Since you read Fan/RotoGraphs, you’re probably in the habit of looking at any number of peripheral stats while searching for your next streamer or roster plug. Allow me to add to the noise as you attempt to talk yourself into taking a chance on certain pitchers, while talking yourself out of others.
You might look at K%, for instance, which is a much better indicator of a pitcher’s usefulness than K/9, though K/9 is a much more common stat in that a number of sites will list it, and is included in lieu of K% on the default leaderboards at FanGraphs itself. Because K% displays strikeouts as a percentage of batters faced as opposed to the number of outs recorded (essentially what K/9 does), it is a better indicator of how efficient a pitcher is at striking batters out.
Understandably, then, K% correlates with IP, W, ERA, and WHIP at a significantly higher rate than K/9 does. Because I am an obsessive baseball lout, like you, reader, I recently explored differences in pitchers’ K/9 and K% just for the hell of it; while the results where not surprising, they were interesting. The results can be found here.1
What was slightly more interesting was when I went started fooling around with F-Strike% and saw how it correlates with certain other 5×5 stats. Generally, it correlates well with those stats, especially WHIP and IP. Eno Sarris explored the effect of F-Strike% on BB% last summer, and found that F-Strike% “explains almost half of the variance in walk rate.” This jives with the -0.401 correlation between F-Strike% and WHIP that I found.1
Read the rest of this entry »