Relief Pitcher Wins and Categories

With so much good fantasy baseball information now available for free on the internet, or at little cost through books, magazines, and subscriptions, it can be hard to get an edge over your competition. One way that I try to find an advantage is by doing small research studies on fantasy-related topics that nobody has looked at. For example, because of the way baseball's scoring system assigns 'wins' to pitchers, I've always thought that relief pitchers would likely get a greater share of the team's wins in home games. Given that teams also win a greater percentage of games at home, I thought that there might be some potential advantage in certain formats to using middle relievers when they're pitching at home.

To test that theory, I put together a list of 18 good middle relievers, and added up their wins at home vs. their wins on the road. Because of the limited sample size, the results were inconclusive - 31 home wins vs. 26 road wins. Considering the rate at which home teams win games overall, that was a little less support for my theory than I expected. However, the study would be worth repeating for a larger group of pitchers.

My reason for expecting more relief wins at home can be illustrated with the following scenarios. Game is tied. Team is at home. Starting pitcher is pulled after 5 innings. Relief pitcher throws scoreless top of 6th inning, and home team scores in bottom of 6th. Reliever gets win. Now, same thing on the road. Starter is pulled after five innings. Team scores go ahead run in sixth inning. Reliever pitches scoreless bottom of sixth. Starter gets win. So basically, the middle reliever gets one extra inning of potential wins at home, and the starter gets on less inning of potential wins.

I think this is something that's worth studying further, using a larger sample size (more pitchers and for multiple years). Also, this effect would affect different pitchers differently, depending on their usage patterns. It would probably be worth categorizing the pitchers into 'types' based on usage, either using the six categories that Bill James used in his "2009 Relief Pitchers" article in The Bill James Handbook 2010
or basing it on some other combinaion of the 22 statistical categories he shows for each pitcher in the same article.