Park Factors

diehardsim · 12-11-2023, 10:34 PM

Hello,

I've simulated over 20 season as the San Francisco Giants and I've come to determine through experience that the park factors are incorrect. Oracle Park turns All Star hitters that hit .300 in every other park into barely above replacement level players hitting .240. When these players are traded to another team they regain their statistics. I've also won multiple Cy Young awards with players like Ross Stripling and Jacob Junis, only to see their performance dip into replacement level on other teams.

When I look up park factors for Oracle Park the analytics indicate it plays very similar to Oakland and Detroit in 2023 on a three year average but the factors are totally different in OOTP24.

https://baseballsavant.mlb.com/leade...t-park-factors

The Giants moved their fences in, changed the height of the fences, and changed the airflow in the archways in 2020. If I had to guess the developers have become lazy with park factors and haven't updated them since before 2020. Distances and fence height are all pre-2020 dimensions.

https://www.mlb.com/news/giants-to-m..._article-share

If this is the case I'd like my money back.

Garlon · 12-12-2023, 02:26 AM

The file with the ballpark factors for OOTP is custom to work properly with the game engine.

The fence distances and heights are cosmetic in the file. The factors are what the game uses.

I do the updates on that file, and I even keep two versions of the ballpark factors in it.

There are what I refer to as ballpark scoring quadratic factors (the game does not load these from the file, but they are in there) which is calculated by taking the R/G at that ballpark and comparing it to the R/G of the road games. This is the total scoring for the home team and road team combined in each situation. This is taken for 3-year spans and then using a Runs Created formula based on the focus season we generate a single value that gets applied to Hits and HR to reproduce the scoring environment of that ballpark. Technically the value is also getting applied to 2B and 3B too as a result of the value affecting overall hits, but in the file 2B and 3B are set to 1.000 for this set of ballpark factors. Again, the game does not load these factors anyway when you play.

The second type of factors that the game does import from the file I refer to as discrete factors. This takes the rate of BABIP, 2B per Hit on Balls in Play, 3B per Hit on Balls in Play, and HR per AB in which a strikeout does not occur and then provides this for RHB and LHB from 1905-2019. Since 2020 the same process is done but we only have the overall discrete values and RHB and LHB are set to the same values. If the required RHB and LHB statistics become available, I can update those since the 2020 season.

The entire file has factors for every ballpark since 1871.

The ballpark factors you see published elsewhere are not appropriate for use with OOTP because they are making a very basic calculation using totals, whereas with the OOTP game engine we need it to be more specific. Putting ballpark factors in from baseball savant or anywhere else would actually be the lazy way to do things.

The 1yr ballpark scoring factors for Oracle Park in recent seasons are:

2017: 0.850
2018: 1.012
2019: 0.798
2020: 1.072
2021: 0.975
2022: 0.968
2023: 0.838

Here are the 2019 factors in the file for Oracle Park for example:

RHB
BABIP: 1.002
2B: 0.917
3B: 1.047
HR: 0.876

LHB
BABIP: .995
2B: 1.021
3B: 1.487
HR: 0.826

As you can see those factors for BABIP will not affect BA that much as they are basically adding 2 more hits for LHB per 1000 hits in an average ballpark and subtracting 5 hits for RHB per 1000 hits in an average ballpark.

If a RHB is expected to hit 30 HR in an average environment but plays half the games at Oracle Park, then his Home HR Factor is 0.826 and his road will be (15-0.826)/14 = 1.0124, and combined we get 0.919. If we take 30*0.919 we get 27.57. The player who is expected to hit 30 HR will end up with about 27 by playing half of their games at Oracle Park in 2019.

However, the neutralized stats account for this and will rate that 30 HR player as 30/0.919 = 32.64 = 33 HR in the neutralized stats for the game so that the neutralized stats work in tandem with the ballpark factors in the file.

Garlon · 12-12-2023, 02:33 AM

If you are getting anomalous results in your game, it is most likely due to the settings you are using. Do you have recalc enabled or are you using player development? Are you letting the game recalc the League Totals Modifiers each season? It could be a variety of other things. Typically, we should expect scoring to be down close to 10% at Oracle Park, so a pitcher who pitches half their games at home at Oracle Park will get an overall reduction of their ERA of roughly 5%. So, a pitcher who is expected to have a 3.75 ERA may and up as 3.56 for the season.

