Table of Contents

Monday, March 17, 2008

Interviews with Padres' Stats Guy

MB, who runs Friar Forecast (among others) and occasionally has commented here, has scored two really interesting interview pieces with Chris Long, who is the San Diego Padres' Senior Quantitative Analyst. They're good reads, and shed some light on some of the advanced analysis work going on in some major league teams.

Here are the two parts of the interview:
And here's my favorite exchange from the second part of the interview:
FF: Are you involved in integrating what I’ll call scouting data (pitch speeds, mechanics, body size, time running from home to first, etc) and more traditional numbers (hits, home runs, runs allowed, etc)? Or do you work more with the traditional numbers and use the scouting-based data separately?

Chris Long: If you aren’t looking at *all* the information you have, and trying to extract the absolute maximum amount of value from that information, you aren’t doing a good job. Hopefully that answers your question.

FF: I imagine one of the main goals of statistical analysis in an mlb front office is to project a player’s future performance. If so, can you tell us a little about how you analyze a player’s past performance to gain insight about what you expect out of his future?

Chris Long: I can’t go into specifics, but again, you need to integrate *all* the information you have about a player. Not just what he’s done, but where and how. Did something change? Was it luck? A simple projection system that’ll outperform almost every human (on average) is fairly easy to build. Building a system that outperforms those systems is harder, but consequently more interesting.
One of the things I've been talking about more and more (most recently here) is the importance of scouting information as another crucially important set of input data when trying to evaluate players. That's why, for example, I explicitly included the Fans Scouting Report data in my composite fielding estimate from my player value series. Some day, we may also be able to include information derived from pitchf/x (or other "scouting" sources) in evaluations of pitchers...and maybe hitters as well. It's the future, man. :)

....

Also, I went to the Cincinnati Reds' front office page to see if they had anyone listed who might be somewhat comparable to Chris Long. Sam Grossman came up as the Baseball Operations Analyst...anyone know anything about him and what he does? I might try asking some of the beat writers about him.