After talking a bit with Eno Sarris about his pitch type benchmarks, we decided to change up the process a bit. Instead of taking the full average of all of the pitches thrown in a category, we limited the pool to only pitchers who had thrown 20 innings from 2011 to 2013. And we limited the number of pitches we counted to ones that had been thrown 50 times. Hopefully this takes out gimmick pitches and small sample anomalies, for the most part. We’re focusing now on regularly-thrown pitches from somewhat-established pitchers.
We also decided to take the median value within each pitch type. This is a better representation of what’s out there — the old way could have allowed the very excellent pitches to pull the benchmarks north of what could actually be considered an average pitch. We also decided to show you where the 40% benchmark was — the ‘good’ but not ‘above average’ pitches. There are a lot of pitches thrown here that are neither ‘show-me’ pitches nor are they ‘strong’ — think ‘useful.’
Eno will take a look at Matt Cain to bring this all into focus after the new table of benchmarks.