If you are winning Cy Young awards, I assume your pitchers are also pitching on the road too. So, they must be performing great in the other half of their IP as well.

Reed · 12-12-2023, 07:20 AM

Garlon,
I don’t have anything to add about park factors since I play with neutralized park factors.
I just wanted to comment on neutralized stats. Tried playing with neutralized stats years ago but wasn’t happy with some of the things in it. I recently started a new league and decided to try neutralized stats again. It is playing exactly as I hoped it would and just wanted to thank you for your work on it.

diehardsim · 12-12-2023, 11:21 AM

Quote:

Originally Posted by Garlon

If you are getting anomalous results in your game, it is most likely due to the settings you are using. Do you have recalc enabled or are you using player development? Are you letting the game recalc the League Totals Modifiers each season? It could be a variety of other things. Typically, we should expect scoring to be down close to 10% at Oracle Park, so a pitcher who pitches half their games at home at Oracle Park will get an overall reduction of their ERA of roughly 5%. So, a pitcher who is expected to have a 3.75 ERA may and up as 3.56 for the season.

If you are winning Cy Young awards, I assume your pitchers are also pitching on the road too. So, they must be performing great in the other half of their IP as well.

I haven't changed any of the game's settings in these results and I'm using player development.

Pitching statistics are dominantly superior for my starters at Oracle Park. In the case of Stripling, he had a 1.34 ERA at Oracle which carried his statistics compared to a 3.45 ERA on the road. I've found that any pitcher that can throw strikes dominates at Oracle and that simply isn't the case in reality.

diehardsim · 12-12-2023, 11:30 AM

Quote:

Originally Posted by Garlon

The file with the ballpark factors for OOTP is custom to work properly with the game engine.

The fence distances and heights are cosmetic in the file. The factors are what the game uses.

I do the updates on that file, and I even keep two versions of the ballpark factors in it.

There are what I refer to as ballpark scoring quadratic factors (the game does not load these from the file, but they are in there) which is calculated by taking the R/G at that ballpark and comparing it to the R/G of the road games. This is the total scoring for the home team and road team combined in each situation. This is taken for 3-year spans and then using a Runs Created formula based on the focus season we generate a single value that gets applied to Hits and HR to reproduce the scoring environment of that ballpark. Technically the value is also getting applied to 2B and 3B too as a result of the value affecting overall hits, but in the file 2B and 3B are set to 1.000 for this set of ballpark factors. Again, the game does not load these factors anyway when you play.

The second type of factors that the game does import from the file I refer to as discrete factors. This takes the rate of BABIP, 2B per Hit on Balls in Play, 3B per Hit on Balls in Play, and HR per AB in which a strikeout does not occur and then provides this for RHB and LHB from 1905-2019. Since 2020 the same process is done but we only have the overall discrete values and RHB and LHB are set to the same values. If the required RHB and LHB statistics become available, I can update those since the 2020 season.

The entire file has factors for every ballpark since 1871.

The ballpark factors you see published elsewhere are not appropriate for use with OOTP because they are making a very basic calculation using totals, whereas with the OOTP game engine we need it to be more specific. Putting ballpark factors in from baseball savant or anywhere else would actually be the lazy way to do things.

The 1yr ballpark scoring factors for Oracle Park in recent seasons are:

2017: 0.850
2018: 1.012
2019: 0.798
2020: 1.072
2021: 0.975
2022: 0.968
2023: 0.838

Here are the 2019 factors in the file for Oracle Park for example:

RHB
BABIP: 1.002
2B: 0.917
3B: 1.047
HR: 0.876

LHB
BABIP: .995
2B: 1.021
3B: 1.487
HR: 0.826

As you can see those factors for BABIP will not affect BA that much as they are basically adding 2 more hits for LHB per 1000 hits in an average ballpark and subtracting 5 hits for RHB per 1000 hits in an average ballpark.

If a RHB is expected to hit 30 HR in an average environment but plays half the games at Oracle Park, then his Home HR Factor is 0.826 and his road will be (15-0.826)/14 = 1.0124, and combined we get 0.919. If we take 30*0.919 we get 27.57. The player who is expected to hit 30 HR will end up with about 27 by playing half of their games at Oracle Park in 2019.

However, the neutralized stats account for this and will rate that 30 HR player as 30/0.919 = 32.64 = 33 HR in the neutralized stats for the game so that the neutralized stats work in tandem with the ballpark factors in the file.

I've heard this before but then why are Comerica Park and the Oakland Coliseum's park factors so different from Oracle Park if they have similar factors at Baseball Savant?

Park factors change from year to year and the game running on data from 2020 is wrong. It seems to me that if OOTP has its own recipe for determining park factors the system isn't matching reality and it is time to fix it.

Garlon · 12-12-2023, 12:13 PM

The ballpark factors file for OOTP has different park factors each season for every ballpark since 1871. They are 3yr factors calculated from what actually happens in that ballpark compared to the road games.

Here are 1yr 2020 scoring factors from retrosheet which is R/G at the home ballpark compared to the road ballparks.

Comerica 0.919
Oakland 0.802
Oracle 1.072

Those ballparks did not play the same that season.

There are many ways to do ballpark factors, 1yr, 3yr, counting totals comparisons, rate comparisons, scoring comparisons, and then the specific ballpark factors made for OOTP which are rate comparisons but done to work with the game engine.

Garlon · 12-12-2023, 12:15 PM

If you do not have the game automatically load the ballpark factors each season then they may not be updating for you as you progress to the next season. Again, this is how you may be setting up your league.

Syd Thrift · 12-12-2023, 12:58 PM

Wait, is the first field "BABIP" now and not just hits? Because it still says hits in the game and that's what I've been manually calculating based on that (which, as an aside, if BABIP is being used in park factors, we should also have K rates - particularly in bygone eras, some batting eyes are better than others - but I'm sure that the vast majority of the time HRs are a much bigger factor).

Garlon · 12-12-2023, 01:55 PM

Yes, that hits factor is actually BABIP as that is how it is applied in the game.

If the game supported it we could have a K factor. However, K rate is probably more influenced by foul territory rather than the batter eye. If you have a lot of foul territory you get more foul outs, which reduces strikeouts. Whereas with less foul territory the ball goes into the stands and there is an additional opportunity to strikeout.

Brad K · 12-12-2023, 05:49 PM

Garlon, thanks for the information.

I understand that it is not possible to neutralize minor league stats. But since when the game takes a career minor leaguer and lets him be good enough to make the majors it's all guess work anyway. So is there some possibility the non neutralized minor league stats could be included in a save with neutralized major league stats?

There's no real life performance target for such players. The target is to let only a few do well in the majors and that those who do have a performance that is remotely plausible.

Garlon · 12-12-2023, 09:06 PM

Does the game currently not let you use neutralized stats while enabling real minor leagues?

I never enable minor leagues and have not tried that.

Brad K · 12-13-2023, 12:06 AM

Quote:

Originally Posted by Garlon

Does the game currently not let you use neutralized stats while enabling real minor leagues?

I never enable minor leagues and have not tried that.

As far as I know that's the case. It would be nice if I were wrong.

LansdowneSt · 12-13-2023, 12:12 AM

As soon as you select the checkbox for enabling the Real Minors, the type of stats locks to "Real Stats" and you can't change it. So, one cannot have Neutralized MLB'ers and real minors in the same save.

mytreds · 12-13-2023, 12:23 AM

I remember getting ootp21, and the game had the RF wall of Oracle at 8 ft instead of 24ft. Not surprised they are still not getting it right.

Garlon · 12-13-2023, 12:50 AM

For all seasons of Oracle Park the fences are listed as heights of 8, 8, 11, 8, 25, 25, 25.

Brad K · 12-13-2023, 07:46 PM

Quote:

Originally Posted by LansdowneSt

As soon as you select the checkbox for enabling the Real Minors, the type of stats locks to "Real Stats" and you can't change it. So, one cannot have Neutralized MLB'ers and real minors in the same save.

Yet if a career minor leaguer makes MLB they are functionally a fictional player, a what if, so having their minor league stats neutralized is irrelevant. So unless there are technical issues preventing it, there is no reason not to have neutralized MLB stats and real minor stats.

Garlon · 12-13-2023, 10:46 PM

The file that contains all of the ballpark factors and wall distances and heights is correct. I have an old save of a 2016 game and the wall heights for Oracle Park did import correctly. I tried creating a 2022 game and wall heights are incorrect. There is possibly something wrong with the .prk file and how the game is referencing the era_ballparks file for some seasons. It is possible a bug got into the process at some later update from when I had had made that 2016 save.

