Blake Snell Tries To Become Extraordinary Again

On Friday night in San Diego, the Padres got an all-too rare glimpse of the version of Blake Snell that they hoped they had traded for in December. Facing the Mets — admittedly, an injury-weakened version of the team, but one that has nonetheless led the NL East for the past four weeks — Snell made his longest start in over two years and threw seven innings of one-hit shutout ball while striking out 10, that after being battered for 12 runs in 6.2 innings over his previous two starts.

Facing a lineup that was without Michael Conforto, J.D. Davis, Jeff McNeil, and Brandon Nimmo due to injuries, and that contained just two regulars with a wRC+ of 100 or better (this qualifier should sound familiar), Snell retired the first 14 Mets he faced before walking Billy McKinney, one of the lineup’s two lefties. He didn’t yield a hit until Francisco Lindor led off the seventh with a single to left field that bounced past Tommy Pham and went to the wall as Lindor took third, a two-base error that put the tying run 90 feet from home. Undaunted, Snell struck out James McCann, retired Alonso on a pop foul, and then struck out Brandon Drury on three pitches to complete a stellar night’s work. The Padres held on to win, 2-0.

Snell’s 101 pitches were the most he’d thrown in a single start this season; he’d gone 95 twice and 97 once, but hadn’t reached 100 since he threw 108 last September 22 against the Mets, one of three times he reached 100 pitches. In terms of innings, Friday’s outing was Snell’s longest since April 2, 2019, when he threw seven shutout innings against the Rockies while allowing just two hits. He didn’t complete six innings in any of his 11 starts last year, and had done so just once this year, on May 18 against an even more moribund version of the Rockies.

In fact, Snell has made it through five innings in just six of his 12 starts as a Padre, and has averaged just 4.50 innings per turn. Among pitchers with at least 10 starts — a crude proxy to select for those whose health concerns are comparatively minimal — only five pitchers have worked shorter:

With the numbers from Friday’s start included, Snell has been just about the most effective of the bunch save for Paddack, his teammate, and Dunning; prior to that gem, his 5.55 ERA, 4.42 FIP, and 0.4 WAR had more in common with Taillon or Keller.

Still, Snell’s performance hasn’t been anything close to the standard he set during his 2018 AL Cy Young-winning campaign, when he led the AL with a 1.89 ERA and placed fifth with a 2.94 FIP. It’s been short of the mixed bag he produced in 2019-20, a stretch that included his missing eight weeks in the earlier season due to in-season surgery to remove loose bodies in his left elbow, and throwing just eight innings over his first three starts of the pandemic-shortened season after being brought along rather slowly. In that two-year span, he pitched to a 3.96 ERA and 3.65 FIP and totaled 3.3 WAR, that while averaging just 4.62 innings per start. Kevin Cash’s decision to pull Snell in the sixth inning of Game 6 of the World Series, while he had shown his dominant form, may live in infamy, but it was hardly out of character with the way he has recently been handled. Once he’s been though the order twice, he’s on borrowed time; indeed, since the start of 2019, Snell’s 19.2 batters faced per start is the lowest of any pitcher with at least 150 innings as a starter.

Both the Rays and the Padres have kept Snell on a short leash of late for a couple of reasons. First, he’s inefficient, burning through his pitch count (such as it is) by going into deep counts more often than just about any pitcher. And second, his performance against righties has deteriorated, a problem given that over three-quarters of the batters he’s facing are righties.

Snell continues to strike out a ton of hitters; his 33.2% rate ranks sixth among NL starters with at least 50 innings, behind only Jacob deGrom, Corbin Burnes, Freddy Peralta, Max Scherzer, and teammate Joe Musgrove. Yet his 13.3% walk rate isn’t just a career worst (up from about 9% in each of the past three seasons), it’s the majors’ second-highest mark behind only John Gant’s 15%. Among all major league starters, only deGrom and Peralta are battling hitters to the point of either a strikeout or a walk more often than Snell’s 46.5% of all batters faced, and both are striking out a larger share of them. Only three pitchers are averaging more pitches per plate appearance than Snell’s 4.24 (plus two in a virtual tie). That’s not great, and ultimately, since he’s putting a lot of guys on base, he’s making a lot of higher-stress pitches. His 85 pitches per start will tire out an observer more quickly than someone else’s 100, that’s for sure.

