Park Factors and Historical Leagues

[Garlon]

DawnBTVS, tell us if you like the results of this file. If so i may be able to add more data and make it available on the Workshop.

[DawnBTVS]

Originally Posted by Garlon

DawnBTVS, tell us if you like the results of this file. If so i may be able to add more data and make it available on the Workshop.

Really enjoying it to be honest. I have an 1871 Historical League up through 1886 that's quasi-based off current MLB (162 Game Sched, 5 Man Rotation, Heavy Bullpen/Closers) and it still manages to feel pretty realistic. More so to me than the normal PF file was doing for me.

Also did a 1961 Historical literally just going year to year without touching anything except Voting HOF. Currently through 1968 but it's great seeing Harmon Killebrew potentially flirting with 500 HR (went from 49/50/49 HR to low 30s as the pitcher era's taking over by 1965). Hank Aaron's been consistent as hell just like real life: 33, 35, 26, 38, 30, 43, 33, and 23 HR thus far.

Sandy Koufax is being Koufax currently sitting at 175-91 with a 2.37 career ERA and has exploded from age 26-32. Bob Gibson's flashed some greatness (20-8 and 21-7 in 1963/1964, 21-9 in 1968) but he's also had some struggling seasons and just got traded to the California Angels in 1968.

I like the feel overall of these. They feel realistic to radically different eras without making me think certain players are "cheated" or playing extremely differently than real life. Maris had 51 HR in 1961 but has consistently been in the upper 20s since then while Mickey Mantle's proven to be a better 30-40 HR range hitter.

[italyprof]

Originally Posted by DawnBTVS

Thanks for the correction, Garlon.

With that said, Roger Hornsby's 2B and 3B were 8.4% of his PA at Sportsman's Park (keeping in mind that he hit there the majority of his career outside of 5 seasons at the end of his full time career). Compare that to other ballparks he hit in.

Sportsman's Park: 8.4%
Robison Field: 6.7%
Cubs Park: 6.9%
Polo Grounds: 6.1%
Braves Field: 7.7%
Baker Bowl: 8.0%
Ebbets Field: 7.4%

That doesn't look like a big difference but it's also pretty clear that Hornsby, as a right handed hitter, took advantage of Sportsman's Park and the propensity for extra bases. That is largely what seems to be missing from OOTP's Historical Leagues especially if every ballpark treats 2B/3B as a league average 1.000 number.

So giving him 308 AB for Sportsman's Park = roughly 26 2B and 3B. Compare that to just 21 at a Robison Field or 23 at Ebbets Field doesn't sound like a big difference but it adds up in BA, OBP, SLG, and career wise.

Over a 12 year run that equates to 312 at Sportsman's Park against 252 (Robison Field) or 276 (Ebbets Field).

Babe Ruth at Yankee Stadium is another extreme example of how ballparks did effect the importance of RH vs. LH but that aspect isn't quite carried over into OOTP which does hurt historical league play through IMO.

He hit 7.1% of his HR per PA at Yankee Stadium compared to a miserable 3.9% at Fenway Park. Even at the Polo Grounds and Navin Field, he hit well over 7.7% of his HR per PA. Keep Ruth in Fenway Park and he doesn't come close to his total career HR numbers because he hit LH in a park that was and still is incredibly tough on power hitting LH.

Sorry, but Babe Ruth hit MORE home runs on the road than at home in his career:

http://research.sabr.org/journals/hi...-a-on-the-road

And when he hit 54 to make the new record, the Yankees played in the Polo Grounds. And only 259 of his 714 were hit at Yankee Stadium. 455 were hit in other ballparks:

http://www.baberuthcentral.com/babe-...tics/#Ballpark

[DawnBTVS]

Originally Posted by italyprof

Sorry, but Babe Ruth hit MORE home runs on the road than at home in his career:

http://research.sabr.org/journals/hi...-a-on-the-road

And when he hit 54 to make the new record, the Yankees played in the Polo Grounds. And only 259 of his 714 were hit at Yankee Stadium. 455 were hit in other ballparks:

http://www.baberuthcentral.com/babe-...tics/#Ballpark

You're right. I was going by Baseball-Reference's career splits and the actual PA within that ballpark not the overall career effect.

So Ruth had 3,633 PA at Yankee Stadium for his career and hit 259 HR there which comes out to a 7.1% rate. That's how I got the rates for the other ballparks e.g. Fenway Park was 1,266 PA and 49 HR, etc. Fun Fact: Ruth hit a HR/PA of 8.8% at the Polo Grounds

https://www.baseball-reference.com/p...ear=Career&t=b

[Garlon]

Here is a new file with discrete park factors 1906-2018. 1871-1905 is still the quadratic factors. The BA factor is really a BABIP factor. The 2B and 3B factors are relative to total hits excluding HR. The HR factor is based on HR per AB not ending with a strikeout.

