April 3, 2003
A Closer Look at the Twins' Outfield Defense
I got two big questions out of my recent article sketching outfield fielding as a percent of fly balls, and both were about the Twins.
"What about foul balls? Weren't the Twins cheating the corner OFs over and letting Hunter cover more ground so the corners could get to fouls?"
That's what I read the Angels said in the post-season, but if that was true, it didn't work--the Twins didn't get more putouts, and Torii Hunter in particular would have been expected to catch a lot more balls than he actually did.
"What about the crappy off-white spotty roof of the Hubert H. Humphrey Metrodome? Wouldn't it be possible for a team to play really good defense in such a hard environment and still look bad?"
Indeed, and that's something I can look at.
Let's check out the Twins' all-outfield putout rate on non-HR fly balls over the last 10 years. If HHH is indeed making it much more difficult for the outfield, we should consistently see them with a PO/FB-HR rate at the bottom of our anticipated range, around 65% rather than the 77% we saw the top teams of 2002 turned out.
Year non-HR FB OF PO PO% 1993 1573 1109 71% 1994 1175 867 74% 1995 1401 998 71% 1996 1723 1223 71% 1997 1564 1035 66% 1998 1711 1200 70% 1999 1643 1122 68% 2000 1649 1137 69% 2001 1700 1222 72% 2002 1745 1202 69%
Only once in the last 10 years did the Twins turn in a defensive year in the outfield that was above the 2002 average, though there are another four that would have been put them at about 2/3 of the way down the list. There are two ways to look at that: First and easiest, no big deal, the stadium doesn't have a huge effect. Or, the stadium has a modest effect that makes it really hard for these blue-collar, hard-working outfielders to even appear average.
Sigh. Always gotta be creating more work for me, don't you? I can't just go to toga parties like Jim Caple, no. I've got to figure out if that kind of statement's true or not.
It's not. I looked at the out rates of outfields led by the same players with other teams, and I didn't see it. Marty Cordova, unfortunately, goes from mostly a regular with the Twins to a half-time player with the Indians and Orioles, and his performance wasn't wildly out of line from his time with the Twins. Otis Nixon's 1998 putout rate in Minnesota was higher, though not significantly so, than what he did in Toronto and Boston, and a little worse than what he turned in back when he was in Atlanta, in 1992. Now, I haven't run historical G/F ratios on all the staffs these guys were behind in, but I'm not seeing any kind of depression of outfielder ability from the Metrodome.
So what's the scoop? Here are the breakdowns for every year from 1996 on, when baseballreference.com has by-position breakdowns:
non-HR Primary Primary Primary Year FB CF CF PO PO% RF RF PO PO% LF LF PO PO% 1996 1723 Becker 471 27% Lawton 381 22% Cordova 371 22% 1997 1564 Becker 434 28% Lawton 249 16% Cordova 352 23% 1998 1711 Nixon 436 25% Lawton 401 23% Cordova 363 21% 1999 1643 Hunter 469 29% Lawton 329 20% Allen 324 20% 2000 1649 Hunter 480 29% Lawton 342 21% Jones 315 19% 2001 1700 Hunter 503 30% Lawton 356 21% Jones 363 21% 2002 1745 Hunter 415 24% Mohr 397 23% Jones 390 22%
The thing I notice right off the bat is that Hunter looks more injured here than I'd suspected. There's an obvious and significant drop in his year-to-year stats, from being pretty good to suddenly being league-worst.
I got another letter that really got me thinking though, from my friend and strikethree.com smart guy David Cameron, which I'll paraphrase for publication:
"So the Twins weren't getting outs from their outfield, and we know that Rivas was bad defensively, so where did their good defensive efficiency at turning balls in play into outs come from? Corey Koskie must have been awesome."
Well, uh, that's a good question there, Dave. You punk. I could be cracking open a beer and watching Baseball Tonight, but nooooooo.
Looking at the fielding stats in the 2003 Baseball Prospectus for last year's Twins, we've got
C A.J. Pierzynski at +4 1B Doug Mientkiewicz at +8 2B Luis Rivas at -10 SS Cristian Guzman at 0 3B Corey Koskie at +19 RF Dustan Mohr at +3 CF Torii Hunter at +5 LF Jacque Jones at +10
Which tells us a couple of things: Despite Dave's tone, Koskie really was awesome, and Rivas was bad. Taking just those Davenport Translation Fielding numbers, the Twins were +21 runs in the infield and +18 runs in the outfield.
Taking a quick look at the other league-worst outfields in the original article, the Phillies' primary OFs came out at -9 runs, the Athletics at 0 runs, the Astros at -2, Yankees at about +10 (I'm estimating White-Mondesi-etc). I'm surprised that the teams who had the worst rate of outfield putouts as a percentage of adjusted fly balls are turning up with such varied defensive ratings in the DTs, frankly.
So maybe raw outfield putouts as a percentage of team non-HR fly balls isn't anything beyond a sort of interesting exercise, though I may press on with related topics: Are outfield arms that important? Is there really a fly-ball distribution difference between infields and outfields, which would seem to go against current thought of batter-determined ball-in-play outcome?
Or is it, as readers suggested, an issue of foul ball territory and home-field adjustments? And most importantly, will I continue to look into this as part of my long-time obsessing over defensive statistics, or will I be distracted by some other shiny object?