As for the righties, here’s a look at Snell’s year-by-year platoon splits:

Snell’s faring worse against righties, and better against lefties, than ever; it’s a good thing that he’s facing fewer of the former and more of the latter than ever, because otherwise his overall numbers might be worse. Lefties actually hit him pretty hard in 2019 and ’20, to the point that he had a weird reverse split in the former year; the fact that he’s still facing more lefties than ever may reflect the way teams view that run of success against him, still believing their lefties can hit him. Given the sample sizes in play, that’s not unreasonable.

The performance gap between righties and lefties against Snell gets us into his repertoire. Like most pitchers, Snell uses a different pitch mix against batters of each hand. Lately against righties he has used a four-pitch mix of his fastball (43.9%), changeup (22.2%), slider (18.1%) and curve (15.7%), while against lefties he’s gone fastball (54.9%), slider (33.6%), curve (11.5%). Distilling some general trends:

• His slider remains a devastating pitch against batters of both hands, and its usage has increased from a low of 6.7% in 2019 (when it was pummeled for a 409 wOBA) to a high of 21.7% this year, when hitters have managed just a .155 wOBA. This year’s 44.8% whiff rate on the pitch, while representing a career low, still puts him in the 80th percentile, and his -5 run value on the pitch in the 92nd percentile.
• His curve, though used less often than a few years ago, remains very effective against righties (.231 wOBA this year) but doesn’t get much play against lefties (just three batted balls this year and two last year); its 45.3% whiff rate this year places him in the 87th percentile, but its run value is just average.
• His changeup, which now gets used exclusively against righties, has been pummeled for a .418 wOBA this year, up from .272 in his Cy Young season; its whiff rate has dropped from 31% to 22% in that span (it peaked at 46.7% in 2019), and its run value has swung from -10 in 2018 to + 5 this year. Eep.
• His fastball — despite remaining fairly consistent in velocity (95.8 mph in 2018, via Statcast, and 95.5 this year) and spin (2,365 rpm in 2018, 2,439 this year, but even higher in ’16 and ’17) — has gotten hit increasingly hard over the past two seasons. Where batters produced wOBAs of .312 and .305 in 2018 and ’19, they were up to .465 last year and are at .412 this year, with slugging percentages above .600 in both seasons. The pitch’s value has swung from -8 runs in 2018 to +6 last year and +2 this season.

While the righties’ meaty actual stats against the heater have been supported by their Statcast expected stats in both seasons, that’s not the case with lefties:

The samples are small enough within these subsets for some randomness, particularly against the lefties. His xSLG against lefties here is consistent from year to year, but the results have not been; Snell’s apparently gotten away with some hard-hit balls that have turned into outs. That said, he’s gotten a consistent whiff rate on the pitch, and has gotten more putaways with it this year.

The pitch is getting thumped by righties, though, where it wasn’t before (.242 AVG/.379 SLG in 2018, .230/.360 in ’19), and I think a good amount of the issue comes down to location. Here are the year-by-year heat maps:

It appears he’s leaving too many fastballs in the middle of the zone this season, and he’s paying the price. He’s getting less horizontal and vertical movement on the pitch than he did in years past, so maybe it has become too predictable. All that’s happening while his changeup gets hit harder, too, and his curve has become somewhat less effective.

Anyway, that’s one view of things. Last week, the Pitcher List’s Michael Ajeto made the case that Snell’s arm slot had dropped from overhand to a more traditional three-quarters slot before the 2020 season, and his curve has suffered by becoming “sweepier” — getting much more horizontal movement and disrupting the complex interplay between the two pitches. That certainly seems plausible, as it’s rare for a single pitch type to be the sole cause of a pitcher’s woes.

On Friday, the interplay was just fine, as Snell set a season high with 10 swings and misses on his fastball (four more than in any other start) and had his second-highest total of called strikes on the pitch (13); meanwhile, his four swings and misses on the curve was his second-highest total as well. Yes, the pitch he allowed for a hit against Lindor (swinging righty) was a fastball the middle of the zone, but he also got plenty of called strikes, swinging strikes, and fouls in that area:

One start isn’t enough to tell us whether Blake Snell has solved his problems or earned a longer leash, but Friday offered a promising step forward. We’ll see where it takes him.