Home run modifiers?

[Sweed]

Originally Posted by locuspc

I can never get league totals to work right. If I want to change things I change the modifiers directly, whenever I mess with the league totals things go screwy.

Well, that is the way it is designed to work. I mean, if you change the LTs you now are asking them to work with LTMs that were set to work with the previous set of LTs. More on the old way of changing LT's below and how counterintuitive it was.

Before autocalc one had to do some math to figure out what to set the league totals too so you could get the "right" stats as there were no League Total Modifiers. Autocalc is what enabled one to change LT to whatever they wanted and then autocalc would use those totals to set the modifiers for you.

In those old days if one wanted more homeruns he had to lower the League Total while leaving the expected At Bats the same. I know, counterintuitive, right? Now if you were making changes to all of the LTs? That took even more work. League Total Modifiers were created to do the heavy lifting for you.

I honestly read this, and many other threads, and am amazed that folks post they don't work for them. I've used them since introduction and have never had an issue. I'm almost done with my current season and most stats are exactly in line with my LTs. I've been waiting to post in this thread when that happens and add some screen shots to "show my work".

Some have said checking the autocalc to run automatically doesn't work or provides bad outcomes. Others have chimed in that manually running autocalc does work while the checkbox version does not. All I can say is I run it manually for my MLB and each MiLB on their respective opening days and have not had problems.

Many comment on HR leaders but other than saying that provide little data. Such as..

did the league homeruns come out to expected results while the leaders were way too high? IOW someone hits 80-90 to lead the league but the league as a whole hit the 4500 (throwing out a number) that was expected. If that's the case then I think you have a ratings distribution problem and not an autocalc problem. League Totals coming out "right" means AC did it's job.

As an example my league HRs should be 11.5% of hits with the way my LTs are set. My team just played game #148 with the rest somewhere around there +\- 3 or so. My HR's today, league wide, are 11.15% of hits. My leader (scouted 7/8 power) today is at 42 and projects to end at 47. Four guys follow that should be between 38-44. If my leader was at 75 and going to 80ish that would point to...
autocalc doing it's job and player ratings being out of whack. Unless this guy was a 1 in a million. But those in these posts suggest it happens "all of the time" or at least "a lot". IOW too many times for it to be a one in a million.

FWIW my league is imported from version to version since v4. So, my players were, for the most part, not created in v23. And those that were are not at the MLB level. What that means is if there is a player creation problem (some guys getting too much power or, whatever else) I wouldn't be seeing it. Is my not seeing this issue another clue? IE if my HR leaders are where they should be, and v23 created leagues have HR hitters that are not, is it a v23 creation problem?

My suggestion\request would be for some of you to post data that we can all look at and try to see where it's going wrong for you. Without seeing screenshots of your LT\LTM screen, final stats for league wide batting, and a list of league leaders one can only guess at the cause. Or what is, or is not working.

If you don't want to post those items at least look at them to see how it all fits together. You may, or may not find an answer.

I play at a slow pace so it may be a bit before I get to the regular season's end. When I do I'll be back with relevant screen shots.

[koohead]

Originally Posted by Sweed

Well, that is the way it is designed to work. I mean, if you change the LTs you now are asking them to work with LTMs that were set to work with the previous set of LTs. More on the old way of changing LT's below and how counterintuitive it was.

Before autocalc one had to do some math to figure out what to set the league totals too so you could get the "right" stats as there were no League Total Modifiers. Autocalc is what enabled one to change LT to whatever they wanted and then autocalc would use those totals to set the modifiers for you.

In those old days if one wanted more homeruns he had to lower the League Total while leaving the expected At Bats the same. I know, counterintuitive, right? Now if you were making changes to all of the LTs? That took even more work. League Total Modifiers were created to do the heavy lifting for you.

I honestly read this, and many other threads, and am amazed that folks post they don't work for them. I've used them since introduction and have never had an issue. I'm almost done with my current season and most stats are exactly in line with my LTs. I've been waiting to post in this thread when that happens and add some screen shots to "show my work".

