Aaron Hill’s Polarizing Defense

The way in which people evaluate defense fascinates me. Not because I find defense fascinating, per se, but because it strikes me as a very different evaluation process than the one most people use when evaluating offensive performance. I think this applies to both casual and die-hard fans, just to varying degrees.

Offensive evaluation is largely based on basic accounting. As a fan, analyst, or someone directly connected to the game, you weigh a player’s offensive actions against one another to develop an idea of their performance. Some people might choose to use a linear weights structure, or some might think only in RBI or RISP, but the methodology is largely consistent for each person. Each plate appearance is weighed based on its importance and not on the outcome.

Some will choose to treat each PA pretty equally, some will choose to place a lot of weight on certain moments, but it’s a predictable process. When evaluating defense, I believe things get much less clear.

I think this is because such a large percentage of defensive plays are routine while a relatively small percentage of plate appearances end in the batter reaching base. As a result, our brains have evolved to measure defenders based upon a small subset of plays that our brains can recall.

It wouldn’t do us much good to distinguish between defenders based on plays made per opportunity because the differences among the majority of players at each position is quite small. Can you really separate a player who made 81% of his plays versus one who made 82%? Maybe, but we haven’t been trained to do that.

Instead, our defensive evaluations typically rely on how many great or terrible defensive plays we can recall a player making. Several terrific catches lead you to think a player is a good defender and several bone-headed plays lead you to think the opposite. We don’t often think about the balance of those two things and what it means for our measurement of defense.

Imagine a player who is asked to field 100 balls in a season. Imagine 85 of them are routine and 15 are difficult. If he makes 80 plays (75 routine, 5 difficult), we might consider him average. But the problem with our brains is that we’re going to remember that 10 easy plays he didn’t convert and the 5 difficult plays that he did. We only take notice when the player does something that runs against our expectations of the play’s difficulty.

Instead, imagine a player who makes 80 plays, but 80 are routine and zero are difficult. Who is the better defender? I don’t think there is an obvious answer, but there is a clear trade-off between a player who makes more routine plays and one who makes more difficult plays. Each play is relevant to evaluating a player, but we don’t treat them equally when making our own judgements. This is why I’m a fan of defensive metrics. The metrics are trying to avoid this memory problem in which we only remember the plays at the end of the spectrum.

I’d like to make use of some Inside Edge data and Diamondbacks second baseman Aaron Hill to illustrate this. Earlier in his career, Hill rated as a terrific defender using UZR and DRS. The metrics disagree a bit, but we’re talking about 10-20 runs above average at times. Over the last couple of years, Hill has been below average, and perhaps even as many as 10 runs below average.

We don’t have to argue about the minor details, but the metrics say he used to be great and his skills are eroding into his thirties. This isn’t surprising.

What is surprising is that Hill made the most remote plays (1-10% chance) of any non-catcher in 2014 according to Inside Edge. Hill made six remote plays at second and one at third, but for our purposes let’s focus on the six at second because we don’t have a long history of data for him at the hot corner.

Aaron Hill made six remote plays in 2014. No non-catcher even made five at one position and only Jean Segura, Josh Donaldson, and Billy Hamilton even made four. Only Alex Gordon, Nolan Arenado, and Andrelton Simmons made three.

We should probably acknowledge measurement error within the Inside Edge data and the pure randomness of a single season, but that’s a group of good defenders and Aaron Hill (with some disagreement about Segura). Those are players you might expect to see on this list, but it also includes a 32 year old whose abilities seem to be in decline.

So the question shifts to the way in which Hill made these six plays and if it can shine a little light on how we evaluate defense? No one made more great plays, as defined by Inside Edge than Aaron Hill. Does that make him great?

April 21

hill1a

May 12

hill2a

June 16
hill4a

August 27
hill6a

You’ve just watched four of Hill’s six best plays to get a sense of what he was able to do. If you casually watched the D-backs in 2014 or followed the NL West in general, you might have a little better memory of Hill’s season, but if you follow another team and division, you probably didn’t see anything except the highlight reel plays.

Now I don’t mean to suggest that the average fan thinks Hill is the best second baseman in baseball simply because he made a handful of really tough plays, but I do think he’s a good example of how great plays can be deceiving.

Defense is difficult to measure and difficult to analyze because we’re dealing with a small sample of plays each season that are anything but routine. Hill seemed to excel at the “step and a dive” reaction, but that will only get you so far at second base. A few great plays don’t offset an otherwise mediocre player but a few great plays might lead an observer to think highly of a player’s ability without recalling the dozen chances in which the same fielder didn’t quite get to a batted ball.

In one way, Aaron Hill was the best defender in baseball last year. But in another, much more accurate way, he wasn’t anywhere close to the best defender. By the sum of his entire performance, he was below average but he made extremely difficult plays more often than anyone else.

This folds neatly into that ceiling versus probability debates we have about young players all the time. What a player is capable of and what they actually produce can be quite different, and Aaron Hill’s 2014 is certainly proof of that. He was a below average defender who gave you occasional glimpses of the great glove man he once was.





Neil Weinberg is the Site Educator at FanGraphs and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.

11 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
David Pinto
9 years ago

Hill was positioned well on all those plays, so as he ages he’s trying to make up for a loss of range (the ability to move a long distance) by better positioning and quick reactions.

Roger
9 years ago
Reply to  David Pinto

Yeah, I think “positioning” is pretty much the one-word answer to all the conundrums (conundra?) raised here — and this jibes pretty well with my eyeball impression of Hill as a fielder too.