2000 MLB park factors

This is my calculation of Major League Baseball Park factors for the 2000 season. A park factor is a coefficient that represents the ratio of how frequently events happen in a particular park with respect to an average park. For example if a park is very extreme like Coors field it could increase runs by 50% over an average park thus it would have a park factor for runs of 1.5.

If the schedule is balanced a park factor is easy to determine because we can compare two subsets of games and compare. One subset would be games in that ballpark. The other subset would be tthe road games for the tenant of the park in question. All the games include the tenant of the ballpark and an opponent.

If the schedule is balanced the all opponents show up in both subsets the same number of times. This makes the relative strengths and weaknesses of the teams playing in the games a minimal factor on the frequency of any event, such as runs scored, when we compare the subsets of games.

The problem is, however, with the coming of interleague play the schedules are no longer balanced. Most interleague opponents will appear in a 3 game series in one subset and not the other and league opponents are played a different number of times with a different number of those games at home than on the road. The problems in comparing the two subsets were increased this year with the unbalanced league schedules. So this is an attempt to use a more complicated algorithm to calculate the park factors. This algorithm also gives us a chance to actually compare the parks in the two leagues because the interleague games are included in the calculation.

The goal of the algorithm is to look at the factors on a game by game basis. Assume that the expected runs in a game are dependent on five factors. These factors are the offensive capabilities of the two teams, the defensive capabilities of the two teams, and the park’s affect of runs scored. Both the offensive and defensive capabilities of the two teams are treated as coefficients just like a park factor. So the expected runs for the home team would be:

Home offensive rating * Road defensive rating * Home Park Factor * League average runs scored

The offensive ratings, defensive ratings, & park factors are calculated to minimize error. So we end up with a very large system of equations which can easily be solved by a computer.

There are obviously a few things that could affect that could affect this outcome such as starting pitchers. Hopefully this evens out over the course of a season. The affect of the DH on runs scored in that park is something that is normally taken out of the equation by eliminating interleague games. For this purpose, and as a first effort on anything to due with sabermetrics, I decided to keep it in since the DH is a consistent rule in a park it is a factor in that park. The factors that can be used in comparing pitcher ERA’s in two different leagues but every pitcher is going to have a different road park factor that is not necessarily close to neutral or even close to neutral after the affect of never pitching in the home park on the road is removed from the equation,

The following is a list of Park Factors related to runs scored for the year 2000.

Bihl Rankings Home