When Mitch Keller made it to the majors, everyone quickly noticed that he threw a lot of fastballs. It obviously didn’t work out well in his first two big league starts, while his third start was an average outing. Since returning to Indianapolis, his pitch usage has changed, though it didn’t happen right away.
The Pirates noted when they sent Keller down that they wanted him working on using his off-speed pitches more often. I’ve watched probably 50 of his minor league starts and it’s always been either the catcher and/or the dugout making the pitch calls, so it seems a little odd to say that they want him to work on mixing his pitches more. If you’ve watched Keller over the years, he’s not shaking off a ton of pitches, and sometimes the shake is just location rather than pitch type.
He has dominated in the minors by going fastball heavy his first time through the lineup, then mixing in his off-speed pitches later. The curve rarely showed up in the first inning, while the changeup (before he added the slider in early May) never showed up that early. The memo for him to change his pitch usage has obviously reached the catchers and coaches, because it has changed drastically in his last three starts. He has made six starts since being sent down though, so that change didn’t happen right away.
After watching his first outing, I decided to get exact numbers on his pitch types being used. Here they are from June 24th:
That was more off-speed pitches than usual for Keller, but it has become more extreme since then. I didn’t see all of the June 29th start, so I’ll skip to the last three starts, beginning with July 11th. Here’s the breakdown from that game, which I’ll note were slightly rough estimates because I didn’t track right away until I noticed things were changing:
As you see, the slider started seeing much more usage and the changeup was almost non-existent. For the first time that I’ve seen him pitch, his off-speed pitches were over 50% usage (and it was well over in this case). The result was six shutout innings on two hits, three walks and eight strikeouts.
On July 17th, there was no estimate because I was tracking from the start. The only thing I’ll note for these last two games is that four specific pitches weren’t clear due to camera work. With his changeup and slider both nearly the same velocity, it’s impossible to say which pitch he threw. More pitches weren’t shown during these games, but the velocity made them easily identifiable. I called two of the unknown pitches changeups (both to lefties), so at worst, my numbers below are off by two on either the sliders or changeups. The total number of off-speed pitches is correct, even if I’m 1-2 off on the actual pitch being thrown.
Here are the July 17th numbers:
Again he is over 50%, but just barely, though the fastball usage deserves a little more explanation. He faced the opposing pitcher twice and threw him eight pitches total, all fastballs. So if you look at the position players in the lineup, they saw 35 fastballs and 44 off-speed pitches. The results were three runs over five innings on four hits and two walks.
What really changed in the July 11th game and has continued through his next two starts was how quickly he went to the off-speed pitches. Batters in the first inning each time were seeing first pitch breaking balls. That is new from him. We wouldn’t see that until the second time through the lineup in the past. If he threw his curve early in the past, it was almost always with two strikes.
Finally, here’s the breakdown of last night’s start:
He gave up one run in six innings last night and that one run came from a slow ground ball single, a stolen base, a wild pitch and a ground out. So he didn’t exactly get hit hard. He didn’t walk any batters and picked up five strikeouts.
The total breakdown from the last three games is as follows:
FB: 132 (47.6%)
CV: 73 (26.4%)
SL: 63 (22.7%)
CH: 9 (3.2%)
Things have obviously changed over these last three starts and it has resulted in four runs over 17 innings, which is fairly impressive in a year when offense is up in Triple-A. Keller has thrown between 43-45 fastballs in each of these games, while using his off-speed pitches more than 50% of the time. He has basically phased out his changeup. He’s throwing the slider a little harder now as well.
It was easy to tell the slider apart from the changeup early in it’s usage because the slider was up in the zone at 86-88 MPH and the changeup was always down at 88-90 MPH. Now he’s throwing the slider down more often and it has hit 90 MPH multiple times. It could be now that the pitches are so similar that he’s going with the one that feels most comfortable. It could also be that his changeup hasn’t been an effective pitch this season. I’ve seen the pitch spiked in front of the plate multiple times over these last six starts, and very rarely has it resulted in anything other than a ball.
The important part now is that he is mixing his pitches much better, still getting strong results this way, and he’s not going fastball heavy early in the game. Any scouting report on him up to this point had to include that batters were going to see first pitch fastballs early in the game. That hasn’t been true these last three games and that should help him going forward.
UPDATE: We usually don’t update articles like this, but I wanted to add his pitch totals from his next start, just to show his trend is continuing. On July 27th, he threw 98 pitches. Here’s the breakdown: