Oakland’s Defense Has Been a Nightmare

This year’s A’s have been something of a mystery to me. No matter how you break them down, they don’t ever look very good, but they’ve felt like a statistical underachiever. Let me explain. You know our BaseRuns standings? The A’s have an actual run differential of -59, which is one of the worst in the game. However, they have an estimated BaseRuns run differential of -2, which is perfectly ordinary. That means the A’s have a difference of 57 runs, where no other team has a difference greater than 35. And while the lineup is a part of it, the run prevention has been worse than the estimate by 0.79 runs per game. No other team has been worse by more than 0.36.

Something has caused the A’s to allow more runs than they arguably should have. Now, in reality, a variety of things have contributed. There’s seldom ever one explanation. Yet the major factor here is the one described in the headline, and it should come as no surprise to anyone who’s watched the A’s on a regular basis. We’re a third of the way into the season, and the Oakland defense has sucked.

For fun, let’s go back in time a week. The A’s lost to the Indians 8-0, while striking out 17 times. They were given little chance by Corey Kluber to do anything, so, it’s not like the A’s were ever on the verge of winning the contest. And yet, the deficit shouldn’t have spiraled out of control as it did. Buckle up. It’s time to walk through a disaster.

In the bottom of the sixth:

It’s not much, but there’s a quick bobble there on a would-be sacrifice bunt. So instead of there being two runners and one out, there were three runners and none out. Following a subsequent bases-loaded walk, this happened:

Not only should the ball probably have been caught by someone — there was a missed opportunity to at least throw out a runner at second. You don’t really get to see it on camera, but that was just a lack of attention. Two batters later, there’s this:

The first baseman leaves the field, assuming the third out. However, the runner wasn’t out, but rather safe, and then another runner came home. What you see in the video is an out call. That was reversed! Carlos Santana was safe. Another blown opportunity. Or, really, an opportunity simply handed over, created from nothing. I so wish we could be done. But in the next at-bat:

Yeah. And then, in the bottom of the seventh:

I didn’t bother making a video from the same game of an Oakland wild pitch. Nor did I bother making videos of two Oakland passed balls. It’s not like this is all representative; that’s probably Oakland’s worst defensive performance of the season. I’ve made them to look as bad as I could. But, in truth, no team defense has been worse. It’s been the big problem, albeit a problem that wasn’t all that hard to foresee.

Oakland has allowed the fifth-most passed balls. They’re tied for second in wild pitches. They’ve allowed the most stolen bases. But, why just look at numbers like that? Here is the whole current MLB picture, in terms of Defensive Runs Saved:

There are the A’s, in dead last. Now here’s the same plot, only this time using Ultimate Zone Rating:

There are the A’s, again, in dead last. And oh, by the way, a year ago, in DRS, the A’s finished dead last. And, in UZR, they finished dead last. The A’s have no shortage of interesting arms, and they’ve assembled a growing crop of interesting bats. Defense is understood to have been a sacrifice, and you should look no further for an explanation of why the A’s have already more or less dropped from the race.

There’s one thing in particular about this A’s defense that blows my mind. This brings us back to the BaseRuns numbers discussed up above. If the defense were simply allowing more hits than usual, because of lousy range, that would show up in the BaseRuns estimates. You know what BaseRuns doesn’t account for? Errors. Simple, traditional errors. Here’s what that looks like:

The A’s have committed 16 more errors than anyone else. We don’t talk about errors very often, because they’re sometimes just judgment calls, and we have better metrics that go beyond counting up regular misplays. But this is insane. The A’s have 59 errors, while the league average is right around 36. Oakland’s error total is three standard deviations higher than the mean. The A’s look terrible by the newer, advanced metrics, but even if you rely on the traditional, familiar stuff, they’re still awful. And, guess what! Historically so, at that.

When it comes to rate stats, errors are a part of fielding percentage. Again, we don’t talk about that much. No great reason to. But we might as well point out when a fielding percentage is out of control. Oakland has a current fielding percentage of .973, while the league is up at .984. That means the A’s have a difference of -11 points. Here are the worst teams, by that measure, since 1950:

If the season ended now, the A’s would be tied for worst. There have been a lot of team-seasons since 1950. No other team has been more error-prone, relative to the league context. And, there’s another way of thinking about this. You won’t be astonished to learn that the A’s lead baseball in unearned runs allowed. They’ve coughed up 41, while the average is about 21. Once again, the A’s are right around three standard deviations worse. If you let me get a little detailed here — I calculated ERA, except using unearned runs instead. I’ve called it unERA. Oakland, right now, has an unERA of 0.71, while the league mark is 0.37. Factor in some rounding and that means the A’s have a difference of 0.35 in the wrong direction. Once again, the worst such teams since 1950:

The only team worse, relatively speaking: the expansion 1962 New York Mets, who finished 40-120. There’s still plenty of time to go, plenty of time for the A’s defense to regress to something less humiliating, but I’d say this is valuable perspective. The A’s, as a team, have played very bad defense. By some measures, they’ve played historically bad defense. It’s no wonder, now, why the estimates think this club should be better — those estimates are overlooking a big giant weakness. For once, we can say it’s the math that’s missing something. Leave errors out and you just don’t get a complete picture of how the A’s play.

I don’t know if it makes it any better that expectations were low, even in March. And I don’t know to what extent the A’s can keep on making these misplays at a similar rate. But, what an achievement. It’s no one single player’s fault, and that’s sort of the problem.