Concerns with Luke Gregerson

For the last few months, it has been no secret that the Cardinals were going to spend a portion of the offseason attempting to upgrade the bullpen. The team lost its last two closers this offseason in Trevor Rosenthal and Seung-Hwan Oh. The two best returning relievers are both lefties in Tyler Lyons and Brett Cecil, and the rest of the returnees are a bit of a mixed bag.

For a guarantee of $11 million over the next two years, Luke Gregerson fits more in the mixed bag category than the good relievers in Lyons and Cecil. Compared to using a pitching prospect in the bullpen, Gregerson might not be an upgrade at all. Compared to guaranteeing more than twice that much for a non-elite reliever or giving up actual prospects to pay for a “Proven Closer”, this guarantee isn’t so bad.

Prior to the 2017 season, Gregerson performed well with very good consistency year to year. Last year, he wasn’t very good, grading out as a replacement reliever despite good strikeout and walk numbers due to the high number of homers.

A large number of homers can sometimes be bad luck, but sometimes it can indicate a loss of stuff. Let’s check out the stuff first. I’ve heard some concerns about Gregerson losing a couple miles per hour off his fastball. That isn’t true. The chart below from Brooks Baseball shows his sinker and slider velocities over the past few years. He only throws the sinker and slider a decent amount so looking at his fourseam fastball velocity and getting concerned over a pitch he throws 6% of the time isn’t going to give you a very good idea of what’s going.

Brooks Baseball

The velocity is fine, particularly on his fastball. There have been some deviations on his slider, but no real changes in five years or so. Gregerson throws multiple types of sliders and he changes speeds on purpose so that can also play a role in the change in velocity, as he told David Laurila of FanGraphs.

I’ve always varied the speed of my slider. It depends on which angle I’m trying to make it break at. I’ll change the speed depending on the hitter and where I’m trying to put the ball. I’ll slow it down if I think a guy is on it. Sometimes I’ll throw a really slow one in there, just to get him off balance. If you can keep a guy on his front foot instead of his back, or his back instead of his front — anything you can do to keep a hitter out of his comfort zone is good.

So velocity isn’t the real problem. Let’s take a quick look at some results. This is whiff percentage on the pitch.

Brooks Baseball

The whiffs were down last season on the slider, although it was still the second-highest percentage of his career. That they went down last season isn’t a great season, but 27% is still a very good number for a slider, or any pitch.

Let’s take a quick look at vertical movement on his pitches as well.

Brooks Baseball

Based on the chart above, it looks like his slider more closely resembled the one from 2012-2015 than the one from 2016. It’s possible that getting more vertical movement on his slider in 2016 resulted in more whiffs, though the whiffs were fine in 2017.

The sinker seems a little more problematic as the pitch appears to be getting less sink than it has in the past. Let’s look at the ground balls in play to see if that had any impact.

Brooks Baseball

As the sinker has gone up a bit, the number of ground balls has gone down. Gregerson has always been a bit of a ground ball pitcher, but it is worth noting that in his first two seasons in Houston, he was more extreme than his days pitching in pitcher’s parks in San Diego and Oakland. In those parks, he threw mostly sliders with his sinker as a secondary pitch. In Houston, he switched things around for the first two years and last year threw the two pitches equally.

If the sinker isn’t as good of a pitch now as it has been in the past, it might behoove Gregerson to go back to the slider-heavy mix as he returns to a more forgiving park. If he can get a little more sink to his fastball, keeping a sinker-heavy or more balanced mix in the juiced ball era could be the way to go.

There are also splits to keep in mind here as he generally uses his slider more against righties and his sinker more against lefties. Using Gregerson more against righties is probably the best way to utilize him. Although he still gave up his share of homers against righties last year, Gregerson was more effective with a 29% k-rate, 7% walk rate and 4.03 FIP with the platoon advantage.

About those homers. Keeping the ball in the park was the main problem for Gregerson last season as if he had put up league-average HR/FB rate, he would have been a pretty competent reliever. We can take a look at Statcast data and see how much of a role luck played with data from Baseball Savant.

Gregerson gave up 13 homers last year. I took a look at the 425 pitchers who gave up at least five homers last season and compared their expected wOBA (xwOBA). This provides an idea of how much damage can be expected based launch angle and exit velocity. A home run’s actual wOBA is going to be 2.000. The expected wOBA is going to take a look at the outcomes of batted balls and determine what the average outcome is for those batted ball regardless of park or conditions. The average xWOBA on homers last year is 1.287. Gregerson’s was 1.055, which means he was pretty unlucky and gave up some cheap homers. His xwOBA on homers ranked 390th out of 425 which means he was in the bottom 10% of the league (FWIW, Carlos Martinez was also down there as well).

Here are the individual homers and the percentage chance they would turn into homers or outs, sorted by home run percentage.

We’ve got four no-doubters, three that are pretty close to 50/50, and then six homers that are pretty cheap. Here’s the spray chart on the launch angle for that last one.

Baseball Savant

That ball goes for a hit most of the time, but rarely gets out of the park. I also checked all batted ball with an xwOBA above .800 to find potential lucky outs. There was one, a long double from Carlos Gomez that gets out of the park 73% of the time. There were another three 50/50 shots and a handful between 10% and 40%.

All in all, I don’t think it is reasonable to assume a league-average home rate based on Gregerson’s performance last year. He probably deserves to give up around 10 homers, which is still less than 13, but not the seven that xFIP would give him.

Gregerson was pretty unlucky last season, but his sinker didn’t perform at the level it had been in previous seasons. There are good reasons to expect a bounce-back performance, and looking at the cheap homers as well as his 3.45 away FIP are positive indicators. That doesn’t eliminate concerns, and it isn’t completely clear the team couldn’t have received a similar performance without guaranteeing any money, but when it comes to relievers, paying a low sum is a much better idea given how often reliever performance shifts.