david pinto has begun posting his 2005 defensive ratings at baseball musings. if you're not familiar with pinto's system -- which he has saddled with the ungainly name of "probabilistic model of range," or PMR -- there's a very good explanation of it here (scroll down past the table). it's based on batted-ball data, with every ball in play catalogued and weighted according to difficulty. as pinto explained the other day: "The idea is to look not just at the balls turned into outs, but how difficult those balls were to turn into outs. Teams or fielders who turn difficult plays into outs do well. Teams or fielders who let easy balls drop for hits (or make errors) do poorly."
the cardinals as a team did very well in 2005, according to pinto. by his data they turned about 100 hits into outs and thereby saved about 75 runs, very roughly speaking -- nearly half a run a game, and in the range of 8 to 10 games in standings. five other teams flashed even better leather, including the astros, who swiped about 110 hits and did it in far fewer balls in play (since opposing teams struck out so much against the astro pitching staff).
pinto began looking at individual positions last night, starting with the shortstops. per his method, david eckstein had the steady solid year ev'ybody was hoping for, ranking 16th among the nearly 40 mlb shortstops who played enough innings at the position (about 250) last year to qualify. that's a far sight better than a lot of us thought he would do when the cards signed him last year. he was slightly below average, according to pinto, but even so li'l eck had the best year with the glove among last year's large class of free-agent shortstops:
| outs below avg |
cost in runs |
cost in games |
|
|---|---|---|---|
| eckstein | 4 | 3 | 0.33 |
| vizquel | 9 | 7 | 0.75 |
| cabrera | 16 | 12 | 1.25 |
| guzman | 17 | 13 | 1.25 |
| renteria | 32 | 24 | 2.67 |
eckstein never shone with the glove, but he delivered on a promise: he made the plays he was supposed to make.
PMR is only one of many defensive measuring systems out there, and it's still being refined. it has the advantage of being based upon actual batted-ball data -- it measures the plays a fielder made against his opportunities to make plays. but since that information is proprietary, the rest of us can't play around with it to see if pinto's conclusions hold up.
what we can do, though, is see if his conclusions are consistent with those yielded by other systems. so here's another little table that gives ratings for the same five shortstops from half a dozen defensive systems. one of them is UZR, or ultime zone rating, developed by cardinals stat consultant mitchel lichtman (aka mgl) and widely considered the best, albeit far from perfect, instrument extant. UZR figures are cardinal property now and no longer publicly available, but mgl can't resist dropping little teasers over at baseball think factory from time to time. i was able to find his ratings for two of these five shortstops.
UZR too is based on batted-ball data and is in fact the mold from which pinto cast his PMR. the other comparison tools i'm looking at here are generated from less finely calibrated data, including zone rating (which is freely available at espn) and basic statistics like assists, putouts, and errors. they include
- Range, developed by david gassko and presented at hardball times. his figures are published in the 2006 THT annual.
- FRAA, or fielding runs above average, developed by clay davenport and the basic defensive measurement used by baseball prospectus
- chone smith's enhanced zone-rating figures, explained here.
- chris dial's similar zone-rating-plus figures, explained here.
| PMR | UZR | Range | FRAA | Chone ZR+ |
Dial ZR+ |
|
|---|---|---|---|---|---|---|
| eckstein | -3 | n/a | +7 | +1 | +5 | -3.6 |
| vizquel | -7 | +5 | +9 | +11 | +14 | +11 |
| cabrera | -12 | +14 | -2 | +6 | +5 | +7 |
| guzman | -13 | n/a | -9 | -10 | -8 | -14.5 |
| renteria | -24 | n/a | -11 | -21 | -15 | -6 |
with a few exceptions, these numbers are pretty consistent across the board. eckstein hovers right around average in all cases; renteria and guzman are uniformly well below average. while the results are slightly mixed on cabrera and vizquel, it's probably safe to assume that they're saving some runs with the glove.
more in this vein to come as pinto's PMR ratings appear for the other positions.