Utley Still a Brand Name Vacuum

(Ed note! This piece is about a week old due to me being bad. Blame not the author. -Trev!)

Let's talk defense, the good and the bad, perceptions and reality. I'd like to cover a lot of current Phillies players, especially Ben Revere and Cody Asche. But here I will start with The Man. Chase Utley has had a long career as one of the best, if not the best, defensive 2B in MLB. Unfortunately for him, much of the professional media were blind to his virtues when he was in his prime. And since then his injuries and natural decline have obscured how good he was and made fans nervous that he will soon be no longer fit for 2B. Is this perception of Utley correct? Has he declined that much from his prime? Does he need to consider surrendering 2B to someone else for the sake of team? I'll say some things about these questions in what follows and incorporate into our conversations a new fielding metric on Fangraphs.

Defense remains the aspect of the game most obscure to rigorous measurement. But the measurements we have are better than anecdotal evidence alone and getting better as we learn from our mistakes. Generally, I approach the evaluation of defensive play by looking at all of the metrics available and trying to watch as many replays as possible to compare my own perceptions to the data. (I know; this is really controversial.) Recently, Fangraphs has provided an even better tool than the previous ones for analyzing how a defender gets his ultimate runs-rating. And this provision has made it even easier to form an informed opinion of a defender's play. The new data, called Inside Edge Fielding, breaks down plays into 6 categories (Impossible, Remote, Unlikely, Even, Likely, Routine) and shows for each defender how many opportunities they've had in each category and the percentage of conversion. The categories are divided according to the "normal" conversion rate of plays. Routine plays are normally converted at a rate between 90-100%, and most plays made are routine. Likely plays are normally converted at a rate between 60-90%, Even 40-60%, Unlikely 10-40%, Remote 1-10%, Impossible 0% (by definition no play made is impossible). Although this sounds rigorous and objective, plays are placed in these bins not based on precise measurements of the ball hit and the position of the player but by the eyes of human beings, specifically scouts. This data is subjective. Nevertheless, it is useful because the judgments are made by people who watch much more baseball than you or I (unless you're Cormican, Jay P, or mattwinks), and they are fine-grained judgments that break down the most important component of advanced metrics such as UZR or DRS, namely fielding range. In what follows, I'm going to use this new data to do just this with our favorite second baseman.

Chase Utley: for a few years now fans have wrung their hands over his degenerating knees. With respect to his fielding, fans have worried that a) Utley does not have the range to play 2B anymore and b) playing 2B takes a higher toll on his knees than other positions would. I don't know what to say about b). I wager fans tend to overestimate the differences in wear-and-tear between non-battery positions. But I have no evidence for this one way or the other. So, I'll let someone with more knowledge than me answer that worry.

Does Utley still have the range to play 2B? The data reveals that the answer varies from "yes" to "why would you even ask that?" If you look at Defensive Runs Saved (DRS, quick explanation, detailed), you will see that Utley was 4 runs below MLB average in 2013 and is currently 1 run below average on the season. That sounds not good. And to be sure Utley in his prime put up absurd DRS numbers, peaking at 30 runs above average in 2008. So, according to DRS Utley has fallen far from his defensive heights. But that does not entail that he is not a suitable 2B any longer. His 2013 DRS, -4, was only 2 runs below the league average and his 2014 DRS, -1, is the current league average. And even single season samples of defensive metrics are unstable. It is better to look at 3 seasons' worth. If we reach back to 2012, the first season after his knee problems arose, we see that Utley was 8 runs above average. Taking all of this together, we should conclude that, according to DRS Utley is still a league average 2B. By DRS, the answer to our question is "yes".

By Ultimate Zone Rating (UZR, explanation) the answer is "why would you even ask that?" By UZR Utley is on pace to be about 9 runs above average for 2014. Granted, in 2013 Utley was only +1.5, but in 2012 he was +6.8 over about half a season. We should have little doubt that according to UZR Utley is still an above average 2B. Combine that with continued excellent offensive production from 2B and we get the dumbfounded answer to the question. Utley has lost a step, but he had so many to begin with he could afford to lose one or two. (Utley's peak in UZR was 18.3. 18.3! ~2 wins!)

"But", the questioner objects, "DRS and UZR are so far apart on Utley. Why should we trust either one? What if they both overestimate Utley's defense? It looks to me like Utley can't get to enough balls anymore. I just don't see him do this anymore:"

Well, I could just tell you your eyes should not be trusted because:

But I could also tell you that we can trust other people's eyes. By using the Inside Edge fielding data we can take a look at whether and how Utley is getting to more balls than other 2Bs and even which metric is closer to the target on Utley's defensive performance.

