Khris Davis has elements of being a very good player. No matter what his reputation might be, he’s perfectly good at tracking down fly balls. He runs fairly well, all things considered, and he’s swinging at a career-low rate of pitches out of the zone. What’s most important, as you already know, is that Davis has power. Major power, big-time power, the kind of power that plays in any park. Davis swats a bunch of home runs. He does so while rarely popping up. He has a career wRC+ of 122, which ties him with Carlos Santana and Yoenis Cespedes.

The A’s are happy to have Davis. Any lineup would be happy to have Davis in it. There’s just — look, it’s not like I take extra pleasure in this. I’ve got nothing against Khris Davis. I just feel obligated to point out some numbers. A short while ago, I wrote about Bradley Zimmer’s outstanding throwing arm. It’s no secret that Davis doesn’t have an outstanding throwing arm, himself. It’s a known weakness, but this year has still been exceptional. In the Zimmer post, I combined DRS and UZR arm-value measures into one. I showed you the current top-10 arms. Here are the current bottom-10 arms.

Just as in the Zimmer post, here’s how 2017 left fielders have done defensively, in terms of holding baserunners, and throwing them out. Davis is the point in yellow.

Davis has recorded one assist. There are 40 left fielders with more than that. And when Davis has had a chance, he’s prevented a baserunner from advancing 52% of the time. The league average for left fielders is a hair over 63%. Obviously, when the surface numbers are bad, the underlying numbers must also be bad. These are bad numbers.

Davis is already at -8.5 runs. That’s the number from the earlier table, and while all these arm-value numbers are estimates, they’re the best we’ve got. That’s the third-worst arm-value rating since 2003, which is as far back as we have both DRS and UZR available. And I’m writing this on July 24. The two players with worse outfield arm seasons played nearly twice as many innings in the outfield as Davis has to date. And their arm values were worse by tenths of one run. To put everything on a consistent scale, why not establish a denominator of 1000 innings?

That’s what I’ve done. I looked at every player-season since 2003 in which someone played at least 500 innings in the outfield. I calculated UZR arm per 1000 innings, and I calculated DRS arm per 1000 innings. I figured I should show them both, since they’re somewhat different. Davis, one more time, is shown here, highlighted.

By UZR arm, no one’s been worse. By DRS arm, no one’s been worse. Combine the two, and 2017 Khris Davis is at -13.3 arm runs per 1000 innings. That would be easily the worst in the sample, with the second-worst season belonging to 2013 Juan Pierre, at -9.7. Davis is taking this and running away with it.

Because the season hasn’t finished, Davis could recover. His arm could play better from here on out. But that’s likely to require some luck, since the arm isn’t good on its own. Davis seems to be having the worst arm season on recent record. It’s usually not the kind of thing you worry about, but in this particular case, it’s been costing the A’s actual runs. I don’t know what you do about it, given that Davis appears too good at running down flies to be a DH. This is just a part of the Oakland reality.

The one assist to 2017 Davis’ name? Jose Ramirez was the victim.

Ramirez lost track of the ball, and froze. Davis identified what was happening, and hit his cutoff. In one sense, it was a gimme assist. In another, was it, really?