The Cody Allen Regression


In an escape from days of grueling bar prep, Mike Hattery has dropped in to share concerns about Cody Allen.

Cody Allen has been a remarkable reliever in his time with the Cleveland Indians. In the five-year stretch starting in 2013 and ending in 2017, Allen was 10th among MLB relievers in fWAR, 23rd in ERA (2.59), and 11th in saves (122). 2018 has been a different story, with a 4.18 ERA and an even worse FIP. This season, he has been, quite simply, not good.

Of course, this happens to relievers from time to time and reacting to struggles in quarter seasons of reliever performances can be dangerous but with an Indians’ bullpen in strife, Allen’s struggles have been particularly consequential. In terms of analyzing a struggling pitcher, the first step is to find the surface-level changes in their profile and once isolated, search for root causes.

For the first step, what are the surface level differences?

  1. Allen is striking out significantly fewer hitters with a K% of 27.4% of 2018 as compared to his career K% of 31.2%.
  2. Allen is walking slightly more hitters with 9.4% in 2018 as compared to a career BB% of 9.1%.
  3. Allen is allowing home runs at a higher than normal rate on fly balls of 13.5% in 2018 as compared to 9.5% for his career.

Now for the second step, what are potential root causes of Allen’s issues in generating strikeouts and limiting hard contact?

The first potential root cause is a degradation of his stuff. With Allen, the arsenal is obviously limited to two pitches, a four-seam fastball which Allen throws roughly 60-65 percent of the time and a curveball which Allen throws 35-40 percent of the time. Time to talk fastball. First, the spin rate has declined 70-80 RPMs off his normal spin rate. However, fastball velocity has declined over the past three seasons as well.

According to Statcast, this is his fastball velocity over the years:

2015: 95.5 MPH

2016: 94.8 MPH

2017: 94.3 MPH

2018: 93.7 MPH

This is problematic and with Allen’s high usage rate, fastball velocity decline limits his margin for error. More concerning, he has been throwing the fastball middle-middle a ton in 2018 as compared to the rest of his career.

Courtesy of Brooks Baseball

Allen is just grooving more pitches in general, with a meatball percentage of 8.2 percent, the highest of his career in the Statcast era, and higher than the league average of 7.2 percent. Before going any farther, the right-hander has lost a tick of velocity in the last year and two ticks since 2015. Even worse, Allen is throwing his fastball over the middle of the plate far more often and is simply not commanding the pitch in or out of the zone.

The curveball remains a plus offering with an insane wRC of 38 but compared to his career that is high with a career wRC+ of 13. The spike curveball has been a special pitch that pairs especially well with a high spin rate fastball from the same slot. But the degradation in spin and velocity gives Allen a lesser arsenal to tunnel and does not put him up in counts.

Allen’s curveball is getting far fewer swings outside the strike zone at just 27.5 percent compared to his career O-Swing% of 36.9 percent. The spin rate and overall movement of the offering look fairly similar but an issue appears that the arsenal gap challenges may have diminished the value of the curveball.

Right now, he looks like a different pitcher with diminished fastball velocity, which allows hitters to be more patient and it makes laying off the curveball a bit easier, which is dangerous for a two-pitch pitcher. Perhaps Allen is just suffering from a dead arm period, perhaps the reliever is experiencing an injury which would be remedied easily by a DL stint, perhaps Allen is simply worn out by early season over-use due to a struggling pen.

In darker light, maybe Allen is simply experiencing the decline of a fastball which is central to creating the margin for error in a two-pitch arsenal. Unless his velocity and command improve on the fastball, it will continue to be a bad season for the former firearm.