Interaction Of Settings With Auto-Calc

[Brad K]

Interaction Of Settings With Auto-Calc


Introduction

Due to my research on auto-calc and pre-calc, which resulted in OOTP creating a new pre-calc file, I have had the opportunity to discuss settings with many people. In some cases, I have received game files from them that I could review. From this I have realized there are many misconceptions about the interaction of settings with auto-calc.

Auto-calc recommendation is presented as set it and forget it. In summary, auto-calc makes every save target the historic output of the selected year. It is highly successful. To do this auto-calc over rides ratings, alters the effect of some settings, and negates the effect of other settings. I have discovered this is not understood by many people.


League Strategy Settings

A large effect occurs with settings on MLB menu > League Settings > Stats & AI. Scrolling down to General Strategic Tendencies some of the settings starting with "Pitcher Stamina" don't work. Those known not to work due to auto-calc are marked in red.

Pitcher stamina is determined by the stamina setting calculated by Auto-Calc. The same is true for stealing bases, batter bunting, pitcher bunting, and catcher framing. The two settings for shifts aren't affected by specific modifiers but if they take away hits the auto=calc hit modifier over rides the effect. There may be an effect on the settings not marked in red but it has not been determined by testing,




Team Strategy Settings

Fortunately most of these work but not in the way expected. As an example, reducing SB attempt for a team will result in in making fewer attempts. But since auto-calc insists on historic league totals other teams will steal more. Yes, a team's strategy affects what other teams do.

The exception to this is when all GMs adopt the same setting, such as greatly reducing SB attempts. This could happen in an online league with all teams having human GMs. In that case, the effect of the setting is zero. Auto-calc will make every team steal enough to meet historic league numbers.

SBs are just an example. This applies to all strategies for which there is a modifier on the output. It also applies when there is an indirect modifier. For example, if increased shifts would reduce the number of hits, auto-calc compensates by increasing the hit modifier.





Talent Differences

There are many ways talent in a league can differ from historic. Some people turn recalc off and use development only. Talent Change Randomness is commonly used. Some people play with full minors which for several versions has allowed many career minor league pitchers to play in MLB. Three year and five year recalc change player ratings from historic for that year. And now many people are excited about the Development Lab.

It's exciting to send a player to the Lab and see his ratings improve. But whatever gains are made are taken from other players by auto-calc. The league overall must hit historic numbers. That can't be done if the average player gets better. The Lab increases talent disparity while keeping the average the same.


Conclusion

Those who want these settings to make a difference can still use auto-calc. The procedure to do so is to 1) Advance to opening day, 2) If auto-calc is on, turn it off, 3) Check to see if current year league totals have imported; if not do it manually, 4) Set everything discussed in this post to default (note a setting named "default" may not be the default for that year, and don't forget to check the strategy of every team), 5) Manually run auto-calc, 6) Set the league and team strategy as desired.

An alternative is to use the most recent pre-calc file which is contained in build 26.4.61. It works correctly if coaching is turned off. If coaching is desired then run auto-calc on a replay league of the desired year and use the modifiers it produces.

Regardless of how auto-calc is run it will still target historic output regardless of talent. Those who don't desire historic output but rather plausible output should use pre-calc.

[Brad K]

Note for those who have already read the first post, I originally forgot the Team Strategy section while including the screen shot. It has been added. I'm sorry if this caused any confusion.

[OutS|der]

So you're saying the Pitcher stamina under General Strategic Tendencies has no effect? Is this a new change because as of 24 it does have an effect as the higher the setting the longer starers will pitch in games. The higher the setting the longer all league starters last, this also leads to K increase for them

At the very least it has a direct impact on day to day game results in regards to league totals. Also "Auto-calc will make every team steal enough to meet historic league numbers." is not totally true. The game will also play with day to day results to achieve these numbers as well, IE you'll see more/less 5+ hit games and Shutouts and/or high K game totals.

Of course league totals are not set in stone. Auto-Cal is also just modifiers on a 3 year sim of your league, running it multiple times will get different results. As well it's generally run at the start of the season and doesn't take into account any talent changes that will happen over that given year. Also doesn't take into account any outlier seasons a player/team can have which can also change the end of league totals.

[Brad K]

Originally Posted by OutS|der

So you're saying the Pitcher stamina under General Strategic Tendencies has no effect? Is this a new change because as of 24 it does have an effect as the higher the setting the longer starers will pitch in games. The higher the setting the longer all league starters last, this also leads to K increase for them.

It's plenty easy to verify what I posted about stamina. Choose a season where stamina is very high or very low. Run auto-calc and note the stamina modifier.

