When Gerrit Cole arrived in the majors, he didn’t use his slider at all in his first two starts. He was throwing almost 80% fastballs, while turning to his curveball and changeup for the other 20% of his pitches. He used the curveball 17% of the time in his debut, then 11% in the second outing.
Cole’s best pitch coming up was his slider. That’s still his best pitch, with a .583 OPS against in his career, including a .568 this year, and an 18% whiff rate.
It didn’t make sense that Cole wasn’t using his best pitch at all. That changed in his third outing, when he threw the slider 19% of the time. But the next start he dropped down to 5% usage, with the curveball getting 11%.
It wasn’t until his start on July 23rd that he started using the slider more consistently. And maybe it’s not just coincidence that his numbers improved from there. Prior to that July 23rd start, he had a 3.89 ERA and a 3.75 xFIP in 41.2 innings, but only a 5.4 K/9. After he started using the slider more often, he had a 2.85 ERA and a 2.80 xFIP, along with an 8.9 K/9.
Now maybe Cole just needed those initial starts to get adjusted to the majors, and the slider wasn’t the only factor. But it’s easy to make a strong connection between slider usage and the success he had, especially when looking at the huge difference in strikeout rates.
This was 2013.
The Pirates have been criticized in recent years for how they handled Cole in 2016 and 2017, relying on the fastball way too often, and not using his best pitch enough. That’s a horrible strategy to have post-2015, during an era when hitters have been doing a better job destroying fastballs, while pitchers have countered with more of their best pitch — typically a breaking pitch.
So it wasn’t totally inexcusable for Cole to be throwing his fastball so often in his 2013-15 seasons. That was the trend back then across the league, and the most successful teams were working off their fastball and going with a two-seamer often to generate ground balls and quick outs.
I wrote about this after Cole’s second start, asking why he was throwing so many fastballs. The article dug into the Pirates’ strategy of throwing a large majority of fastballs to get ahead, get quick outs, last longer in games, and setting up the breaking stuff. Even with all of that, I wondered a few times why Cole wasn’t throwing his slider at all, since it was his best pitch.
When September rolled around, and Cole was in the middle of his dominant stretch over the final two-plus months, I wondered if he was starting to realize his upside, while noting how much of his success came from the slider.
This is 2019.
We’re now in an era where you get destroyed if you throw your fastball even 70% of the time. The league average fastball usage was around 57.7% consistently from 2013-15. It dropped to 56.7% in 2016, then the 55% range the last two years. This year the average pitcher is throwing his fastball 52.6% of the time.
As I said above, when Gerrit Cole was making his debut, it was a different league. It was somewhat acceptable that he relied on the fastball so much. It was also completely puzzling why he didn’t use his slider — again, his best pitch — as often for his first handful of starts.
I bring all of this up because I’m reminded of Gerrit Cole after Mitch Keller’s start last night. Keller didn’t go to a huge extreme with the fastballs, throwing them 62% of the time. But he threw his curveball — his best pitch — just 7% of the time, while throwing the slider 24% of the time.
Look how that played out in the first inning, when Keller threw 38 pitches and allowed three runs.
The first 13 pitches that Keller threw were fastballs. He got Ronald Acuna to ground out to short, but followed that with a seven pitch, all-fastball at-bat that resulted in a single, followed by a single on a slider on an 0-2 count.
Keller went to a 2-2 count or better against four of his next five batters. He struck out Josh Donaldson on a foul tip fastball. He went to 0-2 on Nick Markakis and threw a fastball — his third in as many pitches — that resulted in an RBI single. A 2-2 fastball to Austin Riley led to a two RBI triple.
That was followed by a four-pitch walk (three fastballs, one changeup). Keller got to another 2-2 count, got a foul ball on a fastball, then threw a curveball and a slider for balls to issue another walk. The two curveballs he threw in that plate appearance from Tyler Flowers were the first curveballs of the night, and both ended as balls in the dirt.
Keller finally got out of the inning with a ground out on a fastball, after going 0-1 on another fastball.
Keller did reduce the fastball usage in his next two innings. He threw 33% fastballs and 41.7% sliders in the second, followed by 57% fastballs and 23.8% sliders in the third. He was under 10% in each inning with his curveball usage.
This is almost a repeat of Gerrit Cole. Keller is using his fastball too often (especially in the first inning, when he threw 73% fastballs), and for some reason is throwing a new breaking pitch often while backing away from his best pitch.
I’ll point out here that Cole’s curveball actually ended up a good pitch, ranking second to the slider. But that still doesn’t explain why he ever was in a situation where he wasn’t using the slider at all, or wasn’t using it often. Keller’s slider might end up as a good pitch, but his curveball is his best pitch, and should be a pitch he relies on much more often.
Cole ended up with good results early in his rookie season, despite not using his slider. That’s because it was 2013. Pitchers could have success throwing 70% fastballs, and while it still would have made sense for Cole to use his slider, the lack of use didn’t prevent him from having good results.
Again, this is 2019. The league is different now. Mitch Keller can’t come up and throw 70% fastballs out of the gate while totally ignoring his best pitch. He’s not going to have the results that Gerrit Cole had. He’s going to get hit around a lot, similar to what happened last night.
This is the part where I’ll point out how the Pirates have the second highest fastball usage in the league at 58.1%. Here are the ERA/xFIP rankings for the top five teams:
- Brewers – 17/7
2. Pirates – 28/21
3. Rangers – 23/24
4. Athletics – 13/28
5. Mets – 18/15
The Brewers and Athletics stand out here. The Brewers throw the most fastballs, and have the seventh best xFIP, but aren’t pitching to that xFIP in large part to giving up too many home runs. That’s a big issue today when you’re throwing so many fastballs. The Athletics are the only team here in the top half in ERA, but are one of the worst pitching staffs in the majors by xFIP.
The overall trend is a pitching staff in the bottom half of the league. As you can see by the teams here, you can still be a contender with an approach that leads to poor pitching results. But it’s probably not a good idea to employ this strategy if you’re building around pitching in 2019, like the Pirates are doing.
Back in 2013, the Pirates threw fastballs 59% of the time, only slightly higher than they are now, but that amount ranked them in the middle of the pack in baseball. The league as a whole has changed the way they pitch, but the Pirates are still using the same overall strategy — 58-59% fastballs, around 30% breaking pitches, and about 10% changeups.
The result hasn’t been good for their overall pitching staff this year. It’s also not going to be good for Mitch Keller if he continues to mirror what Gerrit Cole did when he arrived. I don’t want to make too much of one start, but I will say that it would be completely inexcusable for the Pirates to make the same mistakes with Keller that they made with Cole initially in 2013. The results won’t be as good as Cole’s, and we’re more likely to see results similar to Cole’s 2016-17 seasons.
I see Keller as a top of the rotation starter in the majors one day. My concern is that he won’t reach that in Pittsburgh because of their out-dated approach toward pitching, and his tendency to rely on the fastball too often. This needs to change, otherwise the Pirates will waste Keller in the same way that they wasted Cole in his final two years with the team.