The Milwaukee Brewers have developed another elite reliever.
After a 13 inning cup of coffee in 2019, Williams exploded onto the scene in 2020, allowing just one earned run in 27.0 innings and earning a ludicrous 53% strikeout rate. His microscopic 0.33 ERA translated to a borderline incomprehensible ERA- of 7.
The key? A devastating changeup. While the pitch was well above average in ‘19, it was next level in ‘20, as he learned that throwing it more would lead to better success. Over his two major league samples, Williams’ usage of his changeup sprang up 36.6% to 52.7%. As a result, his xwOBA on the pitch plummeted from an already elite .211 to an unfair .110.
Williams was awarded the National League Rookie of the Year for his efforts. And as the 2021 season comes around, the Brewers figure him to be part of a rather formidable one two punch with fellow relief phenom Josh Hader—so long as the league doesn’t figure him out.
As great as he was, it’s fair to wonder a few things. For one, will his 53% K% hold for an entire season? Probably not. Is 27.0 innings too small of a sample to project? Likely. Was Williams helped by the fact that he rarely, if ever, faced the same hitter more than once? Possibly.
What Williams brings to the table are two elite offerings—a fastball that can reach triple digits, coupled with a changeup that drops 25% more than other changeups in its class (similar speed and release point extension). So what’s the problem?
Consider the following side by side video. On the left, a fastball, on the right, a changeup. Do you see any differences a hitter may pick up on?
Admittedly, it may be hard to pick up in real time. Instead, consider the following photo, stopped at release point.
In the photo, it’s a little more clear. The release points between the fastball and change are visibly different. To be fair, we do expect some random variance from pitch to pitch, and it isn’t necessarily detrimental for a pitcher so have slightly different release points between all of his pitches.
But the fault in that thought process is this: Williams is mostly a two pitch pitcher, and with advanced scouting and pitch tunneling being all the rage, it’s also reasonable to think this is something that could get exploited by major league hitters as he makes his next round across the junior circuit.
Here is the video once again, this time overlayed instead of side by side, and slowed down at the time of release.
When a pitcher creates an effective pitch tunnel, an overlayed video would ideally show one release point, as well as one baseball traveling at the same trajectory until they eventually split. In this case, while the baseballs do start in similar paths, we clearly see two arms, and two separate baseballs at release.
Maybe the difference is too subtle for hitters to pick up on. Maybe Williams is too hard to hit, even if they could. But since MLB organizations—and even sometimes pitchers on their own dime—invest and rely so heavily on high speed video, I doubt either party takes that gamble. It’s also possible Williams is aware himself and is taking steps to adjust.
Devin Williams is for real, there is no doubt about that. No matter what, he will probably still regress some in ‘21. But on the chance that hitters are able to take advantage of this difference between pitches—small as it is, it could be prudent to wonder how much it could add to the regression.
Brian Menéndez is a contributing writer for Beyond the Box Score, as well as a senior writer for DRaysBay. Additionally, he has been featured in The Hardball Times. You can find Brian on Twitter at @briantalksbsb.