Contact% and Swing Strike% Correction
It’s come to my attention that the 2010 Contact% and Swinging Strike % numbers looked a bit off. The 4 stats impacted (Z-Contact%, O-Contact%, Contact%, and SwStr%), will be corrected in tonight’s data load are now fixed.
David Appelman is the creator of FanGraphs.
I misread that as correlation rather than correction and got excited for poignant analysis.
I did as well. But this is cool, too. I was curious about the contact% numbers.
Me too. *crosses fingers* for article on contact % and Swing Strike % correlation.
Who notices this kind of stuff and how do you figure out whether or not their suspicions are legit?
Well, with Contact% and Swinging Strike %, it was pretty obvious and there was an unexpected change to one of our source files this year, which kind of broke the code we were using.
Here’s a question for potential analysis: I think its conventional wisdom that a pitcher who faces the same team two starts in a row (like Luke Hochevar facing the Tigs tomorrow, whom he opened the season against last week) will do poorer in his second start since the batters now “know him”. It make sense: the more pitches you see against someone, the better you can begin to get their timing down, read how the ball comes out of their hand, etc. But just because its intuitive, doesn’t mean its true. Can you confirm or debunk? Thanks!
I shall posit, with absolutely no numerical proof, that these back to back starts will correlate heavily to any statistically significant pattern in “first-time through” vs. “second-(or greater)-time through” the order.
Might be a problem with the “pitch type” data as well. Seeing some odd numbers in the XX% column for some pitchers.
Pitch data runs about 1 day behind, so all of yesterdays games’ pitches will be labeled XX.