[pstrickert]

I don't know if this will help, but here's what Garlon said about park factors:

I created the ballparks file myself for OOTP and Markus implemented it into the game. It took me months of work to complete it all. The ballpark factors are indeed accurate and yes OOTP will update the factors and the changes in ballparks as you progress thru history as long as you have "Automatically Adjust League Strategy" selected in the League Setup>Options screen.

The ballpark factors from 1901-1956 were created by running the park's scoring factors thru a quadratic formula along with the league's offensive totals to generate a Batting Avg and HR Factor to reproduce the park's scoring environment. From 1957-present we have the actual park totals of all offensive categories so I was able to calculate the BA/2B/3B/HR factors for the parks directly from the data on a per AB basis. The Bill James' method of weighting the surrounding seasons was also used to provide a bigger data sample size for the park factors.

I suggest playing with Neutralized Player stats when using the park factors.

[BaseballMan]

Originally Posted by pstrickert

I don't know if this will help, but here's what Garlon said about park factors:

I created the ballparks file myself for OOTP and Markus implemented it into the game. It took me months of work to complete it all. The ballpark factors are indeed accurate and yes OOTP will update the factors and the changes in ballparks as you progress thru history as long as you have "Automatically Adjust League Strategy" selected in the League Setup>Options screen.

The ballpark factors from 1901-1956 were created by running the park's scoring factors thru a quadratic formula along with the league's offensive totals to generate a Batting Avg and HR Factor to reproduce the park's scoring environment. From 1957-present we have the actual park totals of all offensive categories so I was able to calculate the BA/2B/3B/HR factors for the parks directly from the data on a per AB basis. The Bill James' method of weighting the surrounding seasons was also used to provide a bigger data sample size for the park factors.

I suggest playing with Neutralized Player stats when using the park factors.

Well Garlon did do a good job for those who wanted to just start up a historical league. But others like me want to fine tune it to their taste. Like Garlon made it easy to start a 19th century league but i use the reams file i edited because i want the chaos of the 19th century in my league.

Anyways should avg and hr be the same? It would seem that hrs should be lower? should ballpark factors have that much effect on league totals. I always assumed it just ment more hrs or 2b would be hit at that park but the league totals would be unchanged. So if i put yankee stadium II in 1961 then the 1961 yankees would not hit a 100 more hrs but maybe just a little more at home.

[StyxNCa]

Originally Posted by BaseballMan

Whats the use in creating historical ballparks if they wont work as they did in real life for that year? Isnt that the point of historical ballparks. Otherwise why not just have all the parks generic.

Good question.

[SteveP]

Originally Posted by BaseballMan

Whats the use in creating historical ballparks if they wont work as they did in real life for that year?

IMHO, this question makes sense only if: 1) you know how those ballparks worked IRL, and you're frustrated by the game's not making it easy for you fix things; or 2) you're convinced that other people know how those parks worked, and they failed to do the job or were ignored. If it's #1, you're best strategy is to provide convincing statistical evidence that the game is getting it wrong.

My view is that Garlon and Markus took their best shot at implementing the available knowledge. What they didn't do was make their method readily transparent to the rest of us. It's possible that there was no simple method for doing that -- and understandably no desire to implement a complicated method that then required extra user support and software maintenance.

[oman19]

Maybe I'm not seeing it right but if you go to 2008 and look at all the stadiums each stadium is differnet but every park effect has the same value. So no L/R difference at all. Same chance at a HR as a double. Hell the doubles and triples info is easy enough to get just go on espn they have that there. The L/R Hits and HR factors do present a challenge but if you had a team of people working on them it wouldn't be to hard. Retrosheet has the info. If I had the computer knowldege to work with their databases I would figure it out myself but alas I download their databases and is all french to me.

[SteveP]

OK, so let's try to fill out the complete picture:

1. Player hitting and pitching ratings, when created using real (unneutralized) stats already incorporate the ballpark factors of the ballparks they played in. If the home park was HR friendly, they hit more HRs, and that fact is incorporated in the Power rating. Ballpark factors in OOTP are not required to produce these results.

2. Garlon undoubtedly assumed (or at least promoted the idea) that when OOTPers used real stats, they would neutralize the ballparks, and that when they used ballpark factors they would used neutralized stats for ratings. He also assumed OOTPers would always use recalc. People don't do those things, of course, but those assumptions aren't unreasonable.

3. Ballpark factors in OOTP probably don't have as much of a distorting impact on resulting stats (when real stats are used for ratings) as Garlon and others might think (because of trades, free agency, etc.). But the effect is to distort the results from historical reality. The ballpark factors do not make for more realistic results.

4. If you are using neutralized stats, and you think that the ballpark factors in the game are not producing proper stats, then you can probably get this fixed by providing the evidence.

5. Item of speculation: it's possible that the ballpark factors in the game do not have L/R splits because the neutralized stats were created in a way that neutralized for L/R stats.

Bottom line: there's nothing wrong with people wanting to experiment with different ballpark factors. But anyone who assumes that the factors in the game are producing "wrong" results needs to consider the context in which these factors are operating.

[spleen1015]

http://www.bb-bbcf.net/fool/stade.xls

[Questdog]

Originally Posted by spleen1015

http://www.bb-bbcf.net/fool/stade.xls

The numbers in that file are ludicrous....

[pstrickert]

Thank you, SteveP, for clarifying the issues. I know Garlon has a neutralized era_ballparks file, which he offered to someone. Does anyone have that file? (Garlon, BTW, is MIA for awhile.)

EDIT: He offered it to PhillieFan. I've contacted him for a copy of the file.

[spleen1015]

Originally Posted by Questdog

The numbers in that file are ludicrous....

How so?

[Questdog]

Originally Posted by spleen1015

How so?

There are too many instances to list, but you will be hard put to find a listing of a ballpark in there that has historically accurate factors that reflect the park's true characteristics. All of them are merely calculated from the fence distances.....

At least one park actually has a negative factor. How is this possible? The park gives triples to the team not at bat?

[SteveP]

Originally Posted by Questdog

At least one park actually has a negative factor. How is this possible? The park gives triples to the team not at bat?

That's easy: you just subtract triples that a player hit in a different ballpark.