Wednesday, September 3, 2008
tRA and the 2008 Yankees
By way of Baseball Think Factory, I was introduced to a new stat called tRA. What is tRA? It’s a way to evaluate pitchers based on their peripheral stats and batted ball types allowed, developed by Graham MacAree who writes for Lookout Landing. He also has a new site called StatCorner which looks like it’s going to be a cool site for more in-depth statistical analysis.
What tRA does is assign run values to each event a pitcher allows, but removing the defense. It’s more in depth than FIP or DIPS which regress BABIP to league average, because it makes use of an important factor that FIP and DIPS ignore. All batted ball types are NOT created equal. A pitcher who allows more line drives should allow a higher BABIP, all other things being equal. So ground balls, fly balls, and line drives are given average run values for those events.
tRA also assigns out values for each event that a pitcher allows. The events are park-adjusted and out values are for what an average defense would do, so what tRA should give us is the net performance of a pitcher in a neutral park with an average defense behind him, but with more information than something like FIP, xFIP, or DIPS gives us. If you want to read more about the actual run values and out values and park adjustments, it’s all discussed at this link.
Here’s what tRA says for the 2008 Yankees.
RV: Run value of all events.
OV: Out value of all events.
tRSAA: Runs saved above average using tRA
tRSAR: Runs saved above replacement using tRA
tWAR: Wins above replacement using tRA (tRSAR/10)
This doesn’t account for leverage, but Mo is still the best pitcher on the Yankees by this measure. The Yankee pitching staff scores decently by this measure, but when you factor in a team defense that’s 48 runs below average overall by zone rating, you go from a run prevention unit that’s 15 wins better than replacement to one that’s 10 wins better.
And lastly, here’s how the AL teams stack up using this measure.
Seems about right. The White Sox raw pitching stats may look middling, but when you factor in their ballpark they may be the best pitching staff in the league. Toronto and Tampa have them beat in ERA+, but that’s most likely a function of defense (Toronto is +48!!! by zone rating as far as runs saved, and Tampa is +8, the White Sox are +3).
I think this a pretty cool way to look at pitching in more depth, but I’d be interested to see what everyone else thinks. The methodology certainly seems sound to me.
Page 1 of 1 pages: