Two interesting sabermetric tidbits that I found share-worthy came up over the last few days:
The fantastic Marc Carig of the Star-Ledger is now writing for Baseball Prospectus, and his first column made an excellent point about defensive metrics:
Which brings me to the idea of defensive metrics. Created in response to the obvious shortcomings inherent in relying on errors and fielding percentage, the array of defensive metrics publicly available have changed the way we view the game, mostly for the better. At the very least, defensive metrics have offered a method to challenge long-held beliefs about players, whose evaluations are too often based upon reputation—a fact we’re reminded of when it’s time to hand out Gold Gloves. With metrics, we can Madden-ize the complicated world of defense on a baseball diamond, or at least try.
Yet, given the broad error bars that surround some metrics, at times it might be more beneficial to take a step backward from our ranking ways, particularly when it comes to defensive metrics. Instead of getting caught up in the degree to which one player may be better than another—no different than assigning letter grades in class—perhaps the metrics may be better used simply determining which players are good and which are not. Ultimately, isn’t this the most important question to answer when it comes to defense? Hot or Not? Pass or Fail? Utley-esque or Uggla-esque?
[...]
For instance, it’s clear that from 2007 to 2010, Chase Utley has been an excellent second baseman while Dan Uggla has been awful. The two represent the extremes at their respective position when it comes to UZR. In the middle, where the ratings are bunched closer together, I’m not sure that the metrics possess the precision to make a distinction between the Aaron Hills and Robinson Canos of the world.
Check out the full article for some quotes from a player on this issue. I think Marc is spot on when he says metrics do not have the sort of precision necessary to truly rank players by their defense. The best we can do at this point is separate players into general categories (clearly great, solid, weak, and clearly bad?) by using all of the metrics and scouting reports at our disposal to try and evaluate a player’s glovework.
Dave Pinto notes that BABIP is down significantly across the league (the last value is 2011):
This could just be a fluke, and it may correct itself over the remainder of the season. However, if it does not, it is important to note that this creates a new baseline for evaluating BABIP and luck on an individual or team basis. If run conditions have changed so as to suppress BABIP across the league, that needs to be factored into any analysis that uses BABIP to call a player lucky or unlucky. While it is a bit early to draw conclusions from this data, I am definitely going to keep my eye on it.
If you enjoyed this article, please consider sharing it!
5 Responses to Saber Stuff: On Defensive Metrics, And BABIP
Leave a Reply
TIQ IQ
While it is a bit early to draw conclusions from this data, I am definitely going to keep my eye on it.
So what you’re saying is that time will tell what this means, right?
Lol. Yes, you could say that.
The Yankees are making Kyle Davies look like a major league pitcher!
[...] Saber Stuff: On Defensive Metrics, And BABIP – The Yankee Analysts [...]
Correct me if I’m wrong, but aren’t defensive metrics evaluated by “eye”? doesn’t a person evaluate whether or not a fielder should have made a play? While a trained person could make a more accurate assessment than a casual observer, this still makes these a bit subjective in my mind, but better than straight fielding perspective.