Some have said checking the autocalc to run automatically doesn't work or provides bad outcomes. Others have chimed in that manually running autocalc does work while the checkbox version does not. All I can say is I run it manually for my MLB and each MiLB on their respective opening days and have not had problems.

Many comment on HR leaders but other than saying that provide little data. Such as..

did the league homeruns come out to expected results while the leaders were way too high? IOW someone hits 80-90 to lead the league but the league as a whole hit the 4500 (throwing out a number) that was expected. If that's the case then I think you have a ratings distribution problem and not an autocalc problem. League Totals coming out "right" means AC did it's job.

As an example my league HRs should be 11.5% of hits with the way my LTs are set. My team just played game #148 with the rest somewhere around there +\- 3 or so. My HR's today, league wide, are 11.15% of hits. My leader (scouted 7/8 power) today is at 42 and projects to end at 47. Four guys follow that should be between 38-44. If my leader was at 75 and going to 80ish that would point to...
autocalc doing it's job and player ratings being out of whack. Unless this guy was a 1 in a million. But those in these posts suggest it happens "all of the time" or at least "a lot". IOW too many times for it to be a one in a million.

FWIW my league is imported from version to version since v4. So, my players were, for the most part, not created in v23. And those that were are not at the MLB level. What that means is if there is a player creation problem (some guys getting too much power or, whatever else) I wouldn't be seeing it. Is my not seeing this issue another clue? IE if my HR leaders are where they should be, and v23 created leagues have HR hitters that are not, is it a v23 creation problem?

My suggestion\request would be for some of you to post data that we can all look at and try to see where it's going wrong for you. Without seeing screenshots of your LT\LTM screen, final stats for league wide batting, and a list of league leaders one can only guess at the cause. Or what is, or is not working.

If you don't want to post those items at least look at them to see how it all fits together. You may, or may not find an answer.

I play at a slow pace so it may be a bit before I get to the regular season's end. When I do I'll be back with relevant screen shots.

Fair enough. And yes, up until '23, I have had no problems with totals or modifiers or whatever. Was always happy. But after 2 seasons with 23 using the exact same totals and modifiers as I did in 22, I get wonky results. Here is one league example...final team totals,along with the league totals and modifiers used for this league (with auto-calc enabled), and player stats sorted by HR totals

[Sweed]

Originally Posted by koohead

Fair enough. And yes, up until '23, I have had no problems with totals or modifiers or whatever. Was always happy. But after 2 seasons with 23 using the exact same totals and modifiers as I did in 22, I get wonky results. Here is one league example...final team totals,along with the league totals and modifiers used for this league (with auto-calc enabled), and player stats sorted by HR totals

Thanks for the screens. Question, why would you use the same LTM's as v22? You never ran autocalc to set LTMs for v23? There are always engine changes between versions and v23 was a massive rewrite. I don't see how numbers that worked in v22 would work in v23. LTs? Yes. LTM? No.

I will share some screens that show what I think is important to see for comparison of LTs vs. resulting stats, and HR leaders since that is the focus of this topic. I'll also add a bit of how my LTs compare with the stats generated using autocalc that was manually run on opening day after the MLB rosters were set.

Context: 144 games played by my team. (sorry, I know I said 148 in earlier post but that was from memory, game not open ).

Three slash stats..
LT .266/.338/.422 My league .256/.327/.403
That is the farthest I can recall ever seeing it between expected output and actual output in my game. Still realistic but in previous versions I don't think I was more than 0.050 off in batting average where, in v23 as we can see it is 0.10. This seems to carry on, as expected, to affect OBP and Slg.

Home Runs..
LT setting HR are 11.49% of total hits
My game HR are 11.16 % of total hits

and to carry onto a couple more to show how close things are..

Doubles
LT 20.03% of hits
My game 19.98 %

Triples
LT 2.01 % of total hits
My game 2.98%

Even with batting average coming in 0.100 lower than expected the ratio of 2b, 3b, and HR to total hits is spot on. Indicates autocalc is working.