Everything has been completely recalculated so there are differences with the BA and HR factors compared to the previous file I posted.

This file is compatible with the Major League Stadiums mod on Steam Workshop. To use this file you need to rename it era_ballparks and put it in the My Documents>database folder. Be sure to save your original era_ballparks file by making a copy or renaming it before installing this one.

[DawnBTVS]

Originally Posted by Garlon

Here is a new file with discrete park factors 1906-2018. 1871-1905 is still the quadratic factors. The BA factor is really a BABIP factor. The 2B and 3B factors are relative to total hits excluding HR. The HR factor is based on HR per AB not ending with a strikeout.

Everything has been completely recalculated so there are differences with the BA and HR factors compared to the previous file I posted.

This file is compatible with the Major League Stadiums mod on Steam Workshop. To use this file you need to rename it era_ballparks and put it in the My Documents>database folder. Be sure to save your original era_ballparks file by making a copy or renaming it before installing this one.

Looking forward to this when I get off work this morning. Really appreciate you going out of your way/above & beyond for this!

[Syd Thrift]

Originally Posted by DawnBTVS

Yeah, that's what I'm saying.

OOTP Example: 1949 Red Sox has Fenway Park at 1.128 for BA and HR (league started in 1948). Yankee Stadium for comparison is set at 0.930 for BA and HR.

But everybody knows the disparity between the two ballparks in real life e.g. trading DiMaggio for Williams and why the "opposite" park would've aided them each more. I'd even be grateful if the OOTP team could just do a "general" park factor adjusting for the fence distances/wall heights rather than needing to be 100% legitimate if it meant closely mimicking the real life park effects. Otherwise it's pointless to have "park factors" if it's just deciding which park is more advantageous to runs ignoring XBH or HR or LH/RH hitting.

Example: Fenway Park RH due to the Green Monster becomes 1.158 for Doubles and 1.070 for RH HR but due to the distance/configuration it makes LH HR just a 0.959 or whatever without trying to factor in the actual player stats and just judging based on distance/wall height.

What I do is I just grab the 3-year factors from seamheads.com. That being said, I like to play fictional/historical for a variety of reasons and one of them is that, well, the player database is not optimized for park factors. Yes, IRL Fenway in the 40s and the 50s was a great place for right-handed hitters and not so much for left-handed ones. However, Ted Williams gets a small *downgrade* to his stats if you bring him in from this era because the database *does* adjust for the overall park factor, and as a result he will probably underperform in your league. Likewise with Joe DiMaggio; although Yankee Stadium was on par a pitcher's park, Joltin' Joe as a righty had to deal with the cavernous Yankee Stadium left field and as a result IRL he hit 213 of his 361 career dingers on the road.

One thing that would be really, really cool is if someone went in and created a "season disk" the way, say, Strat-o-Matic does, with things being a bit more highly curated, the ratings algorithms tweaked to handle realistic park factors, and so on.

[DawnBTVS]

Been utilizing the discrete parks with a 1968 League (and included the DH in 1969 for both AL/NL after Bob Gibson went 23-8 with a 1.03 ERA in 298 IP to "promote offense" ). Also enacted Free Agency in 1972 as players started to demand more fair salary exposure :P Currently through the 1983 MLB Season.

Random Notables
- Denny McLain went 17-8 with a 1.96 ERA in 1968. Then landed with the expansion Montreal Expos via Expansion Draft where he fell off a cliff 33-70 with a 4.81 ERA and 76 ERA+, w00f.
- Bobby Bonds landed with San Francisco and finished with 2,044 Hits and 242 HR along with an OPS+ of 119.
- Hank Aaron finished his career in 1971 with 575 HR while Mickey Mantle retired after 1971 with 594 HR. Willie Mays ranks 2nd All-Time in HR with a very distant 615 to the Babe.
- Roberto Clemente tied Cap Anson for 4th in Hits with 3,418. Pete Rose struggled with injuries, bounced around (Atlanta, St. Louis, New York Mets, Milwaukee, Philadelphia, Chicago White Sox, and Texas Rangers) before retiring with just 2,634 Hits.
- Juan Marichal closed out his career 1 win shy of 300 thanks to a 7-18 record in 1977 between Toronto and Kansas City. Mickey Lolich has the next closest win total at a low 249 for comparison and he's now 43 years old but still pitching in the minors!
- Nolan Ryan retired after spending his career with the New York Mets from 1966-1983. Went 220-156 with 3,445 K in 3,733 IP including a career best season in 1971: 16-6, 1.82 ERA, 266.1 IP with 281 K against 101 BB.
- New York Yankees won another title in 1975.
- The Chicago Cubs broke their curse in 1977(!) led by John Hale (26 HR, 111 RBI), Oscar Gamble (28 HR, 98 RBI), and Claudell Washington (26 HR, 130 RBI) on the offensive side while Burt Hooton went 20-7 with a 3.16 ERA and Jim M. Hughes went 18-12 with a 4.17 ERA. They knocked off the Cincinnati Reds 4-3 then stomped the California Angels 4-1.
- Boston Red Sox remain cursed despite making the playoffs from 1980-1982.

