What’s up with David Robertson?

David Robertson’s calling card, year after year, has been consistently excellent performance. He strikes out lots of guys, doesn’t allow many homers, induces weak contact, and — most critically, as far as 2019 is concerned — doesn’t walk a ton of guys. I went over this in greater detail back in January.

So, it’s a little weird to see him struggle so massively in his first three outings: 15 batters faced, five walks, a home run allowed, just one strikeout, and only six outs recorded. Things were punctuated by a long, drawn-out Wilhelm Scream of an outing Wednesday that resulted in the Nationals winning on a walk-off walk. It was a bad way to lose, but games like these happen to every team.

But is this ragged first week enough to write off simply as a blip, a drop of oil in the ocean that is Robertson’s otherwise stellar career numbers? Yes, of course, for now; but these outings haven’t come without their oddities.

First, a green light: Robertson’s velocity is not down. He’s throwing his cutter as hard as ever (93 MPH average on Wednesday), and his offspeed stuff is coming in identical to 2018 numbers. So, that’s one potential hazard out of the way. If we wanted to dig real deep and look at his pitches’ total spin we’d see that the curve is getting about 150 fewer RPM against last season’s averages so far. But, as explained by Dr. Alan Nathan on Baseball Prospectus a few years back, all spin is not created equal, nor is an increase or decrease in spin necessarily indicative of a pitch’s movement or quality:

Now let’s ask whether a random measurement error of 500 rpm in Ɯt makes sense based on what we know about the Trackman system. Given the relationship between transverse spin and movement, it can be shown that a random error of 500 rpm in Ɯt would be expected if the random error in the movement is about 2 inches.

That’s one part of one paragraph of a far more detailed article that’s plucked out to say, essentially: 150 RPM either way doesn’t mean much here. Nothing we can say with confidence, anyway. So, generally speaking, the stuff is fine! It’s getting smacked around and spat on more than normal, but the inherent quality of the pitches isn’t necessarily dampened.

What is not fine, no matter how in line D-Rob’s pitches’ movement is with his career norms, is Robertson’s command of his pitches. It’s mostly absent right now. He’s thrown just 37 of his 71 pitches for strikes (52 percent, way below his career average of 64 percent), and the ones that have found their way into the zone or garnered a swing haven’t generated many misses. Of note: No swings-and-misses on 15 curveballs, a pitch that normally gets in the vicinity of 40 percent whiffs.

That last bit, small sample and all, sticks out to me. The curve is his bread-and-butter, and it’s done zilch so far? Can’t be right. Let’s try and compare a curve from this season to one from late last year and see what we can see.

In exhibit A, we have one of two registered curve Robertson threw on Wednesday, on 2-1 to Juan Soto. None of the 12 pitches that followed would be curves.

In exhibit B, we a sweeping, slurvy curve to C.J. Cron last September. The camera angle of the center field spot at Nationals Park obscures some of the horizontal movement of the pitch, but it doesn’t move nearly the way the one to Cron does, parallax and all. A comparison of BrooksBaseball’s data from the respective games shows a movement difference of about an inch-and-a-half, and BaseballSavant pegs the total spin rate difference of nearly 400 RPM, far more significant than the ‘19 averages — obviously now known to be buffered by the first two appearances — would initially indicate.

And while those disparate camera angles don’t really lend to visual mechanical comparison, something certainly seems the teensiest bit askew, going by Brooks release point plots. Take a look:

BrooksBaseball

The plot is of all of Robertson’s appearances, charting his average vertical (y-axis) and horizontal (x-axis) release points for all curveballs thrown in that appearance. The relative size of the dot indicates the number of pitches recorded.

I have little more than one half-formed theory, but it’s possible the lower arm slot doesn’t let Robertson be as on top of the ball as he might like. It’s little more than a guess which, I can’t stress enough, is all we really have to go on right now.

Now, again, it’s worth re-stating: He only threw two curves Wednesday. It’s not the only problem. But the curve is his best pitch — his putaway — and right now he’s not getting ahead often enough to properly leverage it, nor is he locating it for strikes when he does use it while even or behind. As for the release, well, that could be much ado about nothing. A lower slot could, possibly, lead to decreased total and effective spin. There isn’t enough here yet to draw any reliable conclusions from. For now, we have to be patient and keep a close eye on things as D-Rob looks to right his ship.