Then change the League Strategy setting for stamina to the opposite extreme and run auto-calc again. The difference from the previous modifier is auto-calc over riding the strategy setting.

If that doesn't convince you run two seasons with similar changes to the SB attempt setting. It's a good one to choose because OOTP compiles the data for you on the accuracy page.

Please post your results. Thanks!

[Matt Arnold]

It's not that those settings don't work, it's that within the auto-calc, if it decides that the modifiers are getting out of line, it may adjust some of the strategy settings to compensate. So if it struggles to get an appropriate pitcher duration modifier, it may adjust the pitcher stamina strategy setting by a notch to better balance and get the accuracy it's aiming for.

[Brad K]

Originally Posted by Matt Arnold

It's not that those settings don't work, it's that within the auto-calc, if it decides that the modifiers are getting out of line, it may adjust some of the strategy settings to compensate. So if it struggles to get an appropriate pitcher duration modifier, it may adjust the pitcher stamina strategy setting by a notch to better balance and get the accuracy it's aiming for.

I'm sorry Matt, but a person who changes the strategy expects something different. And it's reasonable for him to get it. Would you consider testing to verify my statements? As I posted earlier, I suggest SBs because OOTP compiles the data.

[Pelican]

As to the user's expectations, I would fully expect that my choices/changes would be incorporated in a way that did not wreck annual totals and overall league stats to a degree that would distort the game. I always envisioned those sliders as nudging the sim in a preferred direction, understanding the game engine would have to compensate, maybe even pushing back. In a sense that's the game "saving me from myself" if I push too hard. I'm happy seeing starters who can go six innings instead of five, marginally more SB, and similar "tweaks". I appreciate being able to make "adjustments" to the game to bring it closer to what I want. That does seem to happen, although to Brad's point, I have not tested it against a control season with no adjustments. It could be my perception over reality, I admit.

[Brad K]

I urge everyone to test it for themselves. It doesn't take long.

[Brad K]

My posts are intended to increase knowledge of the game. There is no need for anyone to be defensive. State your views and present facts. We discover together.

I have selected SB attempts for this example because, as stated earlier, OOTP compiles the data on the accuracy page. Tests of other settings requires more time.

[Garlon]

In the first year you have let the modifiers autocalc based on those strategy settings and they then provide a modifier value to get the league into historic balance. The actual modifier number value will be very different whether you begin on rarely or very often.

The second time you changed it after the modifiers are calculated. Therefore you changed the underlying baseline variable for which the game had already adjusted for with the modifier and caused it to go out of balance.

There is nothing wrong with the behavior of the game here. You are just doing a combination of moves that causes this to happen. The assumption with autocalc is that you are not going to change the strategy settings after the fact but you can do it if you want but it defeats the purpose.

The SP stamina settings do matter in different eras because if you start to high or too low then the modifier will hit a floor or a ceiling where it cannot do any more work because it is maxed out basically. Therefore is a limit for the stamina modifier of 1.350. I try to select stamina strategy settings g that keep things as close to 1.000 as I can when used in conjunction with the modifiers.

[Brad K]

Originally Posted by Garlon

There is nothing wrong with the behavior of the game here. You are just doing a combination of moves that causes this to happen. The assumption with autocalc is that you are not going to change the strategy settings after the fact but you can do it if you want but it defeats the purpose.

I agree it defeats the purpose of auto-calc. That's what people who change strategy want to do. They want something different from their strategy change. Used as directed, auto-calc prevents them from getting what they want.

You say I did something that prevented auto-calc from doing its job. I did. I admit it. Auto-calc's job in this case is to over ride user strategy settings to achieve historic output. That's the problem I originally identified. It's established as true.

[Brad K]

Hey folks, don't take my word for it. This isn't a matter of faith. Test it for yourself. And if you want to look over my game files let me know.

[OutS|der]

Originally Posted by Brad K

I agree it defeats the purpose of auto-calc. That's what people who change strategy want to do. They want something different from their strategy change. Used as directed, auto-calc prevents them from getting what they want.
.

Auto-Cal is there to "control" league results to match the league totals.
If people want their strategy settings to drastically change final league totals then why would they run auto-cal after making those changes?

[kidd_05_u2]

Originally Posted by Brad K

I'm sorry Matt, but a person who changes the strategy expects something different. And it's reasonable for him to get it. Would you consider testing to verify my statements? As I posted earlier, I suggest SBs because OOTP compiles the data.

There is nothing to test.

You are just discovering how the settings work. The settings work as intended. And I would argue that the settings work as any reasonable person would expect them to.

[Brad K]

Originally Posted by kidd_05_u2

There is nothing to test.

You are just discovering how the settings work. The settings work as intended. And I would argue that the settings work as any reasonable person would expect them to.

You say I am discovering how the settings work. I agree. However I'm being told that what I've discovered isn't correct despite the evidence I've posted. That's why I asked others to test.

[Brad K]

Originally Posted by OutS|der

Auto-Cal is there to "control" league results to match the league totals.
If people want their strategy settings to drastically change final league totals then why would they run auto-cal after making those changes?

I agree they wouldn't do that. But people don't know. Auto-calc is presented as set it and forget and that's what people do.

I'm trying to get the information out that if they want their settings to matter they need to change settings after auto-calc runs or use pre-calc. While your statement indicates you understand auto-calc over rides settings, the majority of the posters in this thread think it doesn't.

[OutS|der]

Auto-Cal also is just the game simming the season 3 times and using modifiers to bring those league totals in line with the totals set. Run it again and you'll see different numbers.

I wish would could get away with league totals, I've never been a fan of it.

[Brad K]

Originally Posted by OutS|der

Auto-Cal also is just the game simming the season 3 times and using modifiers to bring those league totals in line with the totals set. Run it again and you'll see different numbers.

I wish would could get away with league totals, I've never been a fan of it.

Yes, the numbers change a bit. Sometimes more than a bit on the events that don't happen often or that are highly situation dependent.

Better would be an average of many runs, perhaps 50, put into the pre-calc file. I suggested that in the discussion that resulted in the creation of a new pre-calc file.

I don't move saves so I'm playing on 25 and testing on 26. I'm making my own pre-calc mods for 25.

[turtle4499]

Originally Posted by Brad K

I'm sorry Matt, but a person who changes the strategy expects something different. And it's reasonable for him to get it. Would you consider testing to verify my statements? As I posted earlier, I suggest SBs because OOTP compiles the data.

I think that what those modifiers do is not even remotely obvious but they do have an effect even with autocalc. Though TBH the difference in the actual effect when you do or do not use autocalc is 100% not something I think anyone would really expect.


Taking SBs as the easy example, autocalc will target the total number of SBs for that year, the modifier will increase the number of attempts thus autocalc will decrease stolen base success chance. If you do not run autocalc again it will just increase the number of stolen attempts without changing the odds resulting in more stolen bases. Catcher framing impact changes the level of difference between being a 10 framing and an 80 framing. i.e. it really just determines how bad is a bad catcher. This one is easy to observe even just on catcher war.


Pitcher stamina is the only one I am really not sure about, my honest guess is it has a similar effect to catcher framing impact where it just changes the difference in having high stamina vs low stamina. IRC number of relievers is what changes how quickly a starter fatigues as a general thing.


Also TBH that entire section of the game is absolute freaking nightmare fuel in terms of clicking a button and causing crazy stuff to happen. Changing rotation size changes how many days it takes for stamina to reset on a starter lol. I think maybe just adding a checkbox that determines if changing the setting also changes league total modifiers or not would make this less of an issue.


I assume the a million year old bug with autocalc is still there, it doesn't actually use order of events for batting and can result in producing some absolutely insane results because of that. The most obvious place this used to occur was in the DSL where walk rate would go insane and break everything else.


One thing I have never actually checked before, how does expansion impact autocalc?

[Brad K]

Originally Posted by turtle4499

Taking SBs as the easy example, autocalc will target the total number of SBs for that year, the modifier will increase the number of attempts thus autocalc will decrease stolen base success chance.

The second screen shot in post 9 shows SB success declining with increased attempts, and vice versa, as you say. That test was run with pre-calc.

The first screen shot in post 9 shows that auto-calc over rides strategy settings. It does not allow the user to experience the expected results of his settings.

Pitcher stamina is the only one I am really not sure about, my honest guess is it has a similar effect to catcher framing impact where it just changes the difference in having high stamina vs low stamina. IRC number of relievers is what changes how quickly a starter fatigues as a general thing.

I want to research pitcher stamina as both you and another poster have mentioned it. The question is what data to use. Data on Complete Games is compiled by OOTP and would seem to be related.

However when I ran five years on a replay with auto-calc the OOTP Complete Game count was 19.5% below historic. So the game even with auto-calc isn't dialed into reality on that stat.

It might have been usable if the error was consistent, but it wasn't. Annual differences varied from - 10% to - 27.3%. It seems like it would take a lot of runs being averaged to have any hope of a consistent error.

One thing I have never actually checked before, how does expansion impact autocalc?

It doesn't do anything. Auto-calc will always produce very close to historic output regardless of the input. It is designed to over ride everything. It is very successful.