Monday, July 30, 2007

Home Run Derbies and Sophomore Slumps

The "sophomore slump" in sports occurs when the Rookie of the Year winner sees his performance decline in his second season, even though most players improve from year 1 to year 2. Sabermetricians know that this is not a real effect but rather the impact of regression to the mean.

Among all the first-year players in the league, the Rookie of the Year is likely to be one who has played "over his head" with statistics that are better than his inherent level of talent. The next year, he doesn't figure to be as fortunate, and his numbers are likely to decline somewhat. This effect is not limited to rookies; the league leaders in batting average and home runs will usually see those figures decline next year as they regress to the mean.

There is another example of regression to the mean in baseball: the so-called "Home Run Derby curse". I'm not sure if this term existed before Bobby Abreu, but now some players and analysts believe that participation in the Home Run Derby causes that player to suffer a power outage in the second half.

Beyond the Box Score tackled this issue today. Their conclusion--with sufficient sample size warnings--is that participants see their home run output decline slightly in the second half. But this is exactly what we should expect! In order to get an invite to the Derby, you must have a prodigious home run total on July 1. (Or hope that they repeat the 2005 format and you were born in Australia.) If you are among the best of the best, you are likely to slow down in the second half.

I would not be surprised one bit to see that the population of All-Stars declines as a whole in the second half of the season, given that many All-Stars are simply there because they've had the best three months of their life. If you're batting .378 at the break, are you really likely to keep that up?

This kind of "curse" logic is to be expected from an ESPN announcer, but I hope that the guys at Beyond the Box Score know better.

No comments: