Moneyball: The Sequel?

The A's announced yesterday that they had signed Coco Crisp to a 1 year contract for $4.5 M with an option for more in 2011 with a $500 K buyout, so he's guaranteed $5 M in the deal. Coco Crisp? Has Billy Beane lost his mind? This deal certainly has a lot of people scratching their heads, especially since the A's already had 3 OF's and had just acquired Michael Taylor, who appears close to being MLB ready. I don't think Billy Beane has lost it, but to understand this move, we have to understand the evolution of Sabermetrics, the study of baseball statistics.

There are rumors of an upcoming movie based on the book Moneyball by Michael Lewis about how the Oakland A's, led by their GM Billy Beane, approached the 2002 Baseball Free Agent Draft. That book glorified two baseball statistics, OBP and OPS(OBP + SLG%). The gist of it was that it mattered less what "tools", or raw ability, a player possessed than how they performed on the field. It also was about how traditional statistics, such as Batting Average, undervalue certain players, thus creating opportunities for astute GM's, such as Billy Beane, to acquire such players at a relatively lower price, thus enabling low-revenue teams like the A's to compete with richer teams like the LA Angels, for instance. Moneyball proved to be a wildly popular book, in part because it meant a whole bunch of ordinary fans could now use statistics to do their own evaluation of players, and imagine themselves to be superior at player evaluation than their favorite team's GM. It also was a boon to fantasy baseball enthusiasts who believed they now had a leg up on competitors who were less sophisticated about these matters. The book also created a near deification of Billy Beane and the OBP and OPS statistics among a large set of serious baseball fans. BTW, this was not exactly a new concept in baseball. Many years before Moneyball, Earl Weaver, a very smart and successful manager for the Baltimore Orioles said, "baseball games are won by 3 run homers and great pitching", or something to that effect. Well, Michael Lewis might want to get started on Moneyball: The Sequel. Times have changed since 2002!

One major piece missing from Sabermetrics in 2002, and thus from Moneyball, was a good way to measure baseball defense. The practical result was that defense became almost ignored by a new generation of GM's, led by Billy Beane. The corollary was that Beane acquired a whole bunch of big, slow players with big butts who drew a lot of walks, hit the ball hard when they hit it, and who looked more like statues in the field than ballplayers. As time went on, a few keen observers started to realize two things: 1. The A's were getting worse instead of better. 2. Billy Beane was starting to deviate from the script and acquire a more traditional type of player. What few people realized was that rather than abandoning the principles of "Moneyball", Billy Beane had found a new statistic that he quietly started using to evaluate players, UZR or something similar.

Baseball analysts have long known that Errors and Fielding Percentage are terrible ways to evaluate defense. The problem, aside from the inherent subjectivity of judging just what an error is, was that errors are generally only charged on balls the fielder actually touches. A defensive player could stand stock still in the field and let a ball fall on the ground next to him, and it would be counted as a hit rather than an error. But how do you measure what balls a player should be able to make a play on? Modern computers now enable us to track and keep a record of all batted balls hit into the general area that a defensive player is guarding. They can also track which of those batted balls the fielder successfully makes a play on. This data is then processed into a formula that includes errors, successful throws, balls cut off that hold a batter to a single rather than a double or triple, and other factors to produce a defensive statistic called UZR or Ultimate Zone Rating. Raw UZR can then be normalized to 150 games, a typical full season, or UZR/150. UZR purports to be a measure of how many runs a particular fielder saved or cost his team relative to an average fielder. If UZR is to be believed, defense has a far greater impact on the game than was previously believed, at least by "Moneyball" enthusiasts. For instance, Adam Dunn is a Sabermetrician's dream. A great big guy who doesn't look much like a ballplayer, he tends to have low Batting Averages, but draws a lot of walks and hits a lot of HR's producing spectacular OBP's and SLG %'s and thus excellent OPS's. Now, everybody knew that Adam Dunn is a terrible fielder, but nobody really cared because there was no good way to measure it, so nobody knew just how bad he was, and he was a prodigious run producer. Well, according to UZR, Adam Dunn cost his team 47 runs on defense last year, largely negating all those runs he produced on offense!

So, what does this all have to do with the A's signing Coco Crisp? Crisp was injured most of last year, and the A's already had 3 OF's, Scott Hairston, Rajai Davis, and Ryan Sweeney. All 3 are at least as good offensively as Crisp, and possibly better. Hairston hits more HR's, Davis steals a lot more bases and Sweeney projects as a much better hitter overall. What's more, none of these players are perceived as being particularly bad fielders. Lets' look at the data:

Rajai Davis: 2009 UZR +17.9, Career UZR +12.

Ryan Sweeney: 2009 UZR +27.6!, Career UZR +17.1.

Scott Hairston: 2009 UZR +1.8!, Career UZR +6.3, but Hairston plays almost exclusively LF, so his data is based on a comparison with average LF's, which includes a lot of players like Adam Dunn!

Coco Crisp: Career CF UZR + 5.8, Career OF UZR 10.9.

Guess who the odd man out is going to be? The final outcome of all this remains to be seen. It is quite possible that Billy Beane will make more trades before the offseason is over. Most observers are now predicting that if anyone is dealt, it will be Scott Hairston, or else Hairston will become the 4'th OF! It would appear that Billy Beane is going to start the season with 3 CF's playing the 3 OF positions. He will try to win with pitching, speed and defense, the exact opposite of "Moneball"! Instead of basing this approach on an old saw passed down from one generation of baseball managers to another, it will be based on science, Sabermetrics, the study of baseball statistics! Moneyball: The Sequel, if you will!