What’s up with Chad Green’s fastball?

From 2017 to 2019, Chad Green was one of the most consistent arms out of the Yankee bullpen, and one of the better relief pitchers in baseball. Across those three seasons, Green produced the sixth-most reliever fWAR (5.2) in baseball, just a spot behind his current teammate, Aroldis Chapman. In almost 200 innings of work, Green maintained an ERA of 2.67, backed up by a FIP of 2.52 and an xFIP of 3.09.

Green owned the top of the zone with an overpowering, high-spin fastball, complimented by a solid slider. Perhaps most indicative of Green’s reliability was his 29.0% K-BB%, the fifth-best mark in the majors between ‘17 and ‘19, and proof of his ability to prevent batters from putting the ball in play while also limiting free passes as well as almost any other bullpen arm in the game.

Green’s overall effectiveness was buoyed by that K-BB%, but batters still managed to mash his fastball if they were capable of just putting it in play. Like many pitchers who aim for strikeouts with fastballs at the top of the zone as their primary weapon, the pitch can result in a blend of exclusively high-risk and high-reward outcomes: strikeouts and extra base hits (a pattern Yankee fans should be familiar with from watching Gerrit Cole). In each of the aforementioned trio of seasons, Green’s Barrel%, HardHit%, and average exit velocity were often around the very worst in the majors.

Of further interest: It seemed as though batters might have begun to grow accustomed to the rising effect of Green’s four-seamer, as their xSLG against the pitch crept up from .261 in ‘17, to .391 in ‘18, and to .429 in 2019. Nonetheless, Green generated the most wFB of any pitcher in those seasons, with his massive production coming from an extremely high usage (as much as 86.5 percent of the time in 2018), and an ability to limit batted balls via the strikeout.

Then, in 2020, Green ditched his horizontally breaking slider for a slurve/curve with greater North-South action, which appeared to improve the deceptive efficacy of his already awesome fastball. Signs indicated that Green’s two pitches with diametrically divergent breaking actions complemented each other better than ever before. Although he pitched less than half the number of innings of all prior complete big league seasons, they were qualitatively by far the best of his career. Despite having allowed 3.51 earned runs per nine innings pitched, Green’s batted ball data predicted an earned run allowance at a rate of 2.07 per nine.

The most striking departure from the trends across Green’s career to date was the suddenly suppressed quality of contact from opposing hitters. In 2020, Green’s HardHit% fell to a career-low 28.1 percent. After being within the bottom three percent of the majors in each of the prior three seasons, his 2020 mark rocketed him up to the league’s eighth percentile. At the time, Green was so dominant that I argued that he had the best fastball on the Yankees, even more so than that of either Gerrit Cole or Aroldis Chapman.

While it’s possible that Green’s new breaker caught Major League Baseball off guard, it’s all the more likely that the truncated schedule of the 2020 season resulted in some unrepeatable, wonky stats. Further, Green’s xwOBAcon of .293 was more than 100 points better than each of his past two seasons of below league average performance, putting him among the league’s top percent in that category — a result that reeked of unsustainability.

In 2021, the regression monster’s sword has cut Green both ways. He’s allowed far worse contact than last season, but has a far superior ERA of 2.62. After underperforming his xERA by almost a run and a half last year, he’s outperforming it by almost a whole run this season. This comes coupled with his career-high Barrel% of 15.1, way worse than his previous high water mark of 11.6 percent, and almost double last season’s rate of 8.3 percent. Most strikingly, on average, Green’s four-seamer has been hit a