February 24, 2017
Mr. Sale's Arithmetic Class
I think it's something close to common knowledge that Chris Sale pitched to contact last season. Sale talked about it. White Sox pitching coach Don Cooper talked about it. It showed up pretty plainly in the numbers, as Sale racked up a career-high in innings pitched and led the American League in complete games. The idea was a simple one: Sale is good enough to get batters out at an elite rate even without striking them out at an elite rate.
It was true, too. Sale fanned just a hair over a quarter of opposing hitters, instead of nearly a third of them. He posted the worst FIP of his career by a fair margin and gave up more home runs (27) than he ever had. Meanwhile, opposing batters reached base on balls in play at the lowest rate of Sale's starting career. All told, DRA and cFIP do indicate that Sale was slightly worse last season than in the two previous campaigns, but the extra innings he was able to pitch offset that and made him more valuable.
Of course, that alteration in mentality required a well-calibrated change in actual approach. Sale would need to be around the strike zone more and rely less on his secondary pitches. He did both of those things, bumping his Called Strike Probability from 0.470 to 0.493 (or from the 46th percentile to the 90th, in terms of sheer control), and raising his four-seam fastball usage rate to 45.3 percent, the highest it has ever been.
When you throw fastballs in the strike zone, though, you tend to get hit hard. Sale's fastball is an exceptional one—it had the most extreme horizontal movement and the most sink of any four-seamer thrown at least 1,000 times by a given hurler last year—but that didn't mean he would necessarily be an exception to that rule. Sale, therefore, made a small adjustment. He started throwing his four-seamer at almost uniquely variable speeds.
He still threw 34 pitches 97 miles per hour or harder, which is more than many guys who threw harder than Sale on average—including Vince Velasquez, Taijuan Walker, Jake Arrieta, and Jacob deGrom. Obviously, throwing consecutive fastballs is not a plan designed to generate swings and misses, or even a great many called strikes. Most of our pitch tunneling data isn't relevant to it. But when it comes to changing speeds on that pitch, there's something to see.
Our database shows 343 pitcher seasons since 2008 during which a hurler followed a four-seam fastball with another one at least 500 times. Only one of those (Danys Baez's 2009 season) averaged more velocity separation on consecutive heaters than Sale had in 2016. Both pitchers averaged .0063 seconds of flight time difference on consecutive fastballs, by the time they reached the tunnel point. That sounds minuscule, but it works out to perhaps a mile per hour. It's also considerably more than almost anyone else achieves. Of the aforementioned 343 pitcher seasons, in only 14 did hurlers average even .0050 seconds of separation on back-to-back heaters.