OF defensive park adjustments
I don't have them, yet. But I did put together outfield positions totals by ballpark for 1987 to 2006.
OF totals by park and position
Only one park has more than +10 runs per year, and just barely at that. There are a lot more parks with positive values, which is what I expected. No park really makes it easier to catch flyballs, at least by much, but some parks make it very difficult. So while a few parks will have large negative vlaues for ZR, like Fenway, Enron, Camden Yards RF, most parks/positions will show a slight positive value, since this is a zero sum game.
Some of the parks with large negative values are no longer in use, such as Mile High and Exhibition (Blue Jays for you newcomers). Pretty much the current in use parks with negative values are Fenway, Enron, and Joe Robbie LF, and Camden and Metrodome RF, as expected. Others I did not expect are PNC right and left, and all the Coors positions. Great American ballpark CF is almost certainly due to Griffey being the primary player there, not a park effect. PNC might just be due to bad Pirate OF's, but its worth checking out. I though STATS redefined the zones used for Coors, so I'll have to check and see if the low ZR values are coming from early years or if there is still a problem with their data.
Parks may not be listed under their "official" name, as I don't respect corporate naming rights unless I get a share of the money. Sometimes I use the corporate name, its just whatever I like best in any case, for example I prefer "cellular" to "Comiskey II" simply because its easier to distinguish from the original Comiskey.
6 Comments:
I think that PNC might be legitimate, at least in left field. I think there was some talk a few years ago that they'd rather have Bay in LF than CF (when Redman was the CF) because Bay was the better fielder.
Plus, looking at the overhead view, it looks like a tough field to play.
Are those run values only for that park or in divided by two form (e.g., would Manny be 20 runs better than his ZR suggests, or 10 runs)? I'm assuming the latter...
Remember two things: (1) You have to regress to the mean (same way you would with a projection), and (2) You need to account for the quality of the players playing that field (which you already note).
In general, except in a few places (Coors, Safeco, Fenway), I think park adjustments are pretty useless.
Didn't divide by 2, that's just the total ZR runs per X chances for that park and position.
These are NOT park factors, I'm not quite there yet but it will be coming soon.
If I understand this, these are the sum of the "raw" season totals for the home team fielders, rather than the park-factor-like matched innings study you did earlier?
You are correct, Joe.
The matched innings study for all teams is in the works.
Bay is league average at gloving balls in play but significantly below average on the number advancing runners he allows, even with Jack Wilson running halfway to Bay's fielding position and Wilson's cannon arm.
That aside, I don't really understand what you are measuring here Chone. I suppose I haven't read enough of the blog to figure it out yet..
CH INN RV peryr.. I assume this is:
total chances
innings a chance occured in
run value (of what, I'm not sure -
the BiP fielded by the defender?)
average run value per year allowed
am I close?
Jake at Bucco Blog
Post a Comment
<< Home