Milledge's Defense

One of the most important issues for this season in my mind is the question of Lastings Milledge, centerfielder. I know some people are worried about his bat too, but that's a post for another time. For now, let's assume he's going to be at least a solid 15-20 HR, .270 BA type guy for the next 7-8 years or so.

Therein lies the problem. Those numbers just aren't good enough to win with at a corner outfielder spot unless you're getting fantastic offense from premium defensive positions like catcher, shortstop, and second base. So if Lastings can't play center, he becomes either a good 4th outfielder or a real offensive liability in left or right.

So this year we've force fed Lastings into center in the hope that he can improve is jumps and reads off the bat well enough to be at least an average centerfielder. He certainly would appear at least to be athletic enough. Based on my observations watching as a fan, I would say he seems pretty awful--lots of bad jumps, more mental mistakes than you should have, etc. But I thought now would be a good time to look at his defensive statistics and see if we can be a little more analytical and less anecdotal in our evaluation.

Of course, defensive stats are generally pretty unsatisfying. The most basic defensive stats that we all learned to track as kids keeping score at the ballpark, errors, assists, and putouts, really don't tell you much. Errors are very subjective, as any casual fan knows from seeing official scorers post arguable Es or Hs. Worse, errors tell you nothing about range. Putouts and assists begin to get at range, but so much is dependent on the pitching staff--an infielder behind a staff with a lot of strikeout pitchers will never do as well as an infielder behind a staff of sinkerballers, for instance.

But there are a lot of different metrics out there, some pretty complicated, none really perfect. So let's dig in on a bunch of them and see how Milledge compares to other centerfielders. I'll go from the most basic to more complicated. Throughout, I'll refer only to "qualified" centerfielders, meaning those who have played the position in at least 2/3s of their team's games at the position.

• Fielding percentage: This is just the very basic and familiar stat that attempts to show how frequently a player makes plays that "should" be made. It divides putouts + assists by putouts + assists + errors. Milledge's fielding percentage is .987, which ties him for 12th out of 19 qualified CFs.

• Range factor: RF simply adds up the number of putouts and assists a player makes by the number of innings played. This stat is barely an improvement over just looking at the raw assists and putouts by only factoring in the amount of time a player played in the field. Milledge's RF is 2.60, tied with Chris Young and Melky Cabrera for 12th out of 19.

• Zone rating: STATS, inc. invented this stat to try to improve on the more traditional stats discussed so far. The idea is that they divide the field up into zones assigned to each player on the diamond and then simply calculate the number of plays a fielder makes. The system even factors in a narrower range for hard-hit liners compared to lazy pop-ups for instance. One of the problems with ZR is that it is subject to human judgment calls, just like errors. For instance, the fuzzy line where a liner becomes a flyball is determined by a human being watching the game. Another problem is that ZR doesn't account for defensive shifts; when Ryan Howard comes to the plate and Manny positions the shortstop to the right of second base, ZR holds him accountable for covering the same zone as if he was positioned at his normal spot on the diamond. Milledge's ZR is .872, tied with Michael Bourn for 15th out of 19.

• Revised zone rating: This stat, which can only be found on The Hardball Times website, is the same as zone rating but attempts to correct for some of the flaws in the original ZR system by taking "out of zone" plays out of the equation. Milledge's RZR is .884, 18th out of 20 qualified CFs (for some reason I can't figure out Mike Cameron qualifies under Hardball Times' filter but not ESPN's).

• Out of zone plays: This is a corrollary stat to RZR; once you take out of zone putouts and assists out of the formula, you have to credit fielders with those plays with the simple counting stat, OOZ. Milledge has made 44 out of zone plays, ranking 15th of 20 qualified CF.

• Rate2: This is a Baseball Prospectus stat that factors in a bunch of things that aren't really explained on the site and then expresses it as total runs prevented or additional runs allowed per 100 games. The arbitrary baseline is 100 runs; a player with a 110 Rate2 stat prevented 10 runs that would have scored with an average defender. Milledge's rate2 number is 86, 19th out of 20.

• Win Shares: Another complicated stat that isn't explained in any of the sites or books I've read. Bill James invented this to try to create a consistent measure of how many wins a player contributes to his team over time. It attempts to, for instance, measure the relative value of players who play totally different positions. Players' contributions are measured for hitting, pitching, and fielding, and so by just looking at the fielding number we can get another super-sophisticated rating that we don't understand. Milledge's fielding win share number is 2.0, 18th of 20.

There are other really good defensive rating systems that either aren't available for the 2008 season, like David Pinto's Probabilistic Model of Range, which is posted at Baseball Musings. Hopefully David will have these numbers up again by the holidays.

So what does all this tell us? Well, it appears that Lastings is solidly in the very bottom tier of the worst centerfielders in baseball. He isn't in the top half for any stat, and he's right near the bottom for some of the more credible, sophisticated stats. The only other CFs who came up again and again at the bottom of these ratings were Nate McLouth and Joey Gathright. Most other players tend to bounce around the rankings and pop up near the top for at least some stat.

OK, so that's a ton of links and sophisticated stats to tell you something you probably already knew. The next question is whether he's getting better or not. Unfortunately, none of these stats as far as I can tell are available for partial seasons. We could try to compare his 2008 numbers to 2007 to see if he's generally becoming a better fielder, but because he played so little CF last year, there's a real apples to oranges factor in play. Moreover, he just didn't play that much at all, so there's big sample size problem as well.

So we're left with just watching with our own lying eyes to see if he seems to be getting better. The good news in all this is that the cheaper your seats, the better your view of outfield defense play. So next time maybe I'll skip all the links and just get me a $5 seat.