Diving into Luis Cessa’s peculiar projections

For the third straight offseason, BtBS is looking at players where the projections disagree. Check out all the entries here.

Welcome back to Mediating Projections! This week, we’re taking a look at Luis Cessa. The Yankees need starters, and he was a starter in 2016 — just not a very good one. In 70 ¹⁄₃ frames, he tallied a 4.35 ERA (not terrible) and a 5.52 FIP (very terrible), making him exactly replacement-level by fWAR. Will he remain at that level, or become something more serviceable?

Well, that depends. Both systems think his ERA will be a little closer to his FIP, which makes sense. Yet they diverge when it comes to just how good that ERA, and that FIP, will be. Will Cessa sink or swim in his second year as a big-leaguer?

Since both systems project identical strikeout rates, we won’t need to analyze that. For this edition, we’ll examine the four areas of disagreement — walks, home runs, batting average on balls in play, and strand rate — and come to a conclusion about which projected outcome is correct.

Walks — 6.8 percent (ZiPS) vs. 7.3 percent (Steamer)

As a farmhand — first with the Mets, then briefly with the Tigers and Yankees — Cessa stood out because of his control. Across 591 innings in the minors, he walked 5.4 percent of the batters he faced; last season in the majors, he sustained that, lowering his free pass rate to 4.9 percent. Can he keep that up into year two?

In all likelihood, no. The average MLB pitcher threw 63.8 percent of his pitches for strikes in 2016; Cessa was actually lower than that, at 63.5 percent. That wasn’t really a function of poor receiving — Cessa’s expected strike rate, based on his PITCHf/x plate discipline data*, came in at 63.8 percent. Gary Sanchez is more of a bat-first catcher, and while he can frame respectably, he won’t steal a great deal of strikes for his pitcher. Cessa’s walk rate probably won’t stay in the five percent range, as both Steamer and ZiPS project.

*Zone rate and O-Swing rate, via FanGraphs.

But strike-throwing isn’t the only ingredient in a low walk rate. That recipe also calls for efficiency — if hitters keep fouling off your pitches, they could eventually force you to give in and throw ball four. This is where Cessa prospered last year: 33.2 percent of his strikes were balls in play, putting him nearly five percentage points above the big-league average (28.3 percent). Thus, based on the regression run by RotoGraphs’s Alex Chamberlain, he “should have” walked 6.2 percent of the batters he faced, an entirely reasonable figure.

For Cessa to stick in the rotation, he’ll need to limit walks — as we’ll soon see, he has some serious shortcomings elsewhere. And his control may be up to the task: Baseball Prospectus’s Chris Crawford wrote last year that, despite those shortcomings, “[Cessa’s] ability to throw strikes does give him a chance to pitch every fifth day.” If he can keep up the average strike rate, and above-average in-play strike rate, Cessa should live up to what ZiPS predicts.

Home runs — 4.3 percent (ZiPS) vs. 3.8 percent (Steamer)

As is so often the case, the long balls were the source of Cessa’s trouble in 2016: He sacrificed 16 of them, despite facing only 285 batters overall. On a per-nine inning basis, just four other pitchers did worse than Cessa. Whether he’ll improve from there isn’t really up for debate — it’d be hard for anyone to remain that awful. How much he’ll improve is what we’re looking at here.

While it’s tempting to look at Cessa’s minor-league numbers — in his career there, he allowed a home run just 1.6 percent of the time — and write off his rookie struggles as a fluke, that might not be the correct way to look at it. He never posted an especially high ground ball rate in the minors, living in the mid-40s and topping out at 48.4 percent in 2014. Plus, his middling strikeout rate (19.9 percent) ensured he’d have plenty of balls put in play.

What does this mean? Cessa got unlucky last year, sure — but before that, he got really lucky. He notched that minuscule home run rate because his minor-league competition was too weak to square him up; once he got to the major leagues, where hitters muscled up, he ran into some problems. His ground ball rate (43.2 percent) and strikeout rate (16.1 percent) got even lower, meaning more balls in play, and more balls over the fence.

Cessa’s four-seam fastball epitomized why he couldn’t make the transition. The pitch floundered more than anything else in his repertoire — it was responsible for 12 of his 16 home runs allowed — and it’s not hard to see why. It traveled at an average of 95.1 mph, and far too often, it landed in the heart of the plate:

Image via Brooks Baseball

In terms of both horizontal and vertical movement, Cessa’s fastball was in the middle of the pack among starters. A straight, hard heater, right down the gut, can give you some success in the minors, where the opposition can’t catch up to it. In the majors, though, hitters know how to turn on that. Cessa’s one trick — having higher velocity — wasn’t enough, making for a rather unimpressive pony.