[Garlon]

I posted an updated version of the mod on the Workshop. The discrete factors are now the default when the mod gets installed. The original quadratic factors are still available in a file called era_ballparks_quadratic, and there is another file with the ballparks set to 1.000 for every factor called era_ballparks_1000 for those who want to use them. These files will be installed in the database folder and simply need to be renamed era_ballparks, replacing the one currently in the folder. Remember, the era_ballparks file is now the discrete factors, so just reinstall the mod after downloading the new update and they will be the new factors that are used in the game after each season.

There was also an issue with Angel Stadium of Anaheim not loading that I believe is now corrected.

[fredbeene]

Originally Posted by Garlon

The park factors were calculated from retrosheet data for every park every season from 1871 to present. I created the original file and the calculator to generate the factors. This was a very big project. Pstrickert has been using the calculator to update the file these past few seasons (thank you Pstrickert). The factors are based on 3 seasons of data. If a park generated say 14% more offense than the other parks in the league, then using the quadratic formula we are able to find a value for both BA and HR that generates that result, and this depends on how teams were scoring in a given season based on HR/2B/3B from the league stat totals for every season. The BA factor affects 2B and 3B as well so that's why those stay at 1.000 in the file. Retrosheet does not have the L/R data for ballparks so that's why those are the same for both sets of values.

Keep in mind that what you are arguing probably doesn't affect any individual player's HR totals for Home vs Road by more than 1 or 2 in a season. For example, you mention 12% fewer HR. If your player in question has 40 HRs for the season, and roughly 20 at home, that 12% becomes about 2-3 HR that get redistributed to road vs home. But the file itself is fairly subtle in terms of the factors because if you increase BA/2B/3B/HR by say 5% your scoring will probably increase more than 10%.

I am not understanding what you are saying about the 2018 factors for Fenway Park. Both BA and HR should be 1.03056 according to the file. Over the 2017-2018 seasons there was 5.24% more offense at Fenway than on the road and the 1.03056 factor is designed to recreate that in the game as best as possible.

All parkpark is static. forget the players. its all about the design.
a 100 foot line will generate more more homers than 1000 foot line. (over the fence homers)
the modifers should be based solely on the ballpark..
WIDE fould lines gives more chanes for outs.

Everything else is random.
Only the build of park should effect.

Great defenese should have huge advantage in large park.
But it seems all we do is guess what the engine does, sadly

[fredbeene]

Originally Posted by RchW

Park factors are post facto based on actual results. Not sure what you think it should be other wise. In other words a park that allows 15% more HR for RHB would have a HR PF for RHB of 1.15. Distance and wall height are inconsequential.

Did I misunderstand what you said?

One important point is that park factors should always be one season before the the year played. The PF for 2019 won't be available until the end of the season.

This is how I thought i should work:
Pitcher (fly ball pitcher)
Hitter (fly ball hitter)
AI - hit at x velocity to park dimenision xyz.
APPLY park dimensions
Does hit exceed park dimensions? then HR if not apply Fielders outfield ratings.
return result
log all data, data used in stats and pbp and animation.

[Reed]

Park factors for a stadium DO change over time. As teams move to new Cities/stadiums or as there is expansion, those new teams/stadiums will have there own park factors that affects the park factors of the other stadiums. For example Coors field could become a pitchers park if all the other teams moved to a very friendly hitters park.

[DawnBTVS]

Originally Posted by Reed

Park factors for a stadium DO change over time. As teams move to new Cities/stadiums or as there is expansion, those new teams/stadiums will have there own park factors that affects the park factors of the other stadiums. For example Coors field could become a pitchers park if all the other teams moved to a very friendly hitters park.

