it again with draft predictability set to +100%. It's supposed to be doing what it's doing above. If you don't wan that effect, turn it off.

[CatKnight]

I'm not sure I'm explaining myself clearly. I don't mind unpredictability. I kinda like that the peak can and does change.

What I am having a problem with, is that these players are showing no signs of developing - of heading from their base overall value towards their peak.

In the example I gave, in two years my 2007 draftees lost 2, 1 and 4 overall points respectively. My 2008 draftees lost 8, 11 and 10 points in one year. They didn't improve. My A+ farm system HURT them badly.

Are you saying that the default +0 value is for draftees to actually LOSE talent as they age?? In an A+ organization?

I did follow through with your idea though, and got the following results:

Draft Predictability +100%

07 # 1: 53/86 (Draft Day), 53/80 (2008), 53/83 (2009): No improvement
07 # 2: 48/85 (Draft Day), 49/82 (2008), 62/86 (2009): +14 points
07 # 3: 49/78 (Draft Day), 52/76 (2008), 57/76 (2009): +8 points

08 # 1: 50/88 (Draft Day), 61/84 (2009): +11 points
08 # 2: 48/85 (Draft Day), 47/83 (2009): -1 point
08 # 3: 48/80 (Draft Day), 46/76 (2009): -2 points

This is closer to what I expected.

*******

Just for comparison purposes, I also tried:

Amateur Draft Talent +100%

07 # 1: 62/79 (Draft Day), 61/74 (2008), 62/74 (2009): No improvement
07 # 2: 48/87 (Draft Day), 51/86 (2008), 48/83 (2009): No improvement
07 # 3: 49/80 (Draft Day), 50/79 (2008), 48/78 (2009): -1

08 # 1: 60/84 (Draft Day), 58/82 (2009): -2 points
08 # 2: 57/87 (Draft Day), 54/80 (2009): -3 points
08 # 3: 62/75 (Draft Day), 57/69 (2009): -5 points

No dice.

[ohms_law]

Well, Amateur Draft Talent just makes all draftees immediately better, so I'm not surprised there.

I remember looking at some of this a couple of days ago. There might be a reversed mathematical sign, or a multiplier that is being mistakingly added somewhere along the line. Who knows. I'll take another look... probably tomorrow. This type of fix doesn't really fall into my bailiwick right now though, unless I can find something obviously backwards or something like that hidden away someplace.

[CatKnight]

Hm, I wonder if I found a clue.

I thought about your idea (raising predictability), then thought of the modifiers HoustonGM uses and thought I'd give them a try. They are:

Amateur Talent -20
Predictability -30
HoF Career +20

The results:

07 # 1: 67/77 (Draft), 79/82 (2008), 78 (2009), 75 (2010) = +8
07 # 2: 61/68 (Draft), 67/70 (2008), 72/73 (2009), 73 (2010) = +12
07 # 3: 48/80 (Draft), 51/80 (2008), 69/89 (2009), 73/90 (2010) = +25

08 # 1: 55/87 (Draft), 58/91 (2009), 58/91 (2010) = +3
08 # 2: 51/85 (Draft), 50/83 (2009), 55/85 (2010) = +4
08 # 3: 61/84 (Draft), 67/89 (2009), 63/82 (2010) = +2

09 # 1: 52/91 (Draft), 51/85 (2010) = -1
09 # 2: 64/70 (Draft), 65/72 (2010) = +1
09 # 3: 53/77 (Draft), 54/83 (2010) - +1

After the first year's exhuberance things settled down a bit, but the players did progress slowly. Again, this is an A+ organization. However, this is also a good team (the Diamondbacks) so the league might be rebalancing slightly.

Nothing conclusive, but perhaps you're right and this is a matter of finding good settings...

[Coach Owens]

I've never had trouble with the slow development. The only problem I've had is teams with a high draft pick choosing low peak players, such as catchers even though they don't need them. I.e. I started in 2007, in 2009, Atlanta had one of the #5 picks in the draft. They drafted a catcher who was like 78 peak even though they still had 25-year-old McCann who was overall in the 90s and there were many better players left.

[sabersox]

I'll add my two sense to this. Let me see first that I am a guy who both played college baseball (alright I started for a DIII team) and a guy who loves sabermetrics.

I've found that in BM 2008 draft and player development has taken a huge step back. I usually pick from the bottom, but the fact is that there are too many high school high reward / huge risks guys and not enough college high payoff / low risk guys. In RL the top picks are often college pitchers who have dominated. But every year in my SIM there are no college pitchers that will go higher than late 1st round. In BM 2007 there were (at least with my settings) a nice amount of 72/87 college pitchers with exellent sabermetric stats that I would draft. Needless to say these guys formed the backbone of my pitching stafs.

Now these guys are gone.

As many others have said player development has also taken a huge step back in 2008. As a matter of fact it seems almost all picks will see their peak ratings (regardless of farm system rating) drop by 5-10 points once I draft them. This holds true for computer picks as well.

In BM 2008 I find myself being forced into picking RP almost exclusively in the first two rounds, as these guys are the only impact players available. Good luck finding a seasoned college pitcher or position player. Compared to RL, and just for balance, this isn't ideal.

I'd like to see the whole system revamped next year, with greater strides being made to increase the realism in the draft and devolopment process.

[HoustonGM]

Sabersox, this thread should interest you (http://forum.sportsmogul.com/showthread.php?t=152891).

[CatKnight]

So I ran another experiment to test this problem. I tracked six teams - two big market, two in the middle, two small market - through 25 years in 5 year intervals. After 25 years all but maybe a small handful of "historical" rookies should be out of the system and we should be looking at a pure fictional league, with players brought up through the game's farm system.

My theories were:

1) The big market teams might stabilize because any loss in player performance would be offset by economic dominance.
2) Conversely, the small market teams would see a notable crash as they got the chaff of an ineffective farm system.
3) However, we would see a substantial rise in relief quality as the drafts are dominated by them.

Hmm.

Overall ratings of 1) 8 field starters, 2) 5 starters, 3) 6 relievers



Team Year Fld SP RP Total
NYY 2007 696 419 497 1612
2012 673 462 503 1638
2017 705 457 520 1682
2022 682 451 518 1651
2027 694 461 508 1663
2032 694 442 491 1627 - Variance (07-32) +15p (+0.9%)

LAD 2007 645 404 480 1529
2012 709 409 497 1615
2017 722 435 501 1658
2022 660 427 480 1567
2027 702 426 504 1632
2032 713 419 507 1639 - Variance (07-32) +110p (+7.2%)

CHW 2007 661 413 472 1546
2012 668 395 480 1543
2017 643 401 482 1526
2022 663 399 483 1545
2027 668 408 498 1574
2032 675 460 499 1634 - Variance (07-32) +88p (+5.7%)

STL 2007 664 412 493 1569
2012 684 424 508 1616
2017 660 422 489 1571
2022 642 369 472 1483
2027 647 390 503 1540
2032 630 397 516 1543 - Variance (07-32) -26p (-1.7%)

TB 2007 645 411 493 1549
2012 652 416 485 1553
2017 640 410 491 1541
2022 649 431 485 1565
2027 672 395 472 1539
2032 638 425 500 1563 - Variance (07-32) +14p (+0.9%)

PIT 2007 666 405 482 1553
2012 669 417 487 1573
2017 650 427 517 1594
2022 660 448 504 1612
2027 614 431 482 1527
2032 648 374 488 1510 - Variance (07-32) -43p (-2.8%)


First, as we'll see in a minute, player quality did NOT go down as expected. It stabilized fairly rapidly at a higher level than 2007.

Second, and this may be critical for competitive balance questions, while it is clear the bigger market teams could afford to hire better players, the relationship between the haves and have nots does NOT destabilize as badly as we've assumed all along. In 2007, the difference between the Yanks and Devil Rays is 63 points. In 2032 the differnece is 64 points. I'll look at that in a moment, but first look at field players vs. relievers and see if there's a problem there.

*******



---- Year Fld SP RP Total
2007 3977 2464 2917 9358
2012 4055 2523 2960 9538
2017 4020 2552 3000 9572
2022 3956 2525 2942 9423
2027 3997 2511 2967 9475
2032 3998 2517 3001 9516

Field Variance: +21p (+0.5%)
Starter Variance: +53p (+2.2%)
Relief Variance: +84p (+2.9%)
Overall: +158p (+1.7%)



Despite earlier evidence i(ncluding from me), it appears clear that fictional players ARE able to develop just fine.

There does seem to be some favoritism towards the pitchers. It's not severe, but might be worth looking at at some point.

*******

Going back to competitive balance, since at first glance it appears few teams actually become MUCH better or MUCH worse over time (the Dodgers appear to be a big exception), we may have to look at other issues besides pure financials for the current imbalance BBM faces. Here's a look at each time, with first their '07 totals, then 2032.



Team 2007 2032 Dif Percent
LAD 1529 1639 110 +7.2%
DET 1577 1674 97 +6.2%

PHI 1575 1677 92 +5.8%
CHW 1546 1634 88 +5.7%
HOU 1531 1613 82 +5.4%
BAL 1557 1639 82 +5.3%

LAA 1573 1641 68 +4.3%
COL 1523 1588 65 +4.3%
WAS 1523 1588 65 +4.3%
NYM 1594 1660 66 +4.1%
SEA 1555 1619 64 +4.1%

FLA 1557 1618 61 +3.9%
KC 1494 1544 50 +3.3%
CLE 1552 1599 47 +3.0%
CHC 1571 1610 39 +2.5%
AZ 1562 1595 33 +2.1%

TOR 1589 1617 28 +1.8%
SF 1566 1592 26 +1.7% <--League Average
TX 1556 1581 25 +1.6%

NYY 1612 1627 15 +0.9%
TB 1549 1563 14 +0.9%
ATL 1560 1568 8 +0.5%
MIN 1585 1589 4 +0.3%
CIN 1509 1509 0 +0.0%

SD 1562 1554 -8 -0.5%
BOS 1576 1555 -21 -1.3%
STL 1569 1543 -26 -1.7%
PIT 1553 1510 -43 -2.8%
MIL 1566 1507 -59 -3.8%
OAK 1588 1474 -114 -7.2%




Highest Rated Team 2007: NYY 1612
Lowest Rated Team 2007: KC 1494
Difference: 118 (7.3%)

Highest Rated Team 2032: DET 1674
Lowest Rated Team 2032: OAK 1474
Difference: 200 (11.9%)


Only three teams, the Dodgers, Brewers and Athletics, break the League average by more than 5%. Though the difference between the 'haves' and 'have nots' does increase to troubling levels (an average of 11 points per person), for the most part the teams stay relatively stable. Further, some of the teams we'd expect to cause trouble like the Yankees and Red Sox...don't. There were certainly examples of player hoarding (80+ players in the minors), but that was by no means limited to big market teams.

Perhaps the answer to balance is as simple as fixing medical spending. Dunno.