April 18, 2001
The Daily Prospectus
Even More on Park Factors
Last day we'll deal with this, I promise.
One thing I probably should have included in yesterday's column was a brief explanation of how park factors are calculated. A lot of the mail I received attempted to tie Veterans Stadium's high walk park factor to the composition of the Phillies' staff. While that seems to make intuitive sense--and is a common misconception--it is not the case.
A park factor is not influenced in any significant way by a team's abilities. Park factors compare the results of the games played in a team's home park to the results of that team's road games. Park factors include all the statistics generated by both teams; that is, the performance of the Phillies in their 72 (see note below) home games against NL teams counts just as much as the performance of the opponents in those games. The totals in those games are compared to the totals in all of the Phillies road games to arrive at park factors for various occurrences.
(Note that STATS, Inc. only uses games within the league in calculating park factors, surmising correctly that the differences in interleague opponents would skew the factors.)
I can't emphasize this point enough: park factors are independent of a team's composition. Having a good offense or a good pitching staff will not skew a park factor, because the performance of that offense and that pitching staff 1) accounts for only half the performance in a given park and 2) is being compared to the performance against the same set of opponents in the rest of the league.
As an extreme example, let's look at the San Francisco Giants in 2000. The Giants finished third in the league in runs scored, and first in runs scored on the road (in NL games). They had the best offense in the league as measured by Equivalent Average. If a team's performance can have an undue influence on park factor, we would expect that Pacific Bell Park would seems like a hitters' park, or at least a neutral one.
Pacific Bell Park was the third-best pitchers' park in the NL last year. The Giants and their opponents combined for 672 runs in 72 games at the Bell (9.3 per game), and a whopping 831 in 75 games on the road (11.1 per game).
As far as the Phillies are concerned, their pitchers walked 305 hitters at the Vet, and just 258 away from it. Their batters walked 278 times at home, and 261 times on the road. The totals of 583 walks in Philadelphia and 519 elsewhere on the continent yield a walk park factor of 110. The large gap between the Phillies' pitchers totals at the Vet and on the road suggests that the staff isn't the driving factor. They walked more batters in home games than all but two other teams (the Pirates and Brewers), but were 12th in the NL in walks allowed on the road.
I don't want to dwell too much on the Phillies, or on any specific pieces of data. The point I want to emphasize is that park factors are real, and not subject to influence by a team's characteristics.
Finally, did anyone else think it was cool that Barry Bonds's 500th home run was about as clutch as they come, turning 1-2 into 3-2 and a possible loss into a win? Bonds has taken far too much grief over the years for his postseason performance, while not getting enough credit for the otherworldly Septembers that have carried his teams to the postseason. That he got to be a hero with everyone's eyes on him was a great thing.
Joe Sheehan is an author of Baseball Prospectus. Contact him by clicking here.