Last Friday I commented that despite the recent fall in ERAs in MLB since 2000, the variance relative to the mean ERA has risen. Using an assumption from Stephen Jay Gould I assumed that this increased variance in pitching performance indicated an influx of some very bad pitching into the league. However, Skip was a little more skeptical.
Time for a skewness check! If Kurjian & Bagwell are right, the skewness should show up in the right tail of the distribution. If right skewness has increased in recent years, coincident with your increase in std deviation, then they are right. In which case we won’t be seeing those 50 home run years of the recent past. I like Gould, but think Kurkjian is on to something an give him the edge here.
Great idea! While I am no skewness expert, I did run a few tests. The histograms seems to be quite informative. Here are frequency distributions for ERAs in 2000 and 2003.
Skip is right. The skewness indicates that pitchers seem to be improving. While some very bad pitchers still play in the big leagues, overall most pitchers seem to be getting better.
Also, see the post that started it all on The Sports Economist.