Steps for getting more realistic stats in your first season

[SteveP]

The purpose of this post is to describe the process I use for getting correct LTMs for the initial season of a league (or for any single season replay). It uses an undocumented "feature" of the LTM adjustment process that even some veteran players may be unaware of.

In the League Setup portion of Game Setup, on the Strategy Page, there are League Totals (LTs) and League Total Modifiers (LTMs). Please read up about these in the manual, if you are unfamiliar with them and how the game uses them (also how the game adjusts them before each new season).

The LTMs always start out at 1.000 in a new league (this hasn't changed in OOTP10), which means that the stats produced in the first season (and, to some small extent, even in a second season) are likely to be unrealistic. Some of the worst problems in this regard (e.g., SBAs) have been improved in OOTP10, but the problem still exists. The standard recommendation has been that people sim out a couple of seasons (so that the automatic LTM adjustment process can kick in) before they really start playing, but many OOTP gamers don't want to do that. I know I didn't.

Here are the steps:

1. Create your league. I suggest using some other name for your league than the one you ultimately want (because you will be discarding this league eventually and you don't want it pre-empting the name you want to use). Get the league set up for play to your satisfaction. Save it.

2. Create a copy of your league. One way to do this is by using the "Make Backup" function in the game menu. This function makes a complete copy of your game, using the same name, but puts it in a different folder from Saved Games. The other way to do it is to go out to your system and make a duplicate of your league folder in Saved Games (which is why you want to save your league before making a copy). This copy is the version of your league that you will actually end up playing, so be sure you have copied it fully.

3. With the original version of your game still loaded, go to the League Setup Strategy Page and uncheck the "Automatically adjust league totals modifiers after each season for historical accuracy" box. If it was already unchecked for some reason, leave it that way.

4. Select the option which directs OOTP to Auto-play (sim) a complete season until "... PLAYOFFS BEGIN". Since this version of your league is only for the purpose of gathering LTM-related data for your real game, there are a lot of steps you can use to speed up this sim. For example, in the Game Preferences, set Auto-play display to "simple dialog". Turn off injuries, trading, financials, Spring training, all-star games, drafting and anything else not necessary for producing stats. I also find it is faster if I make myself commissioner only, and not a team manager as well.

5. When Auto-play is done, go to the League Setup Strategy page. You will still see 1.000 in all the LTM boxes. At this point check the "Automatically adjust league totals modifiers after each season for historical accuracy". You will see the LTMs change in almost all of the boxes. What has happened is that the game has calculated the LTMs that should have been used to get more accurate results for the season you just simmed. Those are the numbers you want for the league you intend to play. Write them down, making sure that you keep track of which numbers go in which boxes. (Side note: if by some chance you simmed this season but forgot to uncheck the "Automatically adjust ..." box first, it's OK. Just uncheck it, sim one more day, then recheck it).

6. Load the copy of your league, giving it whatever name you want (this is pretty easy with the Restore Backup function). Go to the League Setup Strategy page, and replace the 1.000 LTMs with the numbers you wrote down. Also, check the "Automatically adjust league totals modifiers after each season for historical accuracy" box if it is unchecked. Save. Make a backup (just because it's a good idea). You are ready to play. (Side note: if you want to be really thorough, enter these same LTMs into the appropriate boxes for each of your minor leagues. This is what the game does automatically in the LTM adjustment process. You will get better stats in your minor leagues, though they will still be more unrealistic than many gamers would like. Nothing I can do about that. )

With a fast sim, all this can be done pretty quickly (30-40 minutes perhaps). As a result, I think there is no reason why OOTP gamers should have to compromise on what they get from that first season.

Note for OOTP veterans: the LTMs that you get from this technique are not the same as the adjusted LTMs that the game automatically creates at the start of each new preseason. The latter are calculated based on the previous season's results but adjusted for the coming season's historical stats. You can check this out by getting LTMs using the technique above, then simming forward to the next preseason. You will see that the LTMs have changed. In contrast, the LTMs you get from this technique produce realistic stats for the season just simmed if you use them in a cloned version of that season.

As noted in later posts in this thread, this technique does not seem to work correctly for errors. I don't know why, and perhaps in some future patch this will be fixed. The LTMs you get for errors are better than the 1.000 default settings, because you will get a better distribution of errors by position. But league/team totals for errors will likely be off from historical stats. By checking the historical record, you can calculate a manual adjustment for yourself, if you wish.

[pstrickert]

How hard would it be . . .

1. to do this for every historical season (one at a time, of course)
2. create a file with the resulting modifiers
3. program OOTP to use this file to set LTMs for the initial import season

?

[pstrickert]

Or if all else fails . . . couldn't we create such a file and make it available for download? Then the user could modify the LTMs by hand during setup.

[old timer]

SteveP:

Thanks for the info! I just tried it in a 1974 replay and the league totals are much better compared to before and are generally quite close (I did multiple replays) to RL.

pstrickert:

It would be nice to have templates for every season so if one wants to do a single season replay, it would be quick and easy. I just did a 1974 one and it's working really well.

The difficulty (besides the tedius part of actually doing all of those years) would be finding league settings that most people would want to use. I've fooled with the settings a lot and have found ones that I think work well for replays, but someone else might not like them.

[SteveP]

Originally Posted by pstrickert

How hard would it be . . .

1. to do this for every historical season (one at a time, of course)
2. create a file with the resulting modifiers
3. program OOTP to use this file to set LTMs for the initial import season

?

I think that most people who might consider trying to do this, in a way that was widely usable and reasonably valid, would decide it was way too much work, and that it would be a heck of a lot easier to just sticky this thread.

[pstrickert]

Originally Posted by old timer

SteveP:

Thanks for the info! I just tried it in a 1974 replay and the league totals are much better compared to before and are generally quite close (I did multiple replays) to RL.

1974. I've simmed it countless times. Here's what I've noticed:

1. CAL overperforms, often winning the division title.
2. OAK and BAL and LA seldom approach their RL team ERAs.
3. Each team's FldPct and RF is wildly inconsistent from one sim to the next.

The major problem is defense, I believe. Maybe there's not enough difference between the best and worst defensive teams. Or maybe play outcomes do not take defense into consideration enough. Or the formula for calculating defensive ratings is inadequate.

Does this match up with your experience, old timer?

[old timer]

Originally Posted by pstrickert

1974. I've simmed it countless times. Here's what I've noticed:

1. CAL overperforms, often winning the division title.
2. OAK and BAL and LA seldom approach their RL team ERAs.
3. Each team's FldPct and RF is wildly inconsistent from one sim to the next.

The major problem is defense, I believe. Maybe there's not enough difference between the best and worst defensive teams. Or maybe play outcomes do not take defense into consideration enough. Or the formula for calculating defensive ratings is inadequate.

Does this match up with your experience, old timer?

Actually with v10, those numbers are pretty consistent now over 6 different sims.

One team whose FPct varied quite a bit from RL was the NYY. I looked over their fielding stats to see why. It turns out, the only reason the simulated Yanks did better was because they used better defenders (based on FPct) than did the RL Yanks. The numbers of those who played were very similar to RL.

I have injuries turned off, so players who were injured in RL and who play regularly in the sim will sometimes significantly affect the team results and record. Also, player usage and lineup orders make a big difference as well.

[PhillieFever]

Damn Steve,I wish you would have posted this earlier,you would have saved me a ton of work!LOL Awesome find bro.

[pstrickert]

Originally Posted by old timer

Actually with v10, those numbers are pretty consistent now over 6 different sims.

One team whose FPct varied quite a bit from RL was the NYY. I looked over their fielding stats to see why. It turns out, the only reason the simulated Yanks did better was because they used better defenders (based on FPct) than did the RL Yanks. The numbers of those who played were very similar to RL.

I have injuries turned off, so players who were injured in RL and who play regularly in the sim will sometimes significantly affect the team results and record. Also, player usage and lineup orders make a big difference as well.

OAK team ERA in real life: 2.96
BAL team ERA in real life: 3.28
American League ERA: 3.63
LAD team ERA in real life: 2.98
National League ERA: 3.63

In your sims, is the BAL team ERA ever .35 lower than the league ERA? Is the OAK team ERA ever .67 lower than the league ERA? Is the LAD team ERA ever .65 lower than the league ERA?

In my experience, OAK, BAL, and LAD come near and sometimes exceed the league ERA. But I'm not sure I've ever seen them match their real life ERA.

As for NYY, they do import with a lot of blue fielding ratings (18-20 on a scale of 20). But I'm sure sure all those players merit the high fielding ratings. I'll check it out some more.

[pstrickert]

Rtot (fielding runs above average) - 1974

BOS: 46.1
BAL: 43.6
OAK: 38.1
CLE: 37.0
NYY: 21.7
MIL: 18.8

(every other team is a negative number)

If the defensive model is not flawed in OOTP (though it may be), we would expect to see lower team ERA or R/G for OAK and BAL. But I'm not seeing it when I sim. Top-notch pitching staffs + top-notch defense, as I've said before, does not "translate" adequately into the OOTP world.

[pstrickert]

How about the poor Atlanta Braves? In all of my 1974 sims (100 over the past 2 years), I don't think ATL has ever won the NL West. Part of the problem is that CIN and LAD had terrific seasons. Often, one or the other will be at the top. But ATL had a 3.06 team ERA in 1974 and led the NL in Rtot with 66.7. Oops! I forgot! Top-notch pitching + top-notch defense . . . it doesn't translate too well into the OOTP world.

