For Oliver Drake, success begins with a consistent release point

He’s not the most beloved reliever in the Brewers’ bullpen. That has to be Corey Knebel. He’s probably not the most hated, either. But in one way, I think Oliver Drake is by far the most intriguing. Way back at the beginning of June I tweeted this:

If you click play, you’ll see that one day, Oliver Drake is maintaining his release point when throwing both his four-seam fastball and his splitter. The very next day, his four-seam release point is noticeably different.

I originally looked his release points because I was watching Drake in late May, and could tell when he was throwing his four-seam based on when he utilized a high-three quarters delivery. His splitter comes out of his hand from way over the top, so the difference was noticeable.

Let’s take a look at Oliver Drake’s average release points in 2017:

First, it’s important to note that this is from the catcher’s perspective. Second, bear with me. I realize the gap between the black and purple circles seems insignificant, maybe humorously so, but it isn’t. Over the course of the season, on average, Drake has released his fastball around half a foot (0.52 ft) to the right of his splitter (from his perspective), and a tenth of a foot (0.08) lower.

If we use the distance formula, where A is the distance between the horizontal release point of the splitter and four-seam, and B is the vertical difference, d= √(A²+ B²), then √(.52²+ .08²)=.526 feet=6.3 inches apart.

Six inches may not seem like a lot, until one begins to search for pitchers with Drake’s profile.

Despite a handful of sliders, Drake is a two-pitch pitcher, with the second highest rate of splitter-usage among qualified relievers. The top three relievers on that list are all essentially two-pitch guys, the other two being Hector Neris and Koji Uehara. These pitchers, like most major league pitchers, have more consistent release points than Drake. Let’s look at Uehara’s:

In all fairness, Uehara’s release points are uncommonly consistent. So let’s compare the distance between the fastball and splitter release points of all the relievers who throw splitters at least 25% of the time to those of Drake’s: Uehara, Neris, Blake Parker, and Junichi Tazawa. Parker and Tazawa also regularly spin curveballs, but for the sake of this exercise, that’s mostly irrelevant.

In the table below, “Drake 1” is Drake’s stats from prior to May 25. On that date, the reliever logged his first appearance in which he clearly attempted match his release points; his splitter and four-seam were, on average, just 1.8 inches apart that day. “Drake 2” is every one of his appearances since May 25. “Drake 3” is each of Drake’s appearances since July 25. Every one* of Drake’s appearances since July 25 have featured four-seam and splitter release points which, on average, are less than three inches apart.

So what do all these numbers tell us? For one, Drake’s 6.3 inches of separation on the year are by far the worst, with Parker and Tazawa (who each have curveballs to mix in) coming next. Additionally, Drake’s release points were a whopping 10 inches apart from April 1 to May 25. Since then, he’s clearly made a conscious effort to bring those release points together, especially in August.

So why is Drake making a change? Baseball writers/players will tell you: Maintaining a steady release point generally breeds more success.

From the linked article:

Again, [release point] consistency on a high level means more success, but it’s on a spectrum of importance for different pitchers and pitch types.

This article was written prior to the 2015 season, and mentions pitchers like Max Scherzer, Collin McHugh, and Mike Fiers as a few of the numerous exceptions to the “rule” of consistent release points. There are two notable factors that make Drake unlike those those pitchers, however:

1. Drake’s average release points in 2017 are farther apart (6.3 in.) than any of the aforementioned pitchers’ were in 2014.
2. Drake only throws two pitches. McHugh, Fiers, and Scherzer all threw four consistently in 2014.

This is important, because Drake relies on his spilt-finger fastball, and a splitter’s primary function is to fool a hitter, much like a changeup. John Smoltz described the splitter for MLB.com:

“It’s going to be thrown just like a fastball... The split-finger fastball is supposed to act just like a fastball...and then fade, tumble away.”

The trouble is, I’m pretty sure that hitters can tell which pitch Drake is throwing when his release points are 6+ inches apart. MLB hitters have perhaps the most difficult job in sports, and if they can identify—and crush—pitches in midair, I have to imagine they can tell when Drake drops his arm for a fastball.

Consider:

When his release points are noticeably different, he’s essentially tipping his pitches at the moment of release.

So naturally, I set out to find if Drake’s pitching stats are better when he releases his two pitches from the same spot. I went through each of Drake’s appearances on Brooks Baseball and logged the differences between the splitter and the four-seam’s average horizontal and vertical release points for each individual game.

If the average distance between the release point of the splitter and four-seam for a game was more than .50 feet, or six inches, it went in the “different release point” camp. If the distance was less than six inches, it was logged as “same release point.”

It’s a subjective decision, based on the relative roundness of six inches (or .5 feet) and the fact that six inches of separation is roughly average for Drake in 2017. Interestingly, .40-.60 feet between the release points of the splitter and four-seam was sort of a no-man’s land, with just five of Drake’s 53* appearances landing in that range. So Drake either does a good or terrible job matching the release points of his two pitches. Huh.

Here are my findings:

And this is where the obligatory small sample size warning comes in. Still, its nice this worked out, because it took quite awhile to separate his appearances based on my parameters.

This small sample seems to echo what I said before about Drake “tipping” pitches. When his release points are clearly different, he gives up a lot of extra-base hits, including each of his five allowed home runs. That isn’t particularly surprising when his two pitches are a 92-MPH fastball and a pitch meant to trick you into thinking it’s a 92-MPH fastball. Drake needs to be able to play on a hitter’s uncertainty.

But there’s still the question of why. Why did Drake start the year with a lower arm slot for his four-seam fastball if he’s fully capable of commanding the pitch from the same slot as his splitter?

My best guess is that Drake feels he gets better velocity off of his fastball in the lower slot, while he uses the extreme over-the-top delivery to maximize the drop on his splitter. He utilized a lower slot as a starter in the minor leagues, so it may be more natural for him. (Most deliveries at least look more natural than Drake’s splitter delivery.)

And while there’s some evidence that his fastball slows down when he keeps its release point the same as his splitter, it appears to only be about a half a mile an hour difference.

It seems Drake, or someone on the Brewers staff, has decided that deception is better than a little bit more kick on the heater. I think they’re right.

*Neither Brooks Baseball nor FanGraphs had release point data for Drake’s August 21 appearance against the Giants. I have no idea why. Thankfully, it was a relatively uneventful affair (1 IP, 1 1B, 1 K, 1 BB).

Statistics courtesy of FanGraphs and Brooks Baseball