You'll see my HR leaders are in a believable pack. My HR totals, for the entire league, are running right in line with expected output. This all indicates to me that autocalc is working in my league. If autocalc wasn't working then the output for the entire league would be off.

So, if I had a 90 HR guy AND the HR total, again league wide, was too high that would indicate autocalc was not working.

But if I have a 90 HR guy and the league HR total is correct that, in my mind, would indicate said player has a HR with either a "too high" rating or maybe the engine is off internally when calculating HRs for players at the very top of the scale. A 90 HR guy does not indicate that autocalc isn't working as has been suggested here.

That is my theory. I could be wrong but there's no way to know if users making the assertion that autocalc isn't working are not providing data. Remember as I posted earlier the vast majority of my players were created in previous versions of OOTP. So there is that to consider when looking at my HR leaders. But again, that in my mind, would go to the HR issue being a player creation issue and not an autocalc issue.

And again, I could be wrong but, I only have my game and the stats it produces to go by.

[koohead]

Originally Posted by Sweed

Thanks for the screens. Question, why would you use the same LTM's as v22? You never ran autocalc to set LTMs for v23? There are always engine changes between versions and v23 was a massive rewrite. I don't see how numbers that worked in v22 would work in v23. LTs? Yes. LTM? No.

My bad...I carried the same totals from 22 and have let Auto-Calc set the modifiers (as I always have) so you are correct, those are not the same modifiers as what was used in 22.

And I would absolutely agree that LT and LTM are working for the ML, and even Triple A and probably Double A. It's when you get into the lower minors that the HR totals start getting off. Interestingly the associated batting averages don't seem that far off. some comparison numbers

ML: League totals would forecast .1111 HR per Hit...so far (156 games out of 162 this season) I'm at 4298 HR over 40403 Hits, or a .1063 HR per Hit pace.

In the LowA league that I provided screen shots for earlier, the League totals would forecast .0747 HR per hit. Based on current year stats (138 games in) I'm at 3577 HR against 17096 Hits, or a .2092 HR per Hit pace.

I know that 1.9xx LTM for HR is what is throwing things off, but why is autocalc messing it up only for the lower minors but yet pretty accurate and working for ML and higher minors is my question.

[koohead]

Out of curiosity I ran the numbers for the rest of my leagues. there is a problem with all of the minors, though it seems to get worse the lower you go. Here is what I ended up with. The columns are the league level, the HR and Hit numbers entered into league totals, the ratio expected based on league totals, then the HR and Hit totals for current season (end of the season) and the calculated ratio.

ML: 4700/42295 11.11% 4298/40403 10.64%
Triple A 1: 1680/17518 9.59% 2054/17903 11.47%
Triple A 2: 2228/22173 10.05% 2261/20114 11.24%
Double A 1: 1291/14690 8.79% 1881/14629 12.86%
Double A 2: 937/11977 7.82% 1559/11856 13.15%
Double A 3: 913/9814 9.30% 1160/9911 11.70%
HiA 1: 1298/13377 9.70% 1833/12245 14.97%
HiA 2: 741/9399 7.88% 1574/9479 16.61%
HiA 3: 929/14018 6.63% 2415/14380 16.79%
LoA 1: 1375/18810 7.31% 3736/18962 19.70%
LoA 2: 1257/16823 7.47% 3577/17096 20.92%
SS 1: 322/5188 6.21% 781/4387 17.80%
SS 2: 491/5879 8.35% 1027/4965 20.68%
RL 1: 325/6484 5.01% 1495/5947 25.14%
RL 2: 342/6834 5.00% 1594/6727 23.70%

[koohead]

Originally Posted by Sweed

I'll also add a bit of how my LTs compare with the stats generated using autocalc that was manually run on opening day after the MLB rosters were set.

Question: Do you MANUALLY hit the adjust league total modifiers button every year even with the auto-calc check box checked? I understand the need to do it on opening day when rosters are set...but thought that was what was being done by the Automatically adjust LTM for accuracy check box. And I wonder if the reason it is off for the lower minors is that on Opening Day of the ML, the lower levels rosters are not set by any stretch of the imagination, while the higher you go the more set the roster is. Kind of assumed each league would run the auto adjust modifier on its respected opening day...but perhaps that's where things are going bonkers.