On top of that, even "minor" cosmetic changes can have an effect on how a given ballpark plays for a season or two. Moving fences in from 375 feet to 350 feet would definitely boost home runs irregardless of player's power or pitcher's ability.

https://bleacherreport.com/articles/...ome-run-totals

[Syd Thrift]

Originally Posted by fredbeene

All parkpark is static. forget the players. its all about the design.
a 100 foot line will generate more more homers than 1000 foot line. (over the fence homers)
the modifers should be based solely on the ballpark..
WIDE fould lines gives more chanes for outs.

Everything else is random.
Only the build of park should effect.

Great defenese should have huge advantage in large park.
But it seems all we do is guess what the engine does, sadly

This isn’t how baseball works. First up, the single highest correlation with offense isn’t wall distance or height or some algorithm based on the two but altitude. Period. Coors is one of the biggest stadia in the league and has consistently been by far the most hitter friendly. The stadium the Diamondbacks play in is also big and also allows a ton of homeruns.

Number two is climate, and only once you’ve drilled into climate does the size of the park become a thing. At that, if there’s a prevailing wind it can turn a park that dimension wise ought to have been average into a pitching heaven (see: the Big A). Sometimes the product a wall is built with produces more doubles and so on than the fact that it exists, as was the case with the Green Monster, which went from having a steel framework to being all padded or scoreboard in the 1970s. And then of course there’s the batter’s eye, which is more we’ll maintained now than it used to but which still can vary a bit.

Many times the actual dimensions of a park are the least interesting thing about it in terms of how baseball is played in it.

[BirdWatcher]

Originally Posted by Syd Thrift

This isn’t how baseball works. First up, the single highest correlation with offense isn’t wall distance or height or some algorithm based on the two but altitude. Period. Coors is one of the biggest stadia in the league and has consistently been by far the most hitter friendly. The stadium the Diamondbacks play in is also big and also allows a ton of homeruns.

Number two is climate, and only once you’ve drilled into climate does the size of the park become a thing. At that, if there’s a prevailing wind it can turn a park that dimension wise ought to have been average into a pitching heaven (see: the Big A). Sometimes the product a wall is built with produces more doubles and so on than the fact that it exists, as was the case with the Green Monster, which went from having a steel framework to being all padded or scoreboard in the 1970s. And then of course there’s the batter’s eye, which is more we’ll maintained now than it used to but which still can vary a bit.

Many times the actual dimensions of a park are the least interesting thing about it in terms of how baseball is played in it.

As someone who has watched games at Coors Field since its inception (moved to Denver in the Autumn of 1995, so missed most of first season but made it to Coors for at least one game that year), it is partly the size of the park (or more precisely, the size of the outfield) that makes Coors such a great hitters park. Yes, altitude is the key. But not only because it means the balls fly out of the park much better, but because it is the reason the outfield is vast (to try to mitigate that effect) and as a result Coors yields a great many bloop singles that would be outs in most parks much of the time. It is why outfielders with great range are even more valuable and needed at Coors (and not always a commodity the Rockies are very good at developing or acquiring.) Coors still inflates power numbers, yes, but since the humidor era began it probably even more strongly influences, in a positive direction, batting averages.

While I largely agree with you, Coors Field is a rather extreme example that illustrates that park dimensions can in fact influence offensive production.

[Syd Thrift]

The HR rate is cut in half compared to the pre humidor era, sure, but the extra hits are actually *not* attributable to just the big outfield. Due I’m sure in part to the fact that breaking pitches rather famously don’t break as much in Denver as they do at sea level, K rates are consistently about 90% of the norm through the rest of the league (an effect that unfortunately OOTP does not track).

[BirdWatcher]

Originally Posted by Syd Thrift

The HR rate is cut in half compared to the pre humidor era, sure, but the extra hits are actually *not* attributable to just the big outfield. Due I’m sure in part to the fact that breaking pitches rather famously don’t break as much in Denver as they do at sea level, K rates are consistently about 90% of the norm through the rest of the league (an effect that unfortunately OOTP does not track).

No doubt about breaking pitches being less effective at Coors.
As with anything, many factors combine to create reality. And it's not always easy to untangle every thread.
But I'm arguing, and yes mostly on the basis of observation and anecdote (so I make no pretense of having hard data to back this up) that one of those factors is the great many bloop singles that drop into the vastness of the Coors field outfield without the opportunity for any fielder other than those with tremendous range being able to reach them and convert them into outs.
So to clarify-I am not arguing that all of these extra hits (not just bloop singles, but also line drives into the gaps for extra bases, which admittedly I didn't mention until now)- are ONLY because of the vast outfield spaces. I'm merely arguing that this is one, not insignificant, factor.
And I totally accept that I may not be correct about this. But I'm not without Coors Field observational experience.