fear of flies

the guys over at USS Mariner and Lookout Landing ran a little outfield-defense study the last couple of days, tag-team style; i learned about it over at Tango's blog. the USSM crew counted up the number of putouts by mariner outfielders, year by year, during the 2000s; Lookout Landing counted up the number of flyballs in play year by year going back to 2003, then plugged in the putout data to determined the percentage of flyballs that were turned into outs. not very complicated, but somewhat revealing: the Ms' conversion rate on flyballs has dropped like a stone over the last 5 years.

the cards' terrible outfield defense was one of the biggest (but least recognized) stories of 2007; see the little table at the bottom of this post for a sense of how inept the flycatchers were. even edmonds had a bad year with the glove; john dewan's plus/minus system listed jed as the second-worst centerfielder in the nl. the USSM/LL study was easy to replicate, so i did it using slightly different methodology. Lookout Landing got its ball-in-play data from retrosheet, using a perl search script; i (being lazier and bereft of programming skills) relied on Baseball Info Solutions' gb/fb/ld breakdowns, which are easily accessible over at fangraphs. the putout figures come straight from Baseball Reference and are pretty standard; even sabr-haters consider putouts a legitimate stat. i only went back 3 years, also on account of laziness --- but 2005 gets us back to the days when the outfield defense was (whether judged via the stat sheet or the naked eye) pretty damn good. compare:

zounds.

two years ago they caught 78 percent of the flyballs in play; last year, only 67 percent. that's a stunning dropoff. but it's believable when you consider how the personnel have changed. the 4 outfielders with the most innings afield in 2005 were edmonds (who won a gold glove, his last, that year), larry walker (a 7-time gold glover), reggie sanders, and so taguchi; last year they were the post-foot-surgery edmonds, chris duncan, juan encarnacion, and ryan ludwick. of the latter 4, only ludwick scored above average on the metrics last year.

how many runs did this cost? let's just rough it out --- nothing too fancy here, and no claims of precision. first of all, let's tally the number of flyballs that would have been caught by a competent defense. obviously the 2005 outfield was far more than competent; it's not realistic to use a 78 percent conversion rate as our standard. i could just split the difference and use the 2006 conversion rate; that's the median in our little sample. here, let's try that out:

that's 106 flyballs the 2007 outfield missed that the 2006 outfield theoretically would have caught. even if we assume that every one of the missed flyballs became only a single --- and surely many of them would become doubles, and a handful would be triples --- the cost of those 106 uncaught balls would be nearly 80 runs. if we factor in extra-base hits, the cost inflates to well over 100 runs. is that plausible? maybe. don't forget, the cards' run yield in 2007 (829 runs) was the 3d worst in franchise history --- and we can't blame the inflation on a high homerun total (the cards were average in that regard last year, 8th in the nl) nor a high walk total (the cards allowed the nl's 5th-lowest total). that leaves only one way to explain the increase in runs --- a hell of a lot of balls in play didn't get turned into outs.

now, was the failure to catch all those balls entirely the fault of the fielders? surely not. first of all, as the first table above shows, the outfielders had more flyballs to chase last year than in either of the two previous years --- an increase of 200 over 2006 and nearly 400 over 2005. here's the percentage breakdown by ball-in-play type:

if the pitcher throws a meatball and the hitter scorches a line drive into gap, you can't blame the fielder for failing to run over and catch it. st louis pitchers threw plenty of meatballs last year; we all saw it, and the increased flyball rate indirectly confirms it. so i wouldn't blame the fielders entirely --- but i wouldn't blame the pitchers entirely either. the pitching staff's aggregate rate of line drives allowed was lower last year (19.6 percent) than in either 2006 (20.4 percent) or 2005 (20.3 percent), which suggests that the spike in runs allowed isn't entirely related to lousy pitching. the moundsmen got hit harder last year, but not that much harder --- if they were really that terrible, we'd expect a high line-drive rate and a high home run rate, and last year's staff didn't exhibit either of those symptoms.

it stands to reason that a least some of the balls that weren't caught could and should have been caught. how many? let's make a modest assumptions. suppose the cards allow flyballs at the same rate as last year, but the defenders catch 70 percent of the flyballs instead of merely 67. how many runs would that save? let's see:

a 70 percent conversion rate would eliminate 53 singles, doubles, and triples and turn them into outs. that's a savings of about 50 runs --- 5 wins in the standings. was the outfield defense really so bad that it cost the team 5 or more wins last year? no doubt about it.

so how likely is it that they can increase their conversion rate accordingly? that really depends on who plays. if juan gone makes the team and gets innings out there, the percentage of balls caught might well go down. we probably shouldn't expect much improvement from duncan; he won't be playing with a sports hernia this year, which might help a bit, but he'll always just be a first baseman doing his best to survive out there. ankiel is still a mystery; he did ok last year but is still learning, and he's better in center field than in right. ludwick's good, but he already played more than half the time last year and isn't likely to get substantially more time this season.

the only guy likely to make a real impact, imho, is colby rasmus. by all accounts he's a plus defender; flank him with ludwick and ankiel and you'd have a pretty tight outfield D.

but don't take that to mean that i'm in favor of making rasmus the opening-day CF. i'm still leaning against that idea because i think a premature promotion might set the kid back. but i could be persuaded otherwise if he has a good spring and exhibits an ability to recognize pitches, make adjustments, take the ball the other way when that's what the pitcher gives him. given the likely benefits of his glove, he'd probably only have to hit .250 / .325 / .425 or thereabouts to make a big impact.