[koohead]

couple interesting comments from luckyman and Matt A in this thread regarding Minors and LTM...maybe something is different in the engine between 23 and 22 so will give these suggestions a shot for the following season and report back to y'all.

https://forums.ootpdevelopments.com/...just+modifiers

[Sweed]

Originally Posted by koohead

My bad...I carried the same totals from 22 and have let Auto-Calc set the modifiers (as I always have) so you are correct, those are not the same modifiers as what was used in 22.

And I would absolutely agree that LT and LTM are working for the ML, and even Triple A and probably Double A. It's when you get into the lower minors that the HR totals start getting off. Interestingly the associated batting averages don't seem that far off. some comparison numbers

ML: League totals would forecast .1111 HR per Hit...so far (156 games out of 162 this season) I'm at 4298 HR over 40403 Hits, or a .1063 HR per Hit pace.

In the LowA league that I provided screen shots for earlier, the League totals would forecast .0747 HR per hit. Based on current year stats (138 games in) I'm at 3577 HR against 17096 Hits, or a .2092 HR per Hit pace.

I know that 1.9xx LTM for HR is what is throwing things off, but why is autocalc messing it up only for the lower minors but yet pretty accurate and working for ML and higher minors is my question.

No problem I was wondering if you didn't just misspeak on bringing the LTM's forward.

I know my MiLB stats were fine using AC in v22 and before. I haven't taken the time to look at them yet in v23 as I normally don't until my MLB season is over. I then look at them all in one setting. However with my MiLB seasons already complete I can take a look now.

League-- Season first --- LT second. HR per Hit from all minors.
RA--- .0676 --- .0648
LA--- .0809 --- .0800
HA--- .0919 --- .0848
AA--- .0948 --- .0903
AAA-- .1023 --- .0967

I'll also drop screens for Low A. Rather that with a full schedule than Rookie with a short season. The stats at Low A are pretty good in my league.

Originally Posted by koohead

Question: Do you MANUALLY hit the adjust league total modifiers button every year even with the auto-calc check box checked? I understand the need to do it on opening day when rosters are set...but thought that was what was being done by the Automatically adjust LTM for accuracy check box. And I wonder if the reason it is off for the lower minors is that on Opening Day of the ML, the lower levels rosters are not set by any stretch of the imagination, while the higher you go the more set the roster is. Kind of assumed each league would run the auto adjust modifier on its respected opening day...but perhaps that's where things are going bonkers.

I do not, nor have I ever had the "automatically adjust LTMs for accuracy" box checked. I have always manually run it as I like knowing exactly when it is done.

My MiLB seasons are scheduled to start 7 days after MLB. This allows more time for rosters to straighten out. Though one could argue injuries in MLB might throw off the MiLB rosters when it calls up players as replacements. In any case these are manually run on their opening days too.

My rookie (coastal leagues) start one week after the draft. Yes, manual autocalc on opening day

Originally Posted by koohead

couple interesting comments from luckyman and Matt A in this thread regarding Minors and LTM...maybe something is different in the engine between 23 and 22 so will give these suggestions a shot for the following season and report back to y'all.

https://forums.ootpdevelopments.com/...just+modifiers

I do recall reading that now that you remind me. At the time I had no reason to use that option as my stats appeared to be coming out right. As they have for my present season.
I will watch my MiLB in my next year to make sure things are still working.

I should throw in that I do use ghost players so maybe that has an effect on stabilizing things? Also, FWIW, ghosts don't mean rosters are bare. Most teams will run at least 80% filled with "real" players and most teams with more.

I did not realize you were mostly(?) talking about minor league ball. I think the others in this thread, and other threads too, are talking about their MLB games. In any case this is so much easier to discuss with data. I hope others that are having issues post some screens so we get a more complete picture.

I'll go back to throwing autocalc under the buss, to me, is not the answer. At least it's not with what I see in my game.

