Minor League Splits > About |
[ FAQ ] [ Glossary ] |
Search for Player:
[ Brew Crew Ball ] Park Adjustments
All player splits tables now include a line for park adjusted stats. Those stats aim to show how that player would perform in a "neutral" park in his league. They don't, however, correct for league differences, such as the well-known offensive nature of the Pacific Coast League. The park factors I am using are, in most cases, a weighted mean of the last four years (including 2006). For 2003-05, I am using Dan Szymborski's factors, which are weighted 8/5/3 for 05/04/03. I average his numbers with 2006 factors, resulting in a 16/8/5/3 weighting for 06/05/04/03.
Dan didn't provide rookie league factors, so rookie league stats are adjusted only using 2006 factors. These can be quite volatile.
I am calculating 2006 factors from my own stats, and doing factors for five stats: H, 2B, HR, BB, and K. I don't have handy all the necessary information to do park factors perfectly, so I approximate as follows: [(HE+HEA)/HIP]/[RE+REA/RIP], where HE = home events (e.g. doubles by the team in question at home), HE = home events allowed, HIP = home innings pitched, RE = road events (e.g., doubles by the team in question on the road), REA = road events allowed, and RIP = road innings pitched. As I say, I'm aware this is an approximation and it has its flaws.
To adjust each player's stats, I adjust their home and road stats independently. To their home stats, I apply the home park's factors. To their road stats, I apply a cumulative "road factor" derived from the home team's factors. For instance, if a club in a ten-team league has a HR factor of 110, the appropriate road factor is 100-(10/10) = 99. All numbers are then rounded to the nearest integer value before calculating rate stats.
One last thing, because I know someone will ask. When adjusting for walks, a player's at-bat total can be different after park adjustments. This doesn't mean he's come to the plate more in a neutral park; it means a different number of his plate appearances would be walks, and thus a different number would result in at-bats, as well.
To see the latest park factors I'm using for 2006, click here. These are the weighted factors for 03-06.