Video from my presentation at the 3rd Annual SABR Analytics Conference
Slides from my presentation at this year’s SABR Analytics Conference (click here)
I’ll be presenting some research on the consistency of pitcher performance at the annual SABR Analytics Conference. Here’s a summary of the research:
Talent evaluators and fans alike have long thought that there are differences in how players distribute their performances over the course of a season. Even if two players have identical statistics at the end of the year, how and when that production came about could be very different. It’s the basis for claiming that a player is good, but “streaky”. Or that you “never knew which player you are going to get” on a given day. While most tend to believe that such differences exist there have been few attempts to quantify this phenomena and better understand its causes, if any. The questions this research will seek to answer are: To what extent do pitchers differ in how consistent their daily performances are with their seasonal averages? And, if there are differences among pitchers in how their performances are distributed, what are the causes of consistency and volatility? Are certain types of pitchers more or less prone to being consistent?
A study of into the aging patterns of batted ball distance
Created an interactive spray chart tool, complete with data from the 2012-2013 seasons.
This past August I was a part of the Statistical Analysis panel at the 2013 SABR Conference in Philadephia, along with Dick Cramer, Steve Mann, Vince Gennaro, and Brian Kenny. Here’s video of the panel.
Generally speaking, umpires were less likely to call strikes on the Edge in pitcher-friendly counts and more likely to give those calls in hitter-friendly counts.
While we learned a bit from that analysis, it was really just the tip of the iceberg. There are a number of additional ways to cut the data, and that is the focus of this article. Count is just one dimension when we are thinking about what might influence the likelihood of close called strikes. There are a number of additional dimensions we can layer onto count, and that’s precisely what I show in the (admittedly large) table below.