clock menu more-arrow no yes

Filed under:

A Robbie Erlin postmortem (that no one asked for)

New, 31 comments

No, I’m not forcing you to read about Robbie Erlin

MLB: Atlanta Braves at Boston Red Sox Bob DeChiara-USA TODAY Sports

The 2020 Atlanta Braves are playoff-bound, despite a series of starting pitching experiences that are aptly described as catastrophic. One of those catastrophes: 29-year-old southpaw Robbie Erlin, who made nine appearances (five starts) for the Braves and compiled an unfortunate -0.2 fWAR in the process.

Especially in this shortened season, the temptation to rush to judgment is heightened. If you look on the surface, Erlin was a prime-tier castoff: he was let go by the Padres after a -0.5 RA9-WAR / 126 ERA- season, found his way to the Pirates, and lasted all of two appearances with them. Combine that with the hard-to-dispute reality that he was awful in a Braves uniform, and, well, there’s an easy, satisfying narrative to lean on: Erlin sucked, the Braves signed him to be a warm body, and no surprise, he still sucked. Then he got cut on September 14. Easy peasy. As usual though, easy narratives might be comfortable, and they might even satisfy the basic pattern recognition mandates of our lizard brains… but they’re not very interesting. And that’s why, even though Erlin had little to do with the success of the 2020 Braves, and won’t have anything more to do with that success, this is an article about what went wrong with Robbie Erlin, 2020 Atlanta Brave.

First things first — let’s sucker-punch the starting part of that easy narrative above. In 2019, yes, Erlin’s ERA was elevated. But, like many before him (and many behind him), he was done in by the dual daggers of BABIP (.373) and strand rate (65.3 percent). He had an 82 FIP- and 93 xFIP- in 2019, and even if you want to hang your hat on having to take contact into account, his .335 xwOBA-against was mediocre but not terrible. Overall, Erlin definitely presents a weird case, as he’s kind of an anti-2020 throwback, a soft-tossing lefty who is way less about hyper-specialization than he is about taking the hill whenever he’s needed and pitching to contact with a fairly slow “fastball.” Still, that profile didn’t directly impede his success — by the time the Padres cut him, he had 3.9 fWAR in under 400 innings (mostly in relief, but 38 total starts too) with a 91 FIP- and 93 xFIP-. Yes, the ERA was elevated (120 ERA-), but that was in large part due to strand rate shenanigans — we could have a discussion about whether Erlin actually has some kind of issue pitching with runners on that will enforce an ERA-FIP gap for him, but that’s not what this post is about, and I’m not sure how to go about having that discussion anyway.

Alright, on to actual Erlin-in-the-present things.


During the 2019-2020 offseason, well before we truly knew what 2020 had in store for us, Michael Augustine wrote a pretty cool article about Erlin: https://blogs.fangraphs.com/getting-the-most-out-of-robbie-erlin/. Augustine’s question was a pretty basic one that any team assessing a player probably asks: “So what value can be drawn from Erlin?”

I’m not going to go over the entire article; you should just read it. (It’s cool!) Instead, what I want to focus on are Augustine’s recommendations for, well, getting the most out of Robbie Erlin, and whether they were implemented.

The fastball — throw it up in the zone

In short, here’s what Augustine highlighted about Erlin’s fastball:

  • Erlin kept it low in the zone in 2019, and low is not where you want to go with your fastball.
  • Erlin’s fastball isn’t very fast, but it has decent spin, and decent “rise.”
  • Therefore (perhaps obviously), throw the fastball up in the zone.

Here’s Erlin’s four-seamer location in 2019:

Here’s Erlin’s four-seamer location in 2020:

He threw his fastball up in the zone. Check. In fact, this is more than just a “check.” It might be overboard. In 2020, only two pitchers have thrown more fastballs (proportionately) around the upper edge of the zone than Erlin — Nick Anderson (a reliever) and Sean Newcomb. No pitcher with as many pitches thrown as Erlin has come anywhere near his 40 percent rate; Blake Snell is the closest and he’s three percent behind.

This is part of a different pattern, but I’ll just enumerate it here: there were some clear things (in this case, throwing the fastball up) that Erlin could do to increase his chances of success. He did them.

Stop throwing the two-seamer

Augustine noted that Erlin’s two-seamer was pretty similar to his four-seamer, and didn’t really do anything in and of itself. Combine that with the general league-wide trend against two-seamers/sinkers (for good reason), and this is another no-brainer.

(Note: Augustine doesn’t say this, but one of the specific issues with Erlin’s two-seamer is that it doesn’t actually sink, possibly because of its spin. This effectively gives him two “rising” fastballs; moreover, he didn’t really even start throwing the sinker until 2018, partway through his career, which was definitely not a great time to pick it up.)

This is an easy check. Erlin’s pitch mix in 2019 featured 21 percent sinkers. His pitch mix in 2020 featured either zero sinkers or just a handful (probably just pitch classification errors). Done and done.

