22 April 2014

The Orioles Have an Average Defense

There has been a storyline that has been trumpeted a bit on air and in the media that has not actually been true. You often hear about the club being proficient at fielding and this centering on their ability to avoid errors. At the time of this writing (after a particularly gruesome Sunday), the Orioles had the second-fewest errors in baseball. Only the Houston Astros had less.

On the surface, this looks great. On the surface, our old friend Henry Chadwick (the guy who invented the box score a century and a half ago) would say that errors do not give you the entire story of a player’s or team’s defense.

The problem with only looking at errors, or even the slightly more sophisticated fielding percentage, is that you are only considering batted balls that the player can actually get to. In a previous article, I mentioned the fallacy here as being the Rubbermaid paradox. To restate, if you put a Rubbermaid trashcan out on the field, that trashcan would have no errors and would have cleanly caught a few pop ups giving it a wonderful 1.000 fielding percentage. The paradox here is in the perception. Someone who only considers fielding percentage and errors would be under the impression that the trashcan is a Hall of Fame worthy defender while in actuality, it would be the worst defender that baseball has ever seen. This awareness of the deficiencies in traditional statistics plants a recognized desire to find better quantitative ways to evaluate defense.

Last year saw the Orioles have the fewest errors in baseball history with 54. The team also had the highest fielding percentage at .991. That led people to say that the Orioles had an elite defense. They were right, but the numbers that provided that foundation were not the best ones to use. A better argument would have leaned on more advanced statistics like Ultimate Zone Rating (UZR), which can be found at FanGraphs. How does UZR work? The field is cut up into 64 squares and each batted ball is defined by which square it lands into as well as whether it was a ground ball, line drive, or fly ball. How these scenarios play out provides a baseline or average defensive expectation for each fielder along with the number of runs expected when a certain balls lands a certain way in a certain square.

In other words, Manny Machado plays third base. A number of batted balls come through the zones he covers on the field. MLB as a whole has a certain probability of making plays on each of those balls hit. You then compare Machado’s performance against what the average major leaguer does. For the balls that skate through the hot corner, we know from the MLB database how many runs wound up scoring on average. We then can associate a run value lost for each play Machado misses as well as a run value gained for each he makes. This tells us how many runs saved or lost his performance last season was worth. With respect to 2013, Machado was first with 33.6 runs saved over what an average third baseman would have been expected to save.

How did the team stats look last year? The Orioles ranked third (48.6 runs saved) and were outperformed by only the Diamondbacks (49.7 runs saved) and the Royals (87.6 runs saved). The team also performed well in UZR’s throwing arm (eighth), double play (fifth), range (12th), and error (eighth) metrics. You perform that well on all of those individual measures, you wind up with an elite defensive club.

However, this year’s club currently sits 15th in all of MLB in UZR. They are showing well for throwing arm (eighth), double play (seventh) and error (fifth) metrics, but are 23rd for range. The main problems have been the continued erosion of Nick Markakis’ range, Nelson Cruz’s expected poor coverage, and, surprisingly, J.J. Hardy’s performance. Of those, we should only expect Hardy to get better due to his performances in years past and the likelihood that his range troubles are related to his nagging injuries. In general, things should improve when Machado returns as well. The current Orioles third basemen, Jonathan Schoop and Ryan Flaherty, are performing on a projected value of minus-18 runs or about 50 runs worse than Machado over a full year.

For the Orioles to compete, the defense needs to be elite. Right now, it is not.

No comments: