The Diamondbacks Bullpen Has A Breaking Ball Problem

Yes, the Diamondbacks bullpen is a very hot topic. Jack wrote yesterday about how the bullpen has regressed and (in this author’s words) very much overachieved in the first half of the season. Last week, it was noted that Archie Bradley may not be as good as he appears due to his lack of elite strikeout numbers (or even a second pitch). Brad Boxberger has fallen off a cliff. Et cetera, et cetera.

And as such, this is baseball and it must mean there is data that might be able to help explain what’s happening. And sure enough, it paints a pretty clear picture and one that should have been seen coming all season long.

---

Stats like BABIP, HR/FB%, and LOB% (among others) do a great job of explaining why a player (or team) did better or worse than expectations in a small sample, but they’re not terribly predictive of what’s to come in the future. From looking at Jack’s article yesterday, two things stand out as to why the Diamondbacks bullpen did so well in the first half: a freakishly low BABIP (.259) and a below-average HR/FB% (11.80%). Together, these were two of the main drivers as to why the Dbacks had the MLB-best bullpen ERA.

That ERA (2.85), mind you, was more than a full run better than their FIP (3.97). Now, there is some skill in maintaining a low BABIP (good defense, good pitching) and suppressing home runs (again, good pitching), but there was one massive, glaring flaw to the Diamondbacks bullpen.

They weren’t striking guys out.

I’m sure the strike out thing has been beaten to death. And there are other ways to successfully pitch besides striking guys out. But strikeouts are brought up so much because they are very very valuable. This is why strikeouts are a large component of FIP and ERA. In fact, K% has very strong ties to FIP:

This is narrowed down to exclusively MLB bullpen data for 2018. An R^2 of .6911! That is a tremendous correlation! K% is the single biggest factor that contributes to FIP - and from the chart, you can see there is AZ with the 9th-lowest bullpen K% in the MLB.

But maybe you’re not a fan of theoretical stats like FIP (though you should be). Well, how does it translate to ERA, e.g. the actual results thus far in 2018?

Not as strong as FIP, obviously, but the relationship is still strong with an R^2 of 0.4553. Strikeouts matter for all pitchers. The Diamondbacks starting rotation has the 6th-highest strikeout rate in baseball and has been great. The Diamondbacks bullpen has the 9th-lowest rate and we’re seeing the rails fall off extremely quickly.

So that begs the next question: what actually causes strikeouts?

Now, that is not a simple answer. But there are some things to look at. At the simplest, most basic level: pitch types. Different pitches generate whiffs at different rates - this should be fairly obvious. Fastballs generate the least amount of whiffs and breaking balls generate the most. These are generalizations but across the league as a whole, they’re very applicable.

How have the league-wide whiff rates looked for MLB relievers?

Breaking balls generate whiffs at a rate about 60% higher than fastballs. Unfortunately, there is a problem. The Diamondbacks bullpen is throwing breaking balls less than any other team in the MLB.

The Dbacks relievers are well behind the rest of the MLB (except Toronto) when it comes to throwing breaking balls. Now, some of it is self-serving: Brad Boxberger and Yoshihisa Hirano don’t have a breaking pitch and Archie Bradley barely throws his curveball. Boxberger still puts up elite strikeout numbers (32.1%), though that is partly due to his deception (5th-highest called strike rate in the game), but it does help explain why Archie Bradley and Yoshihisa Hirano don’t get strikeouts quite like you’d expect out of them.

But it’s not just that the Diamondbacks throw the least amount of breaking pitches in the MLB. Their breaking pitches have the 7th-lowest rate of whiffs in the MLB. Effectively, the Diamondbacks bullpen doesn’t have any good breaking pitches.

The usage rate makes sense due to the personnel but the lack of quality breaking pitches is mildly concerning, seeing as the idea was that the bullpen seemed to have some high quality players. However, the spin rate tells a different story:

Diamondbacks relievers have the lowest spin rate on their breaking pitches in the MLB. And while the relationship isn’t super strong, an R^2 of .2419 does indicate a decent relationship between spin rate and whiff% of breaking balls. Of note, the Diamondbacks relievers also rank with the third-lowest RPM of fastballs which is a bit surprising considering how much Archie Bradley and Brad Boxberger rely on their fastballs.

After a bit of digging, it appears that the Diamondbacks bullpen might have been more flawed than the first half performance indicated. Specifically, it really just seems that the Diamondbacks lack relievers with good “stuff”. Last week’s article about Andrew Chafin was, in many ways, showing that Andrew Chafin was better than perceived. However, the overall conclusion that he was just as good “stuff-wise” as Hirano, Bradley, and Boxberger. And therein lies the problem: if it can be reasonably argued that Andrew Chafin is one of your best relievers, it might be time to considering getting some more talent into the bullpen.