Projecting Baseball’s Freakiest Pitcher

We don’t talk a whole lot about minimums, even though we’re always using them. Sometimes the minimums are obvious, but other times, they’re completely arbitrary. For example, let’s use a minimum of 60 innings pitched last season. Going from there, you could say the Yankees have assembled something hilarious, when you sort pitchers by strikeout rate:

1. Aroldis Chapman, 41.7% strikeouts
2. Andrew Miller, 40.7%
3. Dellin Betances, 39.5%

If you were to just bump the minimum down to 50 innings, though — which would be perfectly defensible — then you get an intruder to spoil the party.

1. Chapman, 41.7%
2. Miller, 40.7%
3. Kenley Jansen, 40.0%

Adjusting minimums is one of those ways you get to sort of manipulate the statistics. If not that, directly, then you get to manipulate statistical arguments. And, you know, while we’re playing around with minimums, why don’t we just drop the minimum to 30 innings, for fun, and-

1. Carter Capps, 49.2%
2. Chapman, 41.7%
3. Miller, 40.7%

You forgot about Carter Capps. Un-forget about him. See, at the end of this post, I have a question for you. I want to know your guess for Capps’ 2016 strikeout rate. It’s a simple and straightforward poll, but before we get to that, we need to talk about Capps so you remember fully what he’s all about.

Capps faced 118 hitters. He struck out 58 of them. With one more strikeout, he would’ve finished with exactly half. Alas, 49% will do, and when you get to the poll, that’s going to be the starting point for your brain. But there are, of course, a few factors working against Capps in terms of keeping his rate that high.

You’ve got the fact that Capps had his season end with elbow discomfort. This isn’t an easy thing to fold in, but if nothing else it most certainly isn’t good. The plus side for the Marlins is that tests didn’t reveal any damage to anything important in there. If the Marlins had been playing for something, Capps probably would’ve returned down the stretch. But his elbow hurt, and he didn’t play, and he hasn’t played since the elbow hurt.

And there’s the matter of regression, which tries to bite at everyone. This is where the sample size comes in — Capps threw 31 innings, which is a pretty small amount of innings. Regression tries to pull extreme numbers back toward personal means and population means. You see in the leaderboard above just how easily Capps’ strikeout rate cleared Chapman’s. That’s notable. And Capps’ previous best season strikeout rate was 29%. Now, he made some changes last year. But his history isn’t worth nothing. You’d expect unseen forces to be pulling against that 49% keeping up. That is a very, very high rate of strikeouts.

That’s that stuff. Those are the arguments for Capps’ strikeout rate coming down. Now let’s talk about why it could stay crazy. Why some people would argue it should stay crazy.

I already showed you the top three strikeout rates, with a 30-inning minimum. Capps’ number came with support. Here are the top three swinging-strike rates:

1. Capps, 25.4% swinging strikes
2. Chapman, 19.3%
3. Koji Uehara, 18.6%

Relatedly, here are the top three lowest contact rates:

1. Capps, 52.8% contact
2. Miller, 59.0%
3. Chapman, 59.5%

By any of these measures, Capps’ performance was insane. It was extremely extreme, and the margins matter. I know that Capps barely threw 30 innings, while other pitchers threw plenty more. That does matter, but consider this — the strongest 30-game span of Chapman’s entire big-league career saw him post a swinging-strike rate of 23%. Capps cleared that by more than two percentage points, as a late-inning National League reliever. Chapman at his most unhittable has still been a little more hittable than Capps, once you control for the sample size.

It makes perfect sense why Capps would be so hard to hit. The scouting, that is, lines up with the statistics. Not only does he throw his fastball 98 miles per hour — he has that borderline illegal delivery. He has that hop, like Jordan Walden, and many people would argue it’s actually blatantly illegal, but Capps got his required clarity from the league and to my knowledge there haven’t been any offseason developments. Capps jumps forward off the rubber, and last year he did that more than ever, and that confers an advantage, and Major League Baseball hasn’t asked him to knock it off. It sure seems like he’s in the clear.

Capps is tall, naturally. Even without the hop, he’d get pretty good extension. With the hop, though, you get stuff like this:

On avg Carter Capps perceived fastball velocity is 3.68 MPH faster than his actual velocity Avg Perceived FB: 101.74 MPH Avg Actual: 98.06

— Daren Willman (@darenw) January 13, 2016

That’s how you turn a Carter Capps fastball into an Aroldis Chapman fastball. You effectively allow Capps to release the baseball closer to home plate than anyone else. There’s no reason to think he’s going to stop, barring league intervention. This is just how Capps throws, and it means hitters have to brace for one of two pitches, knowing they’ll have almost no time to react.

This is what a high heater looks like:

This is what a low breaking ball looks like:

Capps isn’t wild, either. Somehow, with that delivery, he can spot his pitches, so even though he’s a complicated pitcher, the way he succeeds is simple. The following images are taken from Baseball Savant. Capps throws his fastball to the upper half of the zone, and he keeps his breaking ball well below. So hitters can’t eliminate anything, and when the count gets to two strikes, which it often does, Capps mixes his pitches almost 50/50. Guess wrong and you’re out of luck. Guess right and you might still be out of luck.

Because Capps throws so hard — and gets that additional boost — hitters have no choice but to prepare for the fastball. They can’t wait to see something in flight, because they just don’t have the time. Even with this, Capps’ fastball still had one of the highest whiff rates in the game. But now consider the breaking ball. I went to the Baseball Prospectus PITCHf/x leaderboards and pulled every pitch of every type thrown at least 100 times last year. Here are the pitches with the highest swinging-strike rates, that being whiffs / all pitches:

1. Carter Capps slider, 40.8% swinging strikes
2. A.J. Ramos changeup, 34.8%
3. Erik Goeddel splitter, 33.0%

By this measure, Capps’ slider was the most unhittable pitch in the game, which makes sense, given that Capps was the most unhittable pitcher in the game. The pitch had a pretty big lead over the runner-up, teammate Ramos’ changeup. That pitch is probably worthy of its own attention, but we’re focusing on Capps, and more than two-fifths of the time he threw a slider last year, he got a swing and miss. Why? In part because of his slider control, and in part just because of his fastball.

I think everything necessary has been covered. I want you to tell me what you expect Capps’ strikeout rate to be. Begin with the fact that, last year, he struck out 49% of his opponents over a decent but limited sample. Consider that he had a bit of elbow discomfort, and consider the unseen but constant pull of regression. Then consider Capps’ actual profile, and the way that he throws. Consider his arsenal, and consider his ability to locate. The poll below is just for you. But I’m going to put my own guess in the comments below this post, so you don’t feel like I’m staying out. I just don’t want my own guess to bias yours.

This is going to be fun.