More Travel Means More Wins?
August 1, 2011 by Jess Coleman · 6 Comments
Realignment has become a serious issue in the ongoing labor talks in Major League Baseball. Everything from turning the Houston Astros into an American League team to abolishing divisions has drawn serious consideration. One issue, though, stands above all in complexity and also seems most logical.
That issue has to do with evening out the amount of miles traveled for each team. Data from Paul Robbins of The New York Times finds that the Dodgers traveled nearly 60,000 miles in 2009 while the Nationals traveled just over 25,000. That is a very wide range, and it is a legitimate concern for teams that are stuck with longer hours on planes and less time to relax and practice.
Initially, one would make the argument that a team that travels more often is hit with an unfair burden. Skipping through time zones and spending more time traveling seems unfavorable. However, the data suggests something different.
Consider the following graph showing miles traveled versus wins for the 2009 regular season:
Shockingly, the data is flipped. Although the correlation isn’t particularly strong, there is a noticeable effect of longer plane rides. But it isn’t what you would think. It turns out that more traveling time actually increases your chances of winning.
I can not figure out why this occurs, and I am open to anyone’s thoughts as to why more traveling miles would actually help a team succeed. Maybe players are less stressed being away from home, or the constant travel helps maintain focus.
In any case, the correlation is there, and, although the data is inverted, there is still a case to be made that miles traveled should be equalized. Surely this isn’t easy, and you cannot change the fact that a team like the Mariners that is isolated in the northwest will need to travel more than a team like the Nationals that is in the middle of the baseball-heavy east.
But MLB can work to manipulate the schedules so a team like Seattle makes only one trip to the east coast. It would mean playing many teams twice within a short period of time, but it can surely be done. Whatever the solution, there is clearly an effect created by traveling time, and it is certainly something MLB should look to resolve.
I think this is probably just a case of the teams that traveled most (west coast teams, etc) happening to have a good year in 2009. If you look at it across multiple years, I would be surprised if the trend was pronounced in either direction.
Which in and of itself would prove that travel distance has essentially no effect on team performance, relative to all the other factors that go into baseball success or failure.
The obvious answer is that there is no reason for those to be related, and they probably aren’t. There is also a strong relationship between sunshine and proportion of the population that voted for George Bush, that doesn’t mean anything. Most likely, the teams with more talent for whatever reason traveled more and thus won more games.
I’m guessing because teams on the coasts in your sample have more talent than teams off the coasts. You should be controlling for talent level. The appropriate method would be to plot the difference between actual wins and expected wins (use the Log5 method and third-order win percentage) against miles traveled.
My guess is that you won’t find anything, that travel probably doesn’t have much of an impact.
One of the few things I remember from statistics class: correlation does not equal causation! If this trend could be demonstrated over say 20 years, then I’d say that something’s afoot.
Clearly teams that are on the west coast and had to travel more had strong seasons, that’s the whole point. That is not just a statistical anomaly. It proves that more traveling does not negatively affect teams, because those teams had strong seasons in spite of their long travels. Whether it increases your chances of winning is a bit more nebulous, but it’s a strong enough correlation that there needs to be more research to prove miles traveled does or does not have an effect.