Yesterday, George Bissell gave a rousing introduction to our bloated American landscape of low innings totals, high earned run averages, and higher ace valuations in turn. With managers increasingly inclined to limit third-time looks, workhorse starters are becoming as rare as split-ticket voters, and an old-world strategic play of bulking up rotation back-ends with average innings-eaters may just be gone forever by the wayside. Before we get too lost in nostalgia, let’s take a look at a few guys who either over- or under-performed in the category of ERA.
Roark may be my favorite non-“elite” pitcher in the National League, if only because watching him work is like reading a textbook about how to get hitters out with underwhelming stuff. His DRA over-performance of 1.62 runs marked the fourth-widest margin of any starting pitcher last year. That’s not the first time it’s happened, either: his career marks deviate by almost a run and a half, and unlike with most guys where you see that kind of gap and reflexively assume over-performance and oncoming regression, Roark actually boasts a profile that pretty-well backs it up. He generated the fifth-most called strikes of any big-league pitcher last year, and he’s shown a consistent ability to generate weak contact – his exit velocity last year rated in the 85th percentile – that depresses his BABIP to well better-than-average levels. He even managed to bump up his whiff rate across the board last year, and I love him as one of the sneakier value plays for pretty much all formats going forward.
I mean, of course Kyle Hendricks is on this list. The NL’s ERA leader had himself a fine season to be sure, and one that was reasonably well-supported (at least in general shape) by his peripherals and underlying metrics. But a 2.13 ERA from a guy coming off a near-four effort in his first full season of starting is likely to drive some out-performance of projections, and that indeed occurred here. Hendricks isn’t actually all that different in his methodology from Roark, insofar as he steals a ton of called strikes and controls contact to generate the heart of his value. He’s just that much better at it. Batters swung less frequently at his strikes than any other pitcher’s, and his exit velocity allowed scraped elite territory in the 94th percentile. He’s also got a borderline-elite secondary pitch in his changeup driving a solid-average strikeout rate, where Roark doesn’t. Basically, if you like Roark’s profile going forward, you should freaking love Hendricks’. Neither posted fluke seasons this year despite peripheral over-performance.
Jake Arrieta, RHP, Chicago Cubs. DRA: 4.02, PECOTA ERA: 3.17, Actual ERA: 3.10
Arrieta ended up delivering pretty much exactly what PECOTA projected for him in the category, but he’s included on this list because the “how” of that delivery did not impress DRA for a second consecutive season. His bugaboo control of yore reared its ugly ahead once again, leading to his whiff and walk numbers sliding in the wrong direction. He got less grounders and gave up more homers. His contact rate crept up, while his chase rate clenched up just that much. His arm slot migrated north a bit this year, and it appeared to take some of the bite out of his pitches, particularly the slider. He managed to overcome at least a small part of this deficit by suddenly developing a devastating changeup that led the majors in whiffs-per-swing, albeit in a still-dramatically-underrepresented percentage of his pitches. If Arrieta continues to struggle with slider consistency, it’ll be interesting to see if the change starts becoming a bigger part of his arsenal in the future.
How about another round of applause for PECOTA, eh? About 30 innings light on the innings total, as is PECOTA’s want when projecting rookies, but on point with the topline ERA production. Snell’s production is a case study in rookie pitcher volatility, however, and anyone who rostered his 1.61 WHIP can tell you a story or three about the price of admission for his whiff and ERA contributions. Everything Snell throws more or less works north and south: his four-seamer and change both have loads of backspin that limits lateral movement, while both his breakers feature well above-average drop with below-average horizontal action. There’s enough velo and one-plane movement to generate a good bit of swing-and-miss, but big-league hitters are pretty good at making adjustments and I don’t think it was necessarily an accident that Snell’s numbers got consistently worse once batters saw him for a second time.
Conley entered the season as one of my favorite low-end targets for average innings (you know, the kind of guy we’re talking a lot about phasing out from your target lists) and a chance for more. Well, instead he got hurt, and when he was healthy he came to embody perfectly our nascent trend of shorter outings for starting pitchers. Hitters chased a little less, made a little more contact against him in-zone this year, and required two more pitches-per-inning off Conley relative to his rookie season. He’s averaged exactly five and third innings per start in both seasons now, and the tail end of those efforts have tended to be bloody affairs, as hitters pummeled him to the tune of an OPS north of .850 and one more walk (21) than strikeouts (20) once they saw him for the third time in a given game. Solid whiff rates with both his slider and changeup suggest some room for growth, but it’d be wise to keep expectations for Conley’s emergence as a standard-league asset in check.
I wrote a good bit about each of these guys (and several others) in this late-August piece on the very subject at hand of pitchers who were most dramatically under-performing their DRA. All four continued to muddle through underwhelming splits through September, save for Nola, whose muddlin’ days were done for the season by then. I don’t have much to add on any of them, and I’d encourage a read (or re-read) of that piece for thoughts on where the poor ERA numbers came from in each case. All of these guys were among the top-six in terms of largest DRA-ERA splits among pitchers who through at least 80 innings, and all four will likely end up popular bounce-back candidates heading into the winter, so buyers beware.
Price logged at least 220 innings for the third straight year, a feat only Max Scherzer can otherwise claim, and posted a sixth-consecutive sub-.300 DRA. Yet his actual ERA ballooned to its highest mark since his rookie year in 2009, and there’s about nine volumes worth of why to be found over in the BP Boston archives. The over-simplified punchline as I see it: he pitched behind more often than he had in the past five seasons, and batters made harder, pull-side contact as a result. His cutter and curve in particular got touched up on the regular, and his fastballs – the two-seamer especially – saw sharp declines in aggregate value as his mistakes just always seemed to get hurt more. The velocity was down a tick pretty much all year, and that ate into the cutter’s movement while re-shaping his curve in line with its pre-2015 shape. The loss of some stuff is marginally concerning here, though there’s still plenty of it and a normalized homerun rate – or even something close to it – would’ve wiped away a good bit of the stank on his season. I’d hold on Price where I’ve got him, and poke around to potentially acquire him where I don’t.
Chris Archer, RHP, Tampa Bay Rays. DRA: 2.92, PECOTA ERA: 3.16, Actual ERA: 4.02
It’s kind of eerie how close Archer’s and Price’s numbers are, and they shared a similar culprit for a good chunk of the DRA-ERA disparities in the long ball. Archer allowed 11 more of ‘em last year in 10 fewer innings than he threw in 2015. Batters all of a sudden started teeing off on Archer’s fastball, with lefties slugging almost .600 against the pitch and righties adding slugging for about 75 more points. He’s another one where his first-pitch strikes retreated all the way back after an aggressive step forward the previous year, and it rendered his strategy of climbing the ladder while ahead much less consistently practical. Still, he basically replicated a solid swinging strike rate and continued to show himself remarkably steady in controlling contact to run regular, predictable BABIPs. Especially given the strength of the slider as an out pitch, there’s plenty here to suggest the proper bet is on a home run regression and Archer’s emergence as the top-flight option his underlying metrics suggest he should already be.