Why Do The Defensive Metrics Hate Howie Kendrick?

Howie Kendrick has worked out pretty well, right? His 111 wRC+ is better than his career average of 108, and just about exactly in line with what he put up in his last few years with the Angels. Among 22 qualified second basemen, that’s eighth, better than Jose Altuve or Ian Kinsler or Brandon Phillips or Robinson Cano. Though he’s slumped lately, like most of the Dodger hitters, he’s generally been a good, steady addition to the team, a good enough one that I don’t think either the Dodgers or the Marlins would undo the Dee Gordon trade if they had the opportunity. (Yes, Gordon has been more productive to date, but I don’t think you have to look too hard to see his regression is in full swing, and the move also brought Enrique Hernandez, Chris Hatcher, and Austin Barnes.)

But something interesting was brought up in the comments the other day, something I hadn’t realized before: Why is Kendrick ranked so poorly on defense? By the eye test, he’s seemed just fine — not stellar, but certainly adequate — but the metrics just don’t back that up.

• DRS: -8 (last)
• UZR/150: -10.7 (20th)
• Defense: -3.6 (21st)

That last one is FanGraphs’ all-inclusive defensive stat that accounts for position, but they all paint the same picture: not great! After all the talk about the Dodgers wanting to improve their-up-the-middle defense over the winter, which they generally did, it’s surprising to see. So, what’s going wrong here?

The obvious answer is “don’t put too much stock into partial-season defensive metrics,” and I fully believe that. UZR/150, for example, is particularly problematic, because it takes his current UZR and spreads it out over 150 games, assuming he’ll do exactly the same thing. It’s like saying that just because Bryce Harper has a .715 slugging percentage right now, he’ll still have it after 150 games. Three months or so of defensive data just isn’t enough to say with certainty that Kendrick has suddenly turned from a consistently above-average defender into a total disaster.

But I also understand that’s not a particularly satisfying answer. So, with the full understanding that we’ll need to revisit this at the end of the year, let’s dig into what exactly the issue seems to be. UZR/150 is formulated with three inputs:

Double Play Runs (-0.3)

Not much to see here, so we’ll just note this and move on. Kendrick ranks 12th of 22 in turning double plays, but there’s just not a big spread here. From the best (Neil Walker) to the worst (Brandon Phillips), the difference is only 2.4 runs. An overwhelming majority of second basemen are within one run of each other. Don’t worry about this part too much.

Error Runs (0.4)

This is generally the first question you ask when you hear that a fielder’s numbers aren’t great: Well, is he making a ton of mistakes? For Kendrick, no, not really. For what little I care about the traditional error stat, he’s booted only three balls all season. This stat can have a larger spread, with the difference between Joe Panik and Addison Russell coming in at 5.8 runs. For Kendrick, his 0.4 ranks 13th of 24, which is average. He’d actually been ever so slightly in the negatives for the last few years. No problems here.

Range Runs (-4.8)

Now we’re getting somewhere. The overwhelming majority of the -4.6 UZR that Kendrick has — which, extrapolated over 150 games, gets him that -10.7 — is in Range Runs. This is the stat with the biggest spread (Russell tops it with 9.8, Johnny Giavotella is at the bottom with -5.6, making for a 15.4 run spread), and it’s basically saying that Kendrick is late-career Derek Jeter: He won’t get to much, but he’ll make the plays on what he gets to.

So is that fair? It’s a considerable drop from what Kendrick had done with the Angels. Between 2011-14, among 22 second basemen with 2,500 innings played, he’d ranked fourth. What this now suggests is that he’s below-average at making anything other than the normal, routine plays, and as it turns out, that’s pretty much what the metrics are saying. On FanGraphs, you can see that he’s made the fourth-fewest “out of zone” plays of any second baseman. Inside Edge grades every defensive play and marks it as falling into one of the following categories, along with the percentage of times that play ought to be made:

• Impossible (0%)
• Remote (1-10%)
• Unlikely (10-40%)
• Even (40-60%)
• Likely (60-90%)
• Routine (90-100%)

Making a “Routine” play doesn’t add a lot of value, but missing one does. Conversely, missing a “Remote” play won’t hurt that much, but making one adds considerable value. Let’s run those categories again, this time showing Kendrick’s percentage made and rank among second basemen.

• Impossible (0%) 0, obviously. It’s impossible.
• Remote (1-10%) 0, tied for 9th with a dozen others.
• Unlikely (10-40%) 33.3%, tied for 11th.
• Even (40-60%) 66.7%, tied for 7th.
• Likely (60-90%) 63.6%, tied for last.
• Routine (90-100%) 98.9%, 6th.

He hasn’t really done much in the three hardest categories to stand out and provide extra value, and while he’s done fine with the “Even” plays, the fact that he’s failed to get to a higher percentage of “Likely” balls than everyone really hurts him. This kind of thing can be influenced by extreme shifts, but as we’ve discussed, the Dodgers really aren’t shifting.

It’s possible, however, that after nearly a decade in the big leagues, Kendrick has simply begun to slow down a little, and you can see that in other areas. Like clockwork with the Angels, he’d be good for 15-20 stolen base attempts a year; with the Dodgers, it’s just 4. FanGraphs’ “Speed” metric is more a fun toy than anything that’s rigorously tested, but at 3.2, it’s down to “poor,” after years of marks between 4.5 and 6.

Again, we’re still dealing with small sample size issues here, and the fact that Kendrick really has made very few mistakes makes this kind of defensive ranking hard to stomach. But the claim that he hasn’t made a ton of outstanding plays compared to other second basemen isn’t hard to believe, either. On defense, Kendrick has been perfectly adequate. It’s just not the same as being a plus. Still plenty of time to change that, though. As we’ve said so many times, partial-season defensive metrics aren’t flawless. This is just an attempt to explain why he ranks where he does.