50 Homers in dead ball era? [Fixed 9.39]

[kalasisking]

This is my first post here, so I hope I'm not trodding over well worn ground with this. I just got the game a few days ago. With version 9.01, I simmed the 1927 season, and with the exception of Ruth and Gehrig, who both hit about 25 homers, everyone else hit no more than 10. I figured it needed a little tweaking, but overall, no big deal.

I just installed 9.38 and simmed from 1901 to 1910. Every year there were guys hitting 40+ homeruns, so I suppose the pendulum has swung back with this issue. Or I could be doing something wrong.

Any thoughts? And again, I hope I'm not the 5000th person to bring this up.

Steve

[HoustonGM]

I dont think the game has an internal engine for each different year. It used to with the stats.txt files (which you can still view in the INPUT folder), but thats now been moved to Simulation Settings within the game. What I do when I sim historically, is I open up the stats.txt file that pertains to hte year I'm in, and edit the Simulation Settings to what the stats file says. I do this every yea.r

[Jeff Olsen]

I dont think the game has an internal engine for each different year.Not each year, no, but it does have an internal engine for pre-WWII.

[kalasisking]

Thanks for the help - I'll check the stats.txt file if I ever try to replay the 20th century. Now that I think of it though - what are the settings going forward?

Now that the steroid era (and possibly the HGH era) is over, will that be reflected in more subdued offensive numbers in the coming years? For my taste, baseball was at its best in the mid 80's where you had a true balance of offense, pitching, and plenty of homers and stolen bases. It was a golden time. (Of course, I was also young and not at all worn out by the demands of life then, so maybe my sepia toned vision of baseball ala '86 is a bit fuzzy?)

[HoustonGM]

Thanks for the help - I'll check the stats.txt file if I ever try to replay the 20th century. Now that I think of it though - what are the settings going forward?
The stats files only go up to 2005. For years after 2005, BB2k6 used 2005 stats files, although you could create your own.

[Clay Dreslough]

FYI, the 19**stats.txt files are no longer used to adjust simulation results in Baseball Mogul (actually they are used, but more as a set of reverse-calibration numbers -- in other words if you RAISE the 'Singles' number in 1901stats.txt it will actually LOWER the number of singles -- although the exact effect is basically undefined at this point).

If you want to tweak simulation results, I would recommend altering the Simulation Settings every few years.

However, I am already aware of this problem. I was just noticing that homers-per-team are about twice as high in the 1901-1905 time period as in real life (although I didn't see anyone hit 40).

I would like for Baseball Mogul to provide more accurate results BY DEFAULT in any year you play. That is, you shouldn't have to edit a file just to get historical realism. And I should be able to get this realism from just properly analyzing the Lahman Data. It's tricky because I have to make sure pitching, hitting and defensive ratings all make sense while still providing results that are appropriate to the era. But I'm confident I can make it work without requiring that everyone edit files to get the best results.

To sum up, "historical accuracy by default" and "roster AI" are the two major issues I think I can address before Version 9.50. That is, I plan to address these in a free patch for Baseball Mogul 2007 in July or August -- instead of putting them off until next year.

But I won't be addressing them in the next few days because I want to post an official patch with the improvements we've made so far. And I don't want to break anything by messing with the engine or AI.

Clay

[kalasisking]

I dont think the game has an internal engine for each different year. It used to with the stats.txt files (which you can still view in the INPUT folder), but thats now been moved to Simulation Settings within the game. What I do when I sim historically, is I open up the stats.txt file that pertains to hte year I'm in, and edit the Simulation Settings to what the stats file says. I do this every yea.r

I just simmed through 1901 and 1902 using your tip and it worked out great. Just one quickie though - those files are say 1900stats thru 2005 stats. Does that mean that when replaying the 1901 season I use the file that says 1900 stats? I'm pretty sure it's yes, but I just want to make sure. Thanks for the help, guys.

And Clay - I look forward to the changes later this summer. I appreciate your dedication to improving this game.

[HoustonGM]

Actually, I think you can just skip over 1900 and start with the 1901 file.

[adam]

I've got a solution idea, but it'd probably have to wait until 2k8, or when clay or a researcher has a lot of time.

I think there should be two sets of ratings pre-1945--an average set, and a leaders set. The leaders were far more above the average than it is today, and that could accurately reflect both levels.

[Clay Dreslough]

Does that mean that when replaying the 1901 season I use the file that says 1900 stats? I'm pretty sure it's yes, but I just want to make sure. Thanks for the help, guys.Actually the 1901 stats file indicates the totals for the 1901 season.

Clay

[SirKodiak]

I've got a solution idea, but it'd probably have to wait until 2k8, or when clay or a researcher has a lot of time.

I think there should be two sets of ratings pre-1945--an average set, and a leaders set. The leaders were far more above the average than it is today, and that could accurately reflect both levels.
Player ratings should handle that though (if based on stats).

As for rookies coming into the league, if the distribution that you propose is true, a single cubic equation with 2 horizontal tangents could mimic this (giving a bell-like curve for normal players, a number of 'leaders', and few in between).

EDIT: Now that I think about it, a sine graph would be a good option too. Better 'bell curve' but cannot be manipulated as well. You could find the cubic equation by plotting number of players vs rating and doing a cubic regression, while making a sine -based plot would be guess work.

[kalasisking]

Actually, I think you can just skip over 1900 and start with the 1901 file.

For some reason I had it in my head that if I played the 1901 season I needed the 1900 modifiers to get the correct stats. Not quite thinking clearly there.

Forgot to ask though - what are the other 2 numbers in the stats file - "Rot" and "Tal"?

[HoustonGM]

Rot is Rotation, and the game automatically does that now.
Tal is talent, and I'm not qutie sure what that was ever used for.

[Clay Dreslough]

Rot is Rotation, and the game automatically does that now.
Tal is talent, and I'm not qutie sure what that was ever used for.
Talent is basically the average talent level, as compared to the superstars. It serves the function that 'adam' mentioned -- keeping talent levels more varied pre-WWII than they are post-WWII.

For example, 'Talent' is 784 (out of 1000) in 1917. That means average talent is low compared to stars. So players like Babe Ruth can blow away the competition by hitting more homers than an entire team.

'Talent' is 956 (out of 1000) in 2005. Average talent is higher (better training, more countries to draw from, no segregation). So Albert Pujols isn't going to dominate the way Ruth did, even though he's the best hitter in the league.

BTW, these numbers may only be used internally at this point. I'm not sure exactly what happens if you change them.

Clay