[old timer]

I'm starting to think that a major problem with regards to pitching not matching is that OOTP doesn't seem to be able to model certain types of pitchers. For example, look at any guy with significantly less hits allowed than IP and if his strikeout totals are low, OOTP can't seem to correctly model him.

Take Phil Niekro and Catfish Hunter. If you go to edit their ratings, you'll see that in order to get the strike out totals to match, the hits allowed totals will be much higher than in RL. If you can edit Phil Niekro and get his expected totals for hits allowed and strike outs to come even close to his 1974 performance, let me know how you did it.

[old timer]

Originally Posted by pstrickert

OAK team ERA in real life: 2.96
BAL team ERA in real life: 3.28
American League ERA: 3.63
LAD team ERA in real life: 2.98
National League ERA: 3.63

In your sims, is the BAL team ERA ever .35 lower than the league ERA? Is the OAK team ERA ever .67 lower than the league ERA? Is the LAD team ERA ever .65 lower than the league ERA?

In my experience, OAK, BAL, and LAD come near and sometimes exceed the league ERA. But I'm not sure I've ever seen them match their real life ERA.

As for NYY, they do import with a lot of blue fielding ratings (18-20 on a scale of 20). But I'm sure sure all those players merit the high fielding ratings. I'll check it out some more.

No, none of those teams have performed like that in my sims either. See my post above.

[old timer]

Find anyone with hits allowed totals that are quite low compared to their IP and unless their strike out totals are very high (yes, I know that's pretty vague), they will not consistently perform well. Their ratings will make their strike out totals match up, but not their HA.

Take Gaylord Perry in '74 or Jim Palmer, Hunter or Niekro in most any year and see how different they perform. They all seem to suffer from this problem.

For more examples, look at Sutton and Messersmith of the Dodgers and compare their '74 RL HA and K totals to your sim results and you'll see what I mean. Their K totals seem to turn out well (they should, if you check their ratings the K totals are good, but not the HA), but not their HA totals.

[pstrickert]

Is HA related to defense, though? If Niekro or Hunter or Messersmith were backed by a stellar defense (as they were), they would give up less hits on balls put into play. Good defense up the middle (e.g., Belanger, Grich, Blair or Campaneris, Green, North) should make a significant difference in HA (and, presumably, ERA).

[Garlon]

Pitchers are rated for strikeouts in terms of K's per 550 Batters Faced. There is no reason you would want to edit their ratings. OOTP assigns ratings appropriately.

OOTP does just fine modeling pitchers.

Catfish Hunter
Stat: Real vs Sim
IP: 3449 vs 3403
K per 1000 BF: 143.38 vs 143.76
BB per 1000 BF: 67.98 vs 68.61
HR Allowed per 1000 BF: 26.65 vs 24.51
HBP per 1000 BF: 3.49 vs 3.01

[old timer]

Originally Posted by Garlon

Pitchers are rated for strikeouts in terms of K's per 550 Batters Faced. There is no reason you would want to edit their ratings. OOTP assigns ratings appropriately.

OOTP does just fine modeling pitchers.

Catfish Hunter
Stat: Real vs Sim
IP: 3449 vs 3403
K per 1000 BF: 143.38 vs 143.76
BB per 1000 BF: 67.98 vs 68.61
HR Allowed per 1000 BF: 26.65 vs 24.51
HBP per 1000 BF: 3.49 vs 3.01

What were his hits allowed? That's what's always off in my sims. The stats you mentioned are always close in my sims too.

[old timer]

Originally Posted by Garlon

Pitchers are rated for strikeouts in terms of K's per 550 Batters Faced. There is no reason you would want to edit their ratings. OOTP assigns ratings appropriately.

Did you check the ratings for the pitchers I mentioned? They are usually way off in HA. No amount of editing (at least as far as I can tell) will fix it either.

[old timer]

After more checking, it seems that the situation is worse for simming a single season (even using the league totals modifiers mentioned above) compared to simming multiple seasons. I'll have to do more testing to see if this is the case.

I haven't simmed multiple season with OOTP X yet, so if someone has, can you check Hunter, Palmer and Phil Niekro and see what their hits allowed totals are compared to RL over their whole career?

[old timer]

Originally Posted by pstrickert

Is HA related to defense, though? If Niekro or Hunter or Messersmith were backed by a stellar defense (as they were), they would give up less hits on balls put into play. Good defense up the middle (e.g., Belanger, Grich, Blair or Campaneris, Green, North) should make a significant difference in HA (and, presumably, ERA).

I've only checked one sim so far, but in this one those players did pretty well defensively. Campaneris was exceptional and Green and North were pretty much the same. Grich was much lower in RF, but the others you mentioned were close. In fact, the A's and O's defenses were pretty close to the same, but pitching was worse. The Dodgers did suffer defensively in this sim however and they also did much worse than in RL (with regards to W-L record).