Sunday, March 27, 2005

On Adrian Gonzalez and spring training stats 

Over at U.S.S. Mariner, there is a great discussion on spring training statistics. In particular, the fact that spring training statistics are largely irrelevant, but spring training performance is not, and the key in evaluating players in the spring is differentiating between the two.

One of the thing that any stathead will tell you is that the reliability of statistics is directly related to sample size...the more data you have to work with, both from the player you are looking at and his peers, the better you can evaluate his performance based on the numbers. The less data there is however, the more important observation and other subjective measures become in evaluating a player. For example, if I were given one game in a vacuum, and had to decide who the best player on a team was based on that one game, I'd be more inclined to trust my eyes than the box score, since anyone can go hitless or have a 3 for 4 game. But as the sample sizes increase, the objective data increases in reliability, making it less necessary to rely so heavily on the subjective.

Which gets us to spring training, where there are two data problems. First, you are talking about only a handful of ABs...for example, the other day, someone was lamenting that they were hoping that Gerald Laird would come into spring training and be hot with the bat, forcing the Rangers to give him the starting catcher job, but that a .250 average wasn't going to get it done. The problem is, though, that at the time, Laird was 7 for 28, barely a week's worth of ABs over the course of a season. Two more hits, and he's got a .321 average, and everyone is talking about how well he's hitting this spring. Anyone who has seen "Bull Durham" remembers Kevin Costner's drunken diatribe, about how one hit a week -- "a bleeder, a gork, a groundball with eyes" -- is the difference between being a major league regular and a bush leaguer. What he says is absolutely true, and highlights the sample size problem with baseball, since difference in a couple of extra bleeders or gorks or groundballs with eyes in the spring is the difference between a .250 hitter and a .320 hitter, and can be the difference between a player heading North and getting shipped out if a manager or G.M. gets to fixated on that data.

Along with the sample size problem, you also have players facing a huge variety of players in the spring...one candidate for a job could be facing mostly major league pitchers in his outings, while another is facing minor leaguers, and a couple of #3 starters who were throwing only breaking pitches to try to get ready for the start of the season. This makes comparing statistics from player to player in the spring much more problematic.

What Derek Zumsteg and Dave Cameron point out, though, in the U.S.S. Mariner piece, is that while spring training statistics shouldn't carry a lot of weight -- particularly in comparison to historic performance -- spring training performance is worth keeping an eye on, since a player who has spent the winter developing a new pitch, or who has gained some strength and power, or who, conversely, has lost a step, can reasonably be expected to have a larger deviation from their historic performance than would be expected. The key, of course, is differentiating between who has legitimately improved their performance, and who is simply on a hot streak.

Where this really becomes relevant for the Rangers is in regards to Adrian Gonzalez, whose torrid spring has the team considering him as their starting DH this season, rather than going with the planned Dellucci/Colbrunn platoon. The putrid performance of the D/C combo this spring, particularly in light of Colbrunn's lack of playing time the past couple of seasons, no doubt plays a role, as well, but Gonzalez has seized the opportunity and has put himself on the map with his play.

Now, I've spent the winter loudly disagreeing with those who have said that Gonzalez should be a candidate for the DH role in 2005, based on the fact that 1) his defense at 1B is one of his greatest attributes right now, and playing him at DH wastes that, and 2) he's not going to hit for enough power to be a productive DH. And at this point, Gonzalez's power, or lack thereof, is the biggest issue with him as a prospect, since how his power develops is going to be the largest factor in determining whether he's going to be the next Rafael Palmeiro (as his proponents suggest he can be) or the next Rico Brogna (as his detractors, including myself, suggest he'll become).

Scouts have felt that Gonzalez has the swing and the body where he can develop power, but it is one of the latest skills to blossom, and for some players, it simply never comes (see, e.g., Sean Burroughs). Gonzalez, however, has shown that missing power stroke this spring, and that is what has gotten him into contention to be the DH.

This is where the spring stats/spring performance differentiation becomes crucial...the fact that Gonzalez is slugging .678 or whatever it is right now is no reason to make him the DH this season. That is, in my mind, almost irrelevant. However, if Gonzalez has gotten bigger and stronger over the winter, if his approach at the plate has improved, if he has made real strides towards becoming the player that scouts thought he could become when he was the first overall pick of the draft, and if this represents a real, permanent improvement in his ability...if that is the case, then he should probably be the DH for the Rangers in 2005.

Again, though, the rub is in making the determination as to whether this is a real leap in ability, versus a hot streak...and that's the decision that Showalter, Jaramillo, and the rest need to make.

I enjoyed you blog about creative memory consultant. I also have a site about creative memory consultant which makes me appreciate this one even more! Keep up the good work!
Post a Comment

This page is powered by Blogger. Isn't yours?