Throw the changeup in the zone

Augustine provides a lot of good information about Erlin’s changeup. In brief, the changeup can definitely be a weapon with a high whiff rate, though it sometimes gets hit hard, making it a bit tricky to use. Augustine noted that in 2019, very few of Erlin’s changeups hit the zone, which may have been intentional — but it was a better pitch in 2018, when it was consistently thrown on the outer part of the plate to righties, rather than missing low a bunch:

2018, changeup to righties:

2019, changeup to righties:

2020, changeup to righties:

If you prefer numbers rather than visuals — well, this one is a little ambiguous, but there was progress.

  • If you go by a hard-and-fast in/out measure, Erlin’s changeup hit the zone about 37 percent of the time in 2020, better than 2019 (36 percent) but much worse than 2018 (46 percent). (Note: if you use Brooks Baseball, as Augustine does, knock the percentages down around seven percent, but the pattern is the same.)
  • If you include borderline pitches as well, 2020 featured 61 percent of changeups in the vicinity of the zone, compared to 54 percent in 2019 but 68 percent in 2018.

So, not a full check, but at least there was progress. Two-and-a-half out of three ain’t bad.

Raise the release point

This is another partial credit one. Augustine noted that Erlin’s release point fell substantially from 2018 to 2019; while it’s hard to directly relate vertical release point to specific outcomes, it’s still interesting that Erlin was a super-good swingman (1.6 fWAR, 84 xFIP-) in 2018 with a higher release point, and trended down as it fell.

Like the changeup zone rate, only very partial gains were made here. Really, even “partial” kind of oversells it. On the changeup, Erlin’s relative release point was 5.84 feet off the ground in 2018, dropping to 5.68 feet in 2019, and not really changing (5.69 feet) in 2020. A by-game analysis for 2020 indicates some variation, but nothing temporally — in other words, while he had some starts closer to the average 2018 release point in 2020, they were scattered with others where they were closer to 2019, or otherwise pretty low.

Is this two-and-a-half out of four? Maybe three out of four, if we’re awarding partial credit? The point is, there was a clear path for Robbie Erlin to improve. Maybe it wasn’t the only path, maybe it wasn’t even the right path. But there were changes, and he and/or the Braves tried to make them a reality.

So, what happened?

Augustine concluded his article with these words:

“With these suggested pitch adjustments, their usage, and how he deploys his arsenal, Erlin could end up being a free agent steal in 2020.”

Robbie Erlin was not a free agent steal in 2020. He lasted all of two outings with the Pirates. The Braves snapped him up, and his very first pitch as a Brave resulted in a grand slam. (I wonder whether that’s ever happened in history — a player’s first pitch with a team being a grand slam.) At the time, there was a joke that Erlin didn’t really have anywhere to go but up at that point (that joke works doubly well when you consider that he threw so many pitches high), but… that didn’t really happen either.

  • In that same outing, Erlin ended up allowing two other homers.
  • He then had a brief oasis-type start in which he threw four scoreless, one-hit, zero-walk innings against the Marlins
  • ...but then gave up four longballs over his next three starts, with a 9/6 K/BB ratio.
  • Then came a relief appearance with no homers somehow, which also happened to be his first outing of the year in which half the balls in play against him were on the ground.
  • But then, he got just five outs while allowing five runs to the Nationals (yes, a homer was involved), and that was that.

Erlin’s totals as a Brave: 23 ⅓ innings, -0.3 fWAR, 187 ERA-, 152 FIP-, 119 xFIP-. The only “steal” he constituted was one of win expectancy — his time on the mound featured the Braves losing 0.71 WPA (only Touki Toussaint has cost the team more among pitchers).

There are a lot of possible answers to “what happened.” One such answer: it was 23 innings! Erlin made some pretty big changes to who he was as a pitcher; it wouldn’t be surprising if working out all the kinks takes longer than a few appearances. But, the 2020 Braves were probably not the place to do that kind of self-diagnosis, anyway.

Here’s another answer. Everything Augustine wrote, and Erlin’s changes, were predicated on a simple premise of: rising fastball good, dropping curveball good, put them together, (that’s) kablamo! And, as we can see by Erlin’s pitch fundamentals in 2019, this idea made perfect sense:

The spiky circles are league-average motion; you can see that despite weak velocity on the four-seamer, it still “rose” (didn’t fall) more than average, while the curveball got a lot of vertical drop without losing much horizontal movement. You can also see why the sinker was a bad idea (why is it “rising” despite being super-slow, Robbie, why?!), why the slider/cutter is a terrible idea, and why the changeup really needs to be spotted well — it has pretty weak “depth.”

The thing is, sometimes you put together a gameplan based on a status quo… and then it shifts under your feet. It’s a shame, but it is what it is. Or, in other words, same chart, now 2020-flavored:

Imagine re-gameplanning your arsenal around a rising fastball that A) lost around 1.5 mph on its already nearly-untenable velocity and B) is no longer really rising nearly as well as it did before. Now imagine still making pitches as though it were, throwing a generic (one might say “flat”), slow fastball up in the zone. That’s a good way to tempt whatever baseball’s version of The Fates are, and it didn’t go well.

Are there lessons to be drawn from this? Maybe, maybe not. After all, it was just 23 innings. But, if one were so inclined, I guess we could file this one under “when making changes deliberately, make sure those changes still work in the context of other, inadvertent changes that happen at the same time.” Boy, that’s a mouthful. I yield the floor for a pithier summation.