12-11-2023, 10:34 PM	#1
diehardsim Minors (Single A) Join Date: Nov 2002 Location: California Posts: 57	Park Factors Hello, I've simulated over 20 season as the San Francisco Giants and I've come to determine through experience that the park factors are incorrect. Oracle Park turns All Star hitters that hit .300 in every other park into barely above replacement level players hitting .240. When these players are traded to another team they regain their statistics. I've also won multiple Cy Young awards with players like Ross Stripling and Jacob Junis, only to see their performance dip into replacement level on other teams. When I look up park factors for Oracle Park the analytics indicate it plays very similar to Oakland and Detroit in 2023 on a three year average but the factors are totally different in OOTP24. https://baseballsavant.mlb.com/leade...t-park-factors The Giants moved their fences in, changed the height of the fences, and changed the airflow in the archways in 2020. If I had to guess the developers have become lazy with park factors and haven't updated them since before 2020. Distances and fence height are all pre-2020 dimensions. https://www.mlb.com/news/giants-to-m..._article-share If this is the case I'd like my money back.

12-12-2023, 02:26 AM	#2
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	The file with the ballpark factors for OOTP is custom to work properly with the game engine. The fence distances and heights are cosmetic in the file. The factors are what the game uses. I do the updates on that file, and I even keep two versions of the ballpark factors in it. There are what I refer to as ballpark scoring quadratic factors (the game does not load these from the file, but they are in there) which is calculated by taking the R/G at that ballpark and comparing it to the R/G of the road games. This is the total scoring for the home team and road team combined in each situation. This is taken for 3-year spans and then using a Runs Created formula based on the focus season we generate a single value that gets applied to Hits and HR to reproduce the scoring environment of that ballpark. Technically the value is also getting applied to 2B and 3B too as a result of the value affecting overall hits, but in the file 2B and 3B are set to 1.000 for this set of ballpark factors. Again, the game does not load these factors anyway when you play. The second type of factors that the game does import from the file I refer to as discrete factors. This takes the rate of BABIP, 2B per Hit on Balls in Play, 3B per Hit on Balls in Play, and HR per AB in which a strikeout does not occur and then provides this for RHB and LHB from 1905-2019. Since 2020 the same process is done but we only have the overall discrete values and RHB and LHB are set to the same values. If the required RHB and LHB statistics become available, I can update those since the 2020 season. The entire file has factors for every ballpark since 1871. The ballpark factors you see published elsewhere are not appropriate for use with OOTP because they are making a very basic calculation using totals, whereas with the OOTP game engine we need it to be more specific. Putting ballpark factors in from baseball savant or anywhere else would actually be the lazy way to do things. The 1yr ballpark scoring factors for Oracle Park in recent seasons are: 2017: 0.850 2018: 1.012 2019: 0.798 2020: 1.072 2021: 0.975 2022: 0.968 2023: 0.838 Here are the 2019 factors in the file for Oracle Park for example: RHB BABIP: 1.002 2B: 0.917 3B: 1.047 HR: 0.876 LHB BABIP: .995 2B: 1.021 3B: 1.487 HR: 0.826 As you can see those factors for BABIP will not affect BA that much as they are basically adding 2 more hits for LHB per 1000 hits in an average ballpark and subtracting 5 hits for RHB per 1000 hits in an average ballpark. If a RHB is expected to hit 30 HR in an average environment but plays half the games at Oracle Park, then his Home HR Factor is 0.826 and his road will be (15-0.826)/14 = 1.0124, and combined we get 0.919. If we take 300.919 we get 27.57. The player who is expected to hit 30 HR will end up with about 27 by playing half of their games at Oracle Park in 2019. However, the neutralized stats account for this and will rate that 30 HR player as 30/0.919 = 32.64 = 33 HR in the neutralized stats for the game so that the neutralized stats work in tandem with the ballpark factors in the file. Last edited by Garlon; 12-12-2023 at 02:28 AM.*

12-12-2023, 02:33 AM	#3
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	If you are getting anomalous results in your game, it is most likely due to the settings you are using. Do you have recalc enabled or are you using player development? Are you letting the game recalc the League Totals Modifiers each season? It could be a variety of other things. Typically, we should expect scoring to be down close to 10% at Oracle Park, so a pitcher who pitches half their games at home at Oracle Park will get an overall reduction of their ERA of roughly 5%. So, a pitcher who is expected to have a 3.75 ERA may and up as 3.56 for the season. If you are winning Cy Young awards, I assume your pitchers are also pitching on the road too. So, they must be performing great in the other half of their IP as well. Last edited by Garlon; 12-12-2023 at 03:20 AM.

