By Father Gabe Costa
» More Columns
Mr. Maxwell McDonnell is a student of sabermetrics and is really into statistics. In this episode of By The Numbers he uses a pretty high-powered approach to address the question of a home field advantage.
Maxwell McDonnell: At first glance, the question of whether or not the home teams maintain an advantage over away teams seems easy to answer. Of course they do! Diehard fans could never accept that their teams could fair equally well on the road as they do with the support of their home city. But is this really the case?
Logically, the effect of the crowd, travel considerations, and stadium conditions are by far going to favor the home team versus the visitor. Based on their travel schedules, players might be jetlagged and fatigued, putting the visiting team at a disadvantage, even before the first batter steps up. When the visiting team actually does take the field, they now have a stadium full of angry fans booing their every move. Combine this with the fact that there are environmental differences to which visiting teams must. For example, say our visiting team is Arizona and is playing in Colorado… now they have to overcome the substantial elevation gain of the Mile High City. And as if this isn’t enough, the home team is given the courtesy of batting last, which, in a close game, can make all the difference. Quite a way to start the day (or night)!
So, when considering all of the above, do these factors make a difference in the performance of home teams versus away teams? After all, the ballplayers are professionals, right?
The answer, for the most part is yes…with one notable exception.
In order to answer this question, and to put the qualms of weary fans to rest, I looked at home and away winning percentages for every team in the AL and NL per season from 2000 onward. The reason for choosing the year 2000 was that there needed to be enough accumulated data per team to conduct the analysis. In addition, I wanted to examine teams throughout the same time period.
With this in mind, in order to meet my standards, teams needed to have played in the same stadium since the new millennium, thus racking up a total 11 seasons worth of data. For this reason the following teams are exempted from my analysis as they do not maintain enough data at their current stadiums: New York Yankees (two seasons), Minnesota (one season), New York Mets (two seasons), Philadelphia (seven seasons), Washington (three seasons), Cincinnati (eight seasons), Milwaukee (nine seasons), Pittsburgh (ten seasons), St. Louis (five seasons), and San Diego (seven seasons).
After the winning percentages were calculated for both the home and away records, I found the difference between these two values for each season. These values were then averaged to compute a value that represented the historical difference in win percentages for that stadium during the last decade. In addition, I found the standard deviation of all the seasons’ differences.
At this point it gets a bit squirrely and you might have to recall topics covered in your high school or college statistics class. (If statistics wasn’t your thing, hopefully you have a little faith!)
To actually see if a team maintained a home field advantage (according to win percentage), I conducted a paired t-hypothesis test on the mean difference. Without getting too technical, all this seeks to do is determine if the difference between the average home and away win percentage is statistically equal or greater than 0. With this in mind, a difference of 0 would signify that a team is equally skilled on their home turf as they are away. Hence the home win percentage subtracted by the away win percentage would be zero.
If the reader accesses the two Excel files (one for each league) and the end of the blog, he or she will see that I calculated a t-value for every MLB team included in my analysis. Once again, if statistics wasn’t your friend, then please bear with me. These values were then compared against my critical value which in this case was 2.718. Note that a significance level of 1% was used. Hence, if a team’s t-value was greater than 2.718, the hypothesis that the mean difference is 0 would be rejected.
Bottom Line: This would signify that they do indeed maintain a home field advantage.
When I actually ran the analysis, it turned out that every MLB team maintains a home field advantage except for the Chicago Cubs. The Cubbies were the only team that had a t-value less than the critical value thus signifying that the difference between the home and away win percentage can confidently be asserted as zero.
So rest assured, unless you are a Cubs fan, know that you presence historically favors your team’s ability. If you are a Cub’s fan, cheer on but know that your efforts are fruitless.
Please note that as part of the t-distribution, the win percentages were assumed to be normally distributed. All data for the home and away wins were provided by www.retrosheet.org/boxesetc/MISC/PKDIR.htm and were crosschecked with www.baseballreference.com