Interesting article in Business Week on measuring player's performance. The buzz on the article on the SABR list and Baseballprimer.com that can be summed up as saying that this technique is nothing new being developed in 1970 and called Player Win Averages (PWA). Bennet and Albert go through the technique and add to it in Curve Ball talked about previously. Interesting that the Mets are said to be interested in hiring these guys to help them. So, my unofficial list of current teams that at least have shown an interest in sabermetrics (the Rangers did way back in the 70s) include:
1. Red Sox - Theo Epstein and hiring Bill James
2. A's - Bill Beane
3. Blue Jays - JP Ricciardi
4. Cardinals - from a post on mlb.com
5. Mets - scuttlebut from this most recent article
I'm drawn to these types of stats because they actually measure what really happened rather than having to assign weights to various events which is always going to be more of an approximation. Of course, stats like PWA (and RC/G for that matter) don't have that much predictive value on a per player basis. I think teams should be using these play by play stats to formulate general principles of strategy for their organization. For example, when should we sacrifice, when should we steal, when should we bring in a closer (ahead or tied and what inning?), when should we pinch hit etc. They should then hand these principles to the manager to implement and evaluate him on those. Play by play data would also be one way of quantifiably evaluating a manager's performance.
Friday, November 14, 2003
What's a player worth?
Posted by Dan Agonistes at 8:38 AM
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment