Michael Lorenzen is probably fine

Michael Lorenzen has appeared in five games out of the Cincinnati bullpen in the first couple weeks of the 2020 season, and nearly all of them have gone terribly. The first one included this homer by Miguel Cabrera:

The next time he was called on to pitch, he gave up another homer, this time to C.J. Cron:

Then the Reds gave him another opportunity to pitch two days later against the Cubs, and Javier Baez did this:

Lorenzen finally turned in a scoreless, homerless outing in his fourth appearance of the season, but when the team asked him to slam the door in a 0-0 game against Cleveland on Wednesday, he once again fell flat. After entering with a runner at second and one out, Lorenzen walked Oscar Mercado, then allowed a base hit to Cesar Hernandez to score the game’s first run. When Jose Ramirez grounded into a forceout in the next at-bat, some heads-up baserunning by Mercado granted Cleveland another run. Lorenzen allowed another base hit to Francisco Lindor before getting out of the inning, his struggle to record just two outs finally put to rest.

Needless to say, this is a concerning development for a Reds bullpen that has collectively has the sixth-worst ERA and fourth-worst FIP in the majors this season, offsetting a rotation that has been the most valuable in the sport. Lorenzen is far from the only one deserving of blame for that, but he is the most easily identifiable face of the unit’s struggles. In 4.1 innings, he’s allowed six runs on six hits including three homers, has walked two and struck out five. Those are shocking numbers considering how Lorenzen fared just a year ago, when he threw 83.1 innings and allowed a 2.92 ERA and 3.66 FIP, with 85 strikeouts and 28 walks to go with nine homers allowed.

So what’s the problem? Well, the simplest (and perhaps most unsatisfying) answer is that the sample is just incredibly small. In each of the last three seasons, Lorenzen has faced at least 340 hitters; this year, he’s faced 21. If he were a starter, this would essentially amount to one bad start. Any reliever is bound to have a bad sample of that size at some point in a season — last year, for example, Lorenzen had a five game stretch from May 26-June 6 in which he allowed seven runs on eights with five strikeouts, a walk and three homers. We didn’t wonder if he was broken, though, because before that, he had a 1.65 ERA in 27.1 innings.

But I didn’t make you click on this story just to hear me say, “it’s too early, go back to bed.” Even though the sample size is so small, there is some information we can look for that doesn’t require a large number of data points to be legitimate. First, let’s try to answer an odd question: Was Lorenzen really good last year, or was he just lucky? We know his run-prevention numbers were great last year, but even one entire season isn’t necessarily a big enough sample to determine a player’s true talent level. Is there any chance his low ERA could have simply been the result of some good luck, making some regression in 2020 seem likely?

Not really! Lorenzen’s FIP was a decent bit higher than his ERA, but his FIP- was still just 81 — meaning his FIP was 81% of what the league average rate is. That still makes him an unambiguously above-average pitcher. But even FIP can hide some lucky breaks here and there — namely home runs per fly ball, which aren’t that sticky from year to year for any individual player. Did Lorenzen give up much fewer homers than the number of fly balls he allowed would indicate? Nope! His HR/FB% in 2019 was 12.5%, just a couple ticks under the league average of 15.3%.

Another way we can check on a player’s luck is by looking at his expected stats at Statcast, which are determined by advanced technology helping to look at quality-of-contact metrics like exit velocity and launch angle to determine how likely any given batted ball is to fall for a hit or leave the park for a home run, and combine that with a pitcher’s walk and strikeout rates to paint a picture of what his production ought to look like in a context-neutral environment. If these metrics don’t line up with what a pitcher’s real numbers, they can help predict a change in his fortune coming in the future. With Lorenzen last year, though, that doesn’t seem to be the case — Statcast seems to think he was exactly as good as he should have been.

So we have our verdict: Traditional stats say he was great. Predictive stats say he was great. Fancy computer tracking stats say he was great. Lorenzen should be great this year. Why isn’t he?

Now we can look at some of the data from this year that doesn’t take several weeks to stabilize, like Lorenzen’s stuff on the mound. A dip in velocity or spin rate could mean he doesn’t have the arsenal to get guys out like he used to. Here, though, we find more good news. Lorenzen is throwing harder than he ever has, and is spinning the ball more than ever.

On paper, Lorenzen’s stuff is better than it’s ever been. That’s an exciting thing to see — Lorenzen is an incredible athlete working with a great arm on the mound, and also seems to be an extremely hard worker, which is pretty much everything the new Cincinnati pitching regime could want. He’s exactly the kind of guy you’d expect to take some kind of step forward with his arsenal this season, and early on, we can see that kind of step being taken with his slider. Lorenzen’s slider has been a work in progress throughout his big league career, having below average movement in his rookie season and slowly building up to about an inch and a half above average break both vertically and horizontally last season. It looked like this:

This year, according to Statcast, that same pitch now breaks 4.4 inches above average vertically, and 5.5 inches above average horizontally. It looks like this now:

Of the 37 pitchers in baseball whose sliders move at least four inches above average on the vertical plane, Lorenzen is one of just eight who get at least four inches more movement than average on both axes (two of the other seven are Trevor Bauer and Sonny Gray). In one offseason, he turned a humdrum slider into one of the most unique pitches in the game.

If Lorenzen’s stuff is in a good place, then what about his location? That’s trickier to analyze after just five appearances, but so far, the results there aren’t out of the ordinary either. His zone rate is the lowest of his career, but only by 1.5% below last year. Despite that, he currently has the highest opponents’ swing rate of his career, along with the lowest contact rate of his career. And when he is in the zone, his meatball pitch% — which Statcast classifies as a pitch in the middle-middle of the plate — is just 4.8%, also the lowest of his career.

Lorenzen is coming off a career year that held up to scrutiny from every statistical angle, and has returned in 2020 with possibly the best stuff he’s ever had, often located right where he wants it. So what’s wrong with him? It could be a pitch selection issue, or something else that will take a while to sort out. More likely, though, the answer is that there isn’t anything wrong, and that even though it will take a lot of games for him to even out his current numbers, he’ll probably eventually get there.