Beyond the Box Score

(An updated version of this article can be read here.)

We are all familiar with the idea that across major league baseball some parks favor hitters while some favor pitchers. The dimensions of the park combined with other features can increase or decrease the run scoring environment.

After opening Citi Field in 2009, the New York Mets saw a significant decline in their ability to score runs. The team averaged 681 runs scored per year from '09-'11 compared to 812 over the previous three seasons. Additionally, management worried that pitchers were developing bad habits while pitching at their spacious home park that led to worse performances on the road.

Let's set aside whether we think that reasoning is legitimate. The question I came away with was whether the kind of home park a team has increases or decreases their chances of winning games. If a team has a pitcher-friendly home park, does that create issues for them on the road that they cannot overcome? What about a hitter-friendly park?

First, I took all ballparks from 2004 through 2010 and calculated their average park factors for run scoring based on ESPN's park factor data. Next, I calculated the average number of wins over that time period for the home teams playing in those parks.

The correlation between run scoring park factors and wins for each of the 35 parks was .11--so essentially, no relationship. The kind of park a team played in did not significantly impact their ability to win games (positively or negatively).

But what about at the extremes? Part of the Mets' reasoning was that their park had become so pitcher-friendly that their hurlers developed approaches that got them into more trouble on the road. There was also talk of hitters changing their swings and approaches based on Citi's unforgiving confines (e.g. David Wright).

To examine this, I bucketed the parks into three categories--hitter-friendly (run park factors >= 1.06), pitcher-friendly (<=.94), and neutral (between 1.04 and .95). There were 7 hitter-friendly parks, 9 pitcher-friendly parks, and 19 neutral parks.

Here are the correlations between each type of park and wins:

Hitter-friendly: -.42

Neutral: -.25

Pitcher-friendly: .10

Now, this is just a rough first-cut, and one based on some very small N sizes, but the results are interesting. For the most extreme hitter-friendly parks, the average number of wins was negatively correlated with the degree to which runs were more easily scored. So the more extreme the run environment, the harder it was for those teams to win games. Neutral parks exhibited a negative relationship as well, but it was half the size of the hitter-friendly parks and the amount of variance it explains was only a third of the hitter-friendly parks as well.

The pitching-friendly parks exhibited essentially no relationship. In terms of a team's average number of wins, pitcher-friendly parks are essentially neutral. This runs counter to the Mets' reasoning for moving in the walls at Citi Field. It also suggests that if you are constructing a new ballpark you might be better served building it so that it plays neutral to pitcher-friendly.

There are, of course, many more ways to look at this question. For example, I wonder whether the park itself isn't as relevant as how well you select players to fit the specific park that you are in, or the extent to which you can identify players whose talents show less variation relative to their environment. That's a longer study to be sure, and hopefully I will get to it one day.

For now, we can say that--at least directionally--teams shouldn't be blaming their ballparks for the lack of wins.