As we discussed this afternoon, Scott Boras has taken to comparing his star free agent of this winter to his star free agent of last winter, as he aims to get Matt Holliday a contract similar to what the Yankees gave Mark Teixeira a year ago. In the comments, the question was raised why I placed Holliday in a lower tier than Teixeira, despite the fact that Holliday has the superior WAR rating.
Before we get into the specifics of Holliday versus Teixeira, let me say something about WAR. Obviously, we’re big proponents of it has an evaluation tool. We think it’s the best single value metric out there for players. We do not think its perfect and infallible, however. There are portions of the game that are currently not measured (non-steal base running, catcher defense, and league differences being the most notable), and we’ve discussed the limitations of UZR ad nauseam (I believe that Teixeira is probably a bit better than his UZR has shown). There is room for discussion when two players are within the same general range.
As it pertains to Holliday and Teixeira, the key here is league differences. For various reasons, we don’t incorporate those into the WAR ratings, but they have been fairly significant for the last five or six years. The AL is just a better group of talent than the NL, and as such, when two players who post equal numbers in the different leagues, the AL player has to be presumed to be better. He’s facing better competition and creating the same results, which is inherently more valuable.
The best estimates for the differences between the AL and NL range between +0.25 and +0.5 wins per season. When comparing players across leagues, this is not an insignificant factor. That Holliday has had the bulk of his success come in the NL requires a downward adjustment to the value of that performance. Likewise, Teixeira’s ability to perform in the AL enhances his value.
Now, that doesn’t mean that Holliday’s going to turn into a pumpkin upon arrival in the AL. After all, he gave the A’s +3 wins of value in 400 PA, which was still quite valauble. However, we just can’t translate his Colorado and St. Louis performances over to the AL without an adjustment, and those are the places where Holliday has performed like a superstar.
Even after the league adjustment, though, Holliday and Teixeira come out as similar in value. They’re essentially the same age. Why would I prefer Teixeira going forward?
All types of past performance are not equally predictive of future performance. One of the driving forces of Holliday’s offensive value is his strong batting average on balls in play. Over the last three years, his .363 BABIP is tied with Chone Figgins for the third highest in baseball, behind only Ichiro Suzuki and Matt Kemp.
BABIP is simply more subject to variance than Teixeira’s preferred method of production – hitting the ball really far. Even though Holliday has been able to sustain a high BABIP, we still have to regress his projcted BABIP further towards the mean than we do Teixeira’s ISO, as the latter has simply shown to be more stable historically. Even though their results over the last few years have produced similar value, Teixeira’s taking the safer path to those results. Holliday’s dependence on a high BABIP for his offensive value makes him a greater risk, and increases in risk drive down value.
I expect that we’ll see these concerns manifest themselves in the marketplace as well. My guess is that Holliday will end up signing for 10 to 15 percent less than what Teixeira got, because teams are also a bit skeptical of unadjusted National League numbers and don’t see him as the same kind of power hitter.
Holliday is a very good player, and certainly the best free agent on the market. But when comparing him to Teixeira, it’s important to not stop at a simple WAR analysis.