Utley's Inside Edge Fielding Data

(_) in the body cells of the table show play opportunities.

This data is not very meaningful in isolation. 98% of routine plays sounds good, but if the league average is 99.5% then Utley is losing ground on the easy ones. Similarly, 96.7% on routine plays in 2014 looks like a decline, but if the rest of the league has seen a drop in its average on routine plays then it might not be a decline at all but an artifact of the reporting. So we need to look at how Utley's percentages compare to other 2Bs. Here comes another table:

MLB 2B Inside Edge Fielding Data

First, an observation. The 2014 data looks markedly different than the 2012-13. The number of total plays seems to have increased. And the numbers of plays classified as Impossible, Remote, and Unlikely in particular have increased. The possible explanations for this change are many. 2Bs could be seeing more action because of the increased use of the shift on both LHB and RHB, moving them into positions where balls would otherwise be classified as hit at the SS or 1B. It could also be that so far more balls are being hit to 2Bs than the past couple of years, shift or no shift. As for the change in the proportion of certain types of play toward the more difficult, that requires even more speculation. Again, it could just be that more of those types of balls are being hit in the 2Bs' vicinities. But I think the table indicates a different, system-oriented explanation. When observers record data, they inevitably learn about their own observational capacities as well as whatever subject they are studying. In this case, we see that Even plays are defined as being made between 40% and 60%. But over 2012-2013 they were being made at a greater than 60% rate. As a result, the scouts might have adjusted how they sort balls in play into the various bins. Specifically, they might have started pushing balls they used to classify as Unlikely into the Even bin and, to compensate, pushed some Remote balls into the Unlikely bin. This is just speculation, and, once again, the increase in shifting probably has a lot to do with the differences we are seeing here. Moreover, the data for 2014 is incomplete. By the end of the year we might find out that the changes are just an aberration. Anyway, enough meta-data hand-wringing, back to Utley.

The Inside Edge data, sadly, lends support to how DRS rates Utley's defensive performance. On Routine and Likely plays, Utley is right around league average since 2012. While Utley's Routine % is slightly down in 2014, that could be a matter of a couple extra errors early in the season. And his Likely % is elevated, although unsustainably so. He also shows well in the Unlikely bin. His rate is above average, but don't overestimate that number. It is at most one play more than the average defender. On the other hand, there are two areas where Utley is losing ground to the competition: Even and Remote. In both of these areas Utlely's conversion rate is below the league average. Of course, the difference between the plays that Utley hasn't converted and the hypothetical league average 2B defender amounts to about 10 plays, valued at about 5 runs, over 3 seasons. If Utley is a below average defender at 2B, it is by a very slim margin.

So, where are we? UZR rates Utley as still among the best defensive 2Bs in MLB, but DRS and Inside Edge suggest he is average or slightly below. The weight of evidence leans toward the conclusion that Utley is now an average defensive 2B, which is pretty good for a 35 year old. However, it is worth bearing in mind that UZR is the only publicly available defensive metric that breaks the field up into territorial shards and compares players within those shards. DRS tends to emphasize great plays in its +/- assignments and Inside Edge does not distinguish the difficulty of plays within its bins. While the evidence leans toward Utley being league average, it is hardly conclusive.

Clearly, Utley is in decline and it is affecting his defense. He is no longer an elite defensive 2B. Some have suggested that as he declines he will have to move to a new position where his reduced range is not such a problem. One position that always comes up is 3B. Let's stop that non-sense. Utley has never had the arm for 3B. Can we stop talking about that now? Good. Another position that has been mentioned is LF, where his arm is not as much a problem and he'll have more time to get to balls. That is at least a plausible move for Utley. But it is not likely to provide more value than having him play 2B. If Utley moved to LF from 2B he would lose about 10 runs on the defensive spectrum. That means you have to think that Utley at 2B plus whatever LF would be at least 10 runs worse than Utley at LF plus whatever 2B. That strikes me as unlikely. It is easier to get hitters who are capable LFs than hitters who are capable 2Bs. While the Phillies are currently bereft of OF talent, they aren't exactly bursting in 2Bs either. Cesar Hernandez and Freddy Galvis might be able to handle 2B defensively, but they have not inspired confidence in their ability to hit MLB pitchers. If a move to LF is in the cards for Utley, it should not be any time soon. For this season and at least the next, let's enjoy The Man march toward the HoF at 2B.