Cessa isn’t a high-strikeout pitcher, or a ground-baller. Those characteristics, together with his bland fastball, predispose him to a high home run rate. Combine them with one of the most home run-friendly stadiums in baseball — particularly for lefties, whom a right-handed pitcher like Cessa will face a ton of — and you have a hurler who gives up taters, to the tune that ZiPS expects.

BABIP — .293 (ZiPS) vs. .286 (Steamer)

Where Cessa couldn’t prevent round-trippers last year, he had no issues with the hits that stayed in the ballpark: His .233 BABIP was the third-lowest in the majors. Obviously, a great deal of luck went into that — remember that whole thing about how pitchers don’t have much control over balls in play — so it wouldn’t be wise to expect him to repeat it. Still, we should wonder if he can maintain a sub-average BABIP, or if he’ll regress toward mediocrity.

His track record would suggest the latter. Cessa always had a high BABIP in the minors; for his career, he was at .316. The nadir came in 2015, when he gave up a .358 BABIP across 139 ¹⁄₃ innings at Double and Triple A. Hitters couldn’t knock the ball out of the yard, but they could definitely find a hole.

Why didn’t that transfer to the Show? It certainly wasn’t because Cessa progressed. Indeed, batters made a lot of solid contact against him last year, whether on a ground ball, fly ball, or line drive:

Cessa’s ground ball rate wasn’t especially low, nor was his popup rate (2.3 percent) very high. While he managed to stay on the BABIP dragons’ good side, they’re notoriously fickle creatures; this year, they could turn on Cessa, just as they did during his minor-league days.

The evidence with regards to BABIP is pretty overwhelming. The hard contact plagued Cessa in the minors, and it stuck around in the majors, although his production didn’t reflect that. Unless Cessa completely overhauls his arsenal or his approach, the hits will start coming, and they won’t stop coming. Even ZiPS might be a little too optimistic here.

Strand rate — 68.9 percent (ZiPS) vs. 69.8 percent (Steamer)

Cessa had a formidable strand rate last year, placing in the upper third of the majors with a 76.8 percent mark. And as with his BABIP, Cessa had a history of underperformance here: In the minors, he stranded 69.2 percent of his baserunners. So he won’t remain in the mid-70s — but how far will he plummet?

Given the poor quality of his peripheral metrics, I’d say pretty far. For one, Cessa didn’t limit stolen bases in 2016 — runners took off in five of their 77 opportunities* against him, which was well above the 5.3 percent major-league baseline. Even with the powerful arm that Sanchez packs, a right-handed pitcher without a strong pickoff move won’t keep runners in place.

*From Baseball Reference: “Plate appearances through which a runner was on first or second with the next base open.”

And more importantly, Cessa fared poorly out of the stretch. In basically every area, he melted down with runners on base:

Most hurlers will get a little worse once someone has reached — in 2016, the average pitcher had a .315 .wOBA with the bases empty, and a .322 wOBA with runners on. To decline by this much, however, suggests a deeper problem. Some pitchers, such as the Angels’ Ricky Nolasco, simply can’t perform as well from the stretch, which caps their clutch ability. Although he hasn’t accumulated much service time in the majors, his record in the minors suggests that could be the case for Cessa.

Somehow — most likely through leaving runners on base for a reliever to put away — Cessa put up a high strand rate in 2016. Given the way he played in 2016, and in the years before that, I foresee a heavy dropoff. As with BABIP, this is an area where ZiPS could end up being a little too rosy.

The total: four victories for ZiPS, zero for Steamer. While that doesn’t tell the full story — in one of those areas (walks), ZiPS took the upside — it nevertheless paints a rather gloomy picture. As a minor leaguer, Cessa compensated for a subpar BABIP and strand rate by suppressing home runs; as a major leaguer, the former two won’t get much better, but the latter will get worse. The Yankees should look elsewhere (perhaps Chad Green?) for their fifth starter.

Ryan Romano is the co-managing editor for Beyond the Box Score. He also writes about the Orioles for Camden Depot, sometimes. Follow him on Twitter if you enjoy angry tweets about Maryland sports.