What is More Important in Historical Leagues: IRL Stats or Ratings?

[Brad K]

LOL. Good to see you fred. A new rhythm to the old tune.

[fredbeene]

Originally Posted by Brad K

Norm Cash, 1961, single season ratings, OOTP 21, 22, 23. Anyone have him hit .400? Cut the homeruns though. That's important.

Most interesting is that there are no new attributes in player editor for 3 years min....i think it is like 10

[uruguru]

Originally Posted by Brad K

Norm Cash, 1961, single season ratings, OOTP 21, 22, 23. Anyone have him hit .400? Cut the homeruns though. That's important.

Well, you definitely piqued my curiosity so I created a 1961 save with settings at what I believe are for the most accurate simulation. I then ran 10 simulations (results in attached image).

As you know, Norm Cash in 1961 hit 41 homers with an astonishing 361/487/662 slash line. In the sims his HR total ranged from 35 to 48, averaging 39.9 per season.

It was actually his batting average that suffered. It ranged from .283 to .367 (averaging .3099)

edit: for completeness, I added an image of his ratings in the sim

Now admittedly this is all a very small sample size but it is random and unbiased so standard deviations for these stats should be calculable.

[fredbeene]

Originally Posted by David Watts

There's Carew and Brett, but again, what about 1961 Norm Cash? Dude hit .243 in 62 and never hit higher than .283 over the next 13 years of his career. In 61 he hit .361 with 41 bombs.

what about campy hitting 17 HRs PEAK after 30.
he will never come close if you import him after age 30

[uruguru]

Originally Posted by fredbeene

Doesn't POTENTIAL handle all that? He has the potential to hit 61 HR
Unless you import him AFTER he turns 30 then POTENTIAL means nothing.

Those are for potential ratings, not resulting stats.

The ratings get "converted", so to speak, into stats when you sim a season. If the ratings are designed to have a player hit 40 homers, then you would expect that over many sims for the same season he would average about 40 homers.

The issue with having Maris average 61 home runs is that the natural variation of the sim engine means that he is going to hit more than 61 home runs half of the time. However, repeated sims show that this is clearly NOT the case so his power ratings for that season are tuned somewhere below 61 home runs -or- there is a governor applied that makes home runs less likely as he approaches 61 (this would be observable by tracking monthly totals).

So while Maris might occasionally hit 61 home runs in a 1961 sim, he is going to average a bit lower. Whether that is the correct approach or not is the point of debate.

[Rain King]

Originally Posted by Brad K

That's what they say but under close questioning they can't tell us what year represents a modern neutral MLB environment. It is quite amazing though how often those resulting stats are very close to historical stats regardless of the year selected.

I'm going to state that resulting stats are reasonably close to a player's real life stats for that year and are valid for this discussion. If you disagree then present your case.

Easy.

Import a modern season. Then change the League Settings to a Deadball season (make sure to lock your league totals).

You now have a situation where your players are rated for a modern season (including the "Resulting Stats" in the editor), but their actual output will be Deadball because that is what is in the League Totals.

Most of the time your Real Stats and League Totals are starting with the same "base" which is why the Editor shows numbers that are close to real life stats. However, those "Resulting Stats" are not reflecting the League Totals at all, the similarity is simply a function of the "Real Stats" and "League Totals" being from the same year/era.

So, when discussing these things those "Resulting Stats" are not truly relevant as they do not actually interact with the League Totals.

[Garlon]

The resulting stats you get in the editor are simply the player stats converted to per 550 AB after the game makes some adjustments with real stats. It is not doing anything else. There is no need to worry about a modern major league environment. It should really say the resulting stats in the environment in which the player actually played during that season.

You need to remember that the results in your league will depend on your league settings and your adjust weaken values. Maris had 590 AB in 1961. What if you set the adjust to 500 AB instead of the default 300 AB? Maris will hit more HR because so many other players will be adjusted, and his HR Power will be even better compared to the rest of the league than if you use the default 300. I do not suggest using 500 AB as your adjust value, I am simply using it to illustrate the situation. This applies to the pitchers too. If you were to reduce the adjust and weaken values of pitchers in the league you will reduce the number of adjusted pitchers, meaning the hitters in the league will in general have a somewhat tougher Ab against a significant amount of pitchers. This means that the rest of the batters in the league will have a tougher time stealing HR away from Maris, Mantle and some other players.

Another option is to use the neutralized stats, which adjust for the ballpark effects but does not use the extra adjustment that gets applied when using real stats.

Maris Neutralized and Cash Neutralized for the 1961 season

[Garlon]

I decided to set up a 1961 season with neutralized stats. Batters 150/50 and Pitchers 19/10 with real lineups and transactions, fielding 3yr, framing 0, disable player personality and morale settings. I should have turned off coaching too, but I missed that.

Maris real: 61 HR in 590 AB with .269/.372/.620/.993
Maris game: 63 HR in 605 AB with .274/.380/.612/.992

Mantle real: 54 HR in 514 AB with .317/.448/.687/1.135
Mantle game: 60 HR in 538 AB with .327/.450/.691/1.142

There are still improvements to be made with replay leagues though in terms of pitching and defense.

[Brad K]

Originally Posted by Garlon

The resulting stats you get in the editor are simply the player stats converted to per 550 AB after the game makes some adjustments with real stats. It is not doing anything else. There is no need to worry about a modern major league environment. It should really say the resulting stats in the environment in which the player actually played during that season.

Thank you. I suspected that was the case. I am grateful that someone in the know has stated it.

[Rain King]

Originally Posted by Brad K

Thank you. I suspected that was the case. I am grateful that someone in the know has stated it.

The fact remains that stat line doesn't "know" the actual league environment. It doesn't know what was used for Adjust/Weaken, etc. It isn't providing a nuanced version of what to expect during the actual simulation.

[chucksabr]

Can anyone help me understand whether an OOTP historical sim that uses real stats instead of neutralized stats really stipulates that 2019 Jose Berrios is a way better strikeout pitcher than 1941 Bob Feller, and that 1930 Don Hurst is exactly as good a hitter as 1968 Carl Yastrzemski? Because if the game’s real stats feature, by definition, doesn’t neutralize player’s stats for era and ballpark when setting up the their ratings for the sim, wouldn’t that be the case? Or is there something I’m not aware of in all this?

[David Watts]

Originally Posted by Garlon

I decided to set up a 1961 season with neutralized stats. Batters 150/50 and Pitchers 19/10 with real lineups and transactions, fielding 3yr, framing 0, disable player personality and morale settings. I should have turned off coaching too, but I missed that.

Maris real: 61 HR in 590 AB with .269/.372/.620/.993
Maris game: 63 HR in 605 AB with .274/.380/.612/.992

Mantle real: 54 HR in 514 AB with .317/.448/.687/1.135
Mantle game: 60 HR in 538 AB with .327/.450/.691/1.142

There are still improvements to be made with replay leagues though in terms of pitching and defense.

Thank you. I really wish the adjust/weaken settings didn't have such a playing the lotto feel to them. I just want to set up a historical game and have OOTP provide the guidance in regards to the settings that will provide the best results. "Garlon" is permanently saved in the forum search function as I'm constantly searching for posts you've made stating what adjust/weaken settings you use. It would be great if you could talk them into adding a feature that actually provides the best settings to use based on the type of recalc we decide to use. Way back when in a post you said you use 154/67 17/14 for 5 year recalc setups. I've been using those setting for random debut's and I really like the results. That being said, what should be used for 3 year leagues?

[Rain King]

Originally Posted by chucksabr

Can anyone help me understand whether an OOTP historical sim that uses real stats instead of neutralized stats really stipulates that 2019 Jose Berrios is a way better strikeout pitcher than 1941 Bob Feller, and that 1930 Don Hurst is exactly as good a hitter as 1968 Carl Yastrzemski? Because if the game’s real stats feature, by definition, doesn’t neutralize player’s stats for era and ballpark when setting up the their ratings for the sim, wouldn’t that be the case? Or is there something I’m not aware of in all this?

We need more context about your situation. Are these players in the same league at the same time? How was that done?

[Brad K]

Originally Posted by Rain King

The fact remains that stat line doesn't "know" the actual league environment. It doesn't know what was used for Adjust/Weaken, etc. It isn't providing a nuanced version of what to expect during the actual simulation.

You selected a great word with nuanced. A subtle effect doesn't matter here.

[uruguru]

Originally Posted by chucksabr

Can anyone help me understand whether an OOTP historical sim that uses real stats instead of neutralized stats really stipulates that 2019 Jose Berrios is a way better strikeout pitcher than 1941 Bob Feller, and that 1930 Don Hurst is exactly as good a hitter as 1968 Carl Yastrzemski? Because if the game’s real stats feature, by definition, doesn’t neutralize player’s stats for era and ballpark when setting up the their ratings for the sim, wouldn’t that be the case? Or is there something I’m not aware of in all this?

I think the easiest way to find that out is to create a 1941 save and a 2019 with the desired settings and then to manually inspect the ratings of Feller & Berrios.

Because you are correct IF 'real stats' don't neutralize for era and ballpark then it logically follows that Feller's K rate in 1941 will pale in comparison to most pitchers in the modern era.

[jcard]

Originally Posted by chucksabr

Can anyone help me understand whether an OOTP historical sim that uses real stats instead of neutralized stats really stipulates that 2019 Jose Berrios is a way better strikeout pitcher than 1941 Bob Feller, and that 1930 Don Hurst is exactly as good a hitter as 1968 Carl Yastrzemski? Because if the gameÂ’s real stats feature, by definition, doesnÂ’t neutralize playerÂ’s stats for era and ballpark when setting up the their ratings for the sim, wouldnÂ’t that be the case? Or is there something IÂ’m not aware of in all this?

I do not wish to confuse the issue with misinformation, so I hope that someone better versed on the matter than I will correct or confirm the ensuing as necessary:

My understanding was that the neutralized database comprises adjusted totals that have been translated into a specific uniform context, such that if you use a neutral ballpark file for your league the relative performances of the players will reflect the numbers in this database. The real stats database, conversely, comprises raw statistics that upon game creation are run through the OOTP historical ballpark and era files to adjust for the context in which the raw data was produced. So, ultimately, either database is in effect neutralized, but the neutral context and methodology of translating to it differ somewhat.

[chucksabr]

Originally Posted by uruguru

I think the easiest way to find that out is to create a 1941 save and a 2019 with the desired settings and then to manually inspect the ratings of Feller & Berrios.

Because you are correct IF 'real stats' don't neutralize for era and ballpark then it logically follows that Feller's K rate in 1941 will pale in comparison to most pitchers in the modern era.

Right, which is why I am asking whether someone who has had experience in this regard can tell me whether that real stats hypothesis is correct. Because if it's truly about the real stats, then yes, 2019 Berrios is way better than 1941 Feller, and if I need strikeouts, I am taking Berrios, and probably hundreds of other modern pitchers, over Feller.

In terms of context, I am starting participation in an historical league that will make available every major leaguer from 1871 through 2022, and it is practically certain that at some point Feller and modern pitchers like Berrios will show up at the same time. Since the ratings will be based on real stats, my question is whether to treat Feller as a K-challenged pitcher when taken against modern options, specifically because of the raw stats, or whether there is an additional, perhaps undocumented, moderating influence beyond the documented neutralization filter (which is not being implemented in our sim) that would elevate Feller's strikeout ability beyond what raw real stats would suggest.

[chucksabr]

Originally Posted by jcard

I do not wish to confuse the issue with misinformation, so I hope that someone better versed on the matter than I will correct or confirm the ensuing as necessary:

My understanding was that the neutralized database comprises adjusted totals that have been translated into a specific uniform context, such that if you use a neutral ballpark file for your league the relative performances of the players will reflect the numbers in this database. The real stats database, conversely, comprises raw statistics that upon game creation are run through the OOTP historical ballpark and era files to adjust for the context in which the raw data was produced. So, ultimately, either database is in effect neutralized, but the neutral context and methodology of translating to it differ somewhat.

If this is true, the $64 question would be how real stats are so neutralized. Would this mean that + stats such as wRC+, FIP-, K9+, etc., still be valid informers of OOTP ratings in a real stats-based game? Meaning, because of K9+, is 1941 Feller with the 6.8 K9 still a much better strikeout pitcher than 2019 Berrios with the 8.5 K9?

And the $65 question, of course, is whether this is true in the first place, because this would suggest that real stats are also neutralized, just not as neutralized as neutralized.

[jcard]

Originally Posted by uruguru

Those are for potential ratings, not resulting stats.

The ratings get "converted", so to speak, into stats when you sim a season. If the ratings are designed to have a player hit 40 homers, then you would expect that over many sims for the same season he would average about 40 homers.

The issue with having Maris average 61 home runs is that the natural variation of the sim engine means that he is going to hit more than 61 home runs half of the time. However, repeated sims show that this is clearly NOT the case so his power ratings for that season are tuned somewhere below 61 home runs -or- there is a governor applied that makes home runs less likely as he approaches 61 (this would be observable by tracking monthly totals).

So while Maris might occasionally hit 61 home runs in a 1961 sim, he is going to average a bit lower. Whether that is the correct approach or not is the point of debate.

I just want to say that everything you have written in this thread is spot-on, particularly in your framing of the issue as whether the goal is to represent the player (thus, the expression of his given underlying talent absent stochastic external variables) or his statistics (thus, the expression of that given talent and a single season sample of those external variables).

[jcard]

Originally Posted by chucksabr

If this is true, the $64 question would be how real stats are so neutralized. Would this mean that + stats such as wRC+, FIP-, K9+, etc., still be valid informers of OOTP ratings in a real stats-based game? Meaning, because of K9+, is 1941 Feller with the 6.8 K9 still a much better strikeout pitcher than 2019 Berrios with the 8.5 K9?

And the $65 question, of course, is whether this is true in the first place, because this would suggest that real stats are also neutralized, just not as neutralized as neutralized.

If you have not already, I strongly suggest reading the linked thread from 2021; the early posts are less relevant, but soon the discussion focuses on exactly the types of questions you are asking:

https://forums.ootpdevelopments.com/...d.php?t=331958