[koohead]

Thanks much for your data and feedback. Couple extra steps but think the manual hitting of that button for each league on its respective opening day is the answer to my problems. Half tempted to check that box for automatically adjusting the game engine but for the sake of not having too many variables change in a test will leave it off and see if manually calculating LTM for the minors does the trick.

Good data point too regarding ghost players. That 10-20% might account for the small percentage points below your expected rates, but definitely encouraged by your results...tells me the engine works, just something I'm doing or not doing is throwing my results off. Thanks for taking the time to discuss!

[Lukas Berger]

For the minors, you'll generally get better (i.e. less extreme) results if if you have the box to 'automatically control in-game engine' checked.

[Sweed]

Originally Posted by Lukas Berger

For the minors, you'll generally get better (i.e. less extreme) results if if you have the box to 'automatically control in-game engine' checked.

Thanks Lukas.

[koohead]

Originally Posted by Lukas Berger

For the minors, you'll generally get better (i.e. less extreme) results if if you have the box to 'automatically control in-game engine' checked.

Thanks. And for the minors, less extreme variance is all I would be going for.

[RchW]

Originally Posted by Lukas Berger

For the minors, you'll generally get better (i.e. less extreme) results if if you have the box to 'automatically control in-game engine' checked.

All minors? Seems to me the original recommendation was for the lowest minor league level in-game.

[dodgerblue88]

With regards to the skewed HR leaders, could the issue be that LTM's are being applied at the individual player level as opposed to being evenly spread? For example, if I have a LTM of 1.5 for home runs:

Player A - expected stats: 40 HR in the editor
40 * 1.5 HR LTM= 60 HR (+20 Hr difference)
Player B - expected stats: 4 HR in the editor
4 * 1.5 = 6 HR (+2 hr difference)

So instead of evenly spreading the additional 50% of home runs across all players, it is only widening the gap because the base hr to calculate upon is higher for the guys with more power.

[MathBandit]

Originally Posted by dodgerblue88

With regards to the skewed HR leaders, could the issue be that LTM's are being applied at the individual player level as opposed to being evenly spread? For example, if I have a LTM of 1.5 for home runs:

Player A - expected stats: 40 HR in the editor
40 * 1.5 HR LTM= 60 HR (+20 Hr difference)
Player B - expected stats: 4 HR in the editor
4 * 1.5 = 6 HR (+2 hr difference)

So instead of evenly spreading the additional 50% of home runs across all players, it is only widening the gap because the base hr to calculate upon is higher for the guys with more power.

My (admittedly limited) understanding of LTM is that's exactly how it is *supposed* to work.

In your case Player A is hitting about 10x more HRs than Player B, so if the HRs were spread out evenly (rather than proportionally), then maybe instead he's only hitting 5x more HRs than Player B and is therefore much less valuable than he was before.

[dodgerblue88]

Yikes.. That's a problem for counting stats though isn't it? Going back to my example... Even in an environment with 50% more HR's the difference in value between a 60hr and 6hr player is much wider than that between a 40hr and 4 HR player. That could be why the gap in WAR among starters gets wider the further you go into a sim.

I wonder what would be the effect of keeping all LTM 's at 1 and only checking the "lock league total stats" button? Anyone know?

Originally Posted by MathBandit

My (admittedly limited) understanding of LTM is that's exactly how it is *supposed* to work.

In your case Player A is hitting about 10x more HRs than Player B, so if the HRs were spread out evenly (rather than proportionally), then maybe instead he's only hitting 5x more HRs than Player B and is therefore much less valuable than he was before.

[dodgerblue88]

So my understanding is... LTM's are pretty good at doing their job so long as its within +/- .20 or so. Once you get outside of those ranges things start to get wonky with stats distribution.

What if instead of adjusting LTM's up to match HR totals, we adjust the League total HR's up and have the LTM's adjust down (<1.0), that should help spread things out right?

I imagine this would require some sort of spreadsheet tool to "back in" to a League Total HR number that keeps LTM's as close to 1.0 as possible, while keeping all the other League Total #'s in line