12-12-2023, 12:13 PM	#7
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	The ballpark factors file for OOTP has different park factors each season for every ballpark since 1871. They are 3yr factors calculated from what actually happens in that ballpark compared to the road games. Here are 1yr 2020 scoring factors from retrosheet which is R/G at the home ballpark compared to the road ballparks. Comerica 0.919 Oakland 0.802 Oracle 1.072 Those ballparks did not play the same that season. There are many ways to do ballpark factors, 1yr, 3yr, counting totals comparisons, rate comparisons, scoring comparisons, and then the specific ballpark factors made for OOTP which are rate comparisons but done to work with the game engine. Last edited by Garlon; 12-12-2023 at 12:16 PM.

12-13-2023, 12:12 AM	#14
LansdowneSt Global Moderator Join Date: Feb 2004 Location: From Duxbury, Mass residing Baltimore Posts: 8,027	As soon as you select the checkbox for enabling the Real Minors, the type of stats locks to "Real Stats" and you can't change it. So, one cannot have Neutralized MLB'ers and real minors in the same save. __________________ Complete Universe Facegen Pack 2.0 (mine included) https://www.mediafire.com/file_premi...k_2.0.zip/file Just my Facegen Pack: https://www.mediafire.com/file_premi..._Pack.zip/file

12-12-2023, 07:20 AM	#4
Reed Hall Of Famer Join Date: Oct 2013 Posts: 2,610	Garlon, I don’t have anything to add about park factors since I play with neutralized park factors. I just wanted to comment on neutralized stats. Tried playing with neutralized stats years ago but wasn’t happy with some of the things in it. I recently started a new league and decided to try neutralized stats again. It is playing exactly as I hoped it would and just wanted to thank you for your work on it.

12-12-2023, 12:15 PM	#8
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	If you do not have the game automatically load the ballpark factors each season then they may not be updating for you as you progress to the next season. Again, this is how you may be setting up your league.

12-12-2023, 01:55 PM	#10
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	Yes, that hits factor is actually BABIP as that is how it is applied in the game. If the game supported it we could have a K factor. However, K rate is probably more influenced by foul territory rather than the batter eye. If you have a lot of foul territory you get more foul outs, which reduces strikeouts. Whereas with less foul territory the ball goes into the stands and there is an additional opportunity to strikeout.

12-12-2023, 05:49 PM	#11
Brad K Banned Join Date: May 2016 Location: St Petersburg Florida USA Posts: 6,693 Infractions: 0/2 (4)	Garlon, thanks for the information. I understand that it is not possible to neutralize minor league stats. But since when the game takes a career minor leaguer and lets him be good enough to make the majors it's all guess work anyway. So is there some possibility the non neutralized minor league stats could be included in a save with neutralized major league stats? There's no real life performance target for such players. The target is to let only a few do well in the majors and that those who do have a performance that is remotely plausible.

12-12-2023, 09:06 PM	#12
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	Does the game currently not let you use neutralized stats while enabling real minor leagues? I never enable minor leagues and have not tried that.

12-13-2023, 12:23 AM	#15
mytreds All Star Starter Join Date: Nov 2019 Posts: 1,183	I remember getting ootp21, and the game had the RF wall of Oracle at 8 ft instead of 24ft. Not surprised they are still not getting it right. __________________ “Baseball isn’t statistics; it’s Joe DiMaggio rounding second.” “Once, centuries ago, it was the beloved national pastime of the Americas, Wesley. Abandoned by a society that prized fast food and faster games. Lost to impatience.” “ The term ‘WAR’ should be replaced by ‘WAG’. WAR isn’t an actual measurement; it’s just a wild-ass guess” -Bill James RIP National League 1876-2022 Floreat semper vel invita morte. I make custom ballparks.

12-13-2023, 12:50 AM	#16
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	For all seasons of Oracle Park the fences are listed as heights of 8, 8, 11, 8, 25, 25, 25.

12-13-2023, 10:46 PM	#18
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,347	The file that contains all of the ballpark factors and wall distances and heights is correct. I have an old save of a 2016 game and the wall heights for Oracle Park did import correctly. I tried creating a 2022 game and wall heights are incorrect. There is possibly something wrong with the .prk file and how the game is referencing the era_ballparks file for some seasons. It is possible a bug got into the process at some later update from when I had had made that 2016 save.