Saturday, July 5, 2008

Finding Stability in Ground Ball Rates

Of all the pitching stats out there, ground ball rates seem to be the most stable across all levels and throughout a pitchers career. Unfortunately, we don't have ground ball rates for a lot of histories greatest pitchers, but we do know that high ground ball rates can only help a pitcher, and usually translates into 0.8 outs per every ground ball hit (whereas a strike out translates into 1.0 outs).

A lot of baseball fans -- and even players and teams themselves -- tend to overlook ground ball rates and don't give them the credit they deserve, but guys like Greg Maddux, Brandon Webb, and Fausto Carmona are slowly changing all of that. High strike out rates are nice, but they experience a sharp decline during the first few years a pitcher is in the majors, and high strike out totals can often lead to increased pitch counts, which in turn lead to arm fatigue, injury, and shorter career spans. Ground ball rates usually do not significantly decrease once a player is called up, and they can occur during any point of an at-bat, leading to lower pitch counts and lengthier careers.

The only problem I have with ground ball rates is their rare unpredictability during the move from A ball to higher levels of competition. While most pitchers don't see a wide range of variation, occasionally a player comes along that experiences a 10% or more drop off and I still can't figure out which players are most likely to see this happen and why it happens. For instance, Michael Bowden had a 57% GB% after 108 innings at A ball, but has since posted a GB% of 45%, 45%, 38%, and 43% in stints at A+ and AA ball. I can't find many pitchers that experience such a heavy decline, but the fact that it happens so inexplicably bothers me.

Despite the slight hiccups that GB% rates can undergo, I still think they are perhaps the second most important stat behind K%. There is sufficient research to back up my opinions, but research aside GB% rates provide a level of stability that is hard to find in other stats.

No comments: