Thursday, September 27, 2007

2007 MLB Year To Date Fielding Data

I had a request to post my +- runs translations of Hardball Times' zone rating data for players to date. Here it is:

All stats current through yesterday, 25 September.

I'll update it next week once the final data are in, but I thought folks might like to have access to this spreadsheet. If anyone wants the original excel doc, just shoot me an e-mail.

My methods for these translations are described in detail here. Conversions from plays to runs are based on Chris Dial's runs/play values in this post.

Have fun!

Update: John Dewan's Stat of the Week today released the top 5 and bottom 5 defensive shortstops in baseball using 2007 data. Here they are, as well as how they stack up to the above fielding ratings (remember, Dewan's stats are reported in +-plays, while THT's data are translated to +-runs...that's why Dewan's numbers look more extreme):

Name
Dewan+/-
THT+/- (rank)
Tulowitzki
+34
+28 (1st)
McDonald
+26
+18 (3rd)
Vizquel
+19
+14 (4th)
Bartlett
+19
+6 (10th)
Pena
+18
+23 (2nd)
Eckstein
-15
-7 (12th from last)
Young
-15
-14 (6th from last)
Harris
-19
-1 (~average)
Ramirez
-35
-22 (2nd to last)
Jeter
-35
-27 (last)
Not perfect agreement, but pretty darn consistent. Correlation coefficient = 0.95, which makes me plenty happy with the match. Biggest discrepancies are with Brendan Harris and Jason Bartlett.

2 comments:

  1. Awesome, Justin! Thanks for putting this stuff out there.

    Have you ever thought about looking at how teams leverage their fielders? Say, for instance, that you have an awful fielding left fiedler, but you have a staff that doesn't allow a lot of balls in play to left (at least when compared to average) -- or something along those lines.

    I'm not sure how you'd do it or if it's even interesting, but I was thinking about it the other day. Maybe you could look at balls in zone for an individual player (or position) per overall balls in play or something.

    I don't know. Just throwing something out there. Anyway, great job.

    ReplyDelete
  2. That's a really neat idea. Might be hard to answer every question, but it would be interesting to do as you say and just quantify how teams vary in BIZ to different positions. It'd be especially neat that could be correlated to something predictable, like strikeout rate. Or if there was some year-to-year correlation going on (though that could potentially be park effects too).

    Another approach would be to look at weighted frequency of innings by left and right handed pitchers, and compare that to BIZ distributions. You'd expect that if opposing teams were leveraging platoon splits, an all right handed pitching staff would have more balls hit to right field than a pitching staff that is 50/50 right/left.

    Anyway, cool idea! Hope I can get to it relatively soon!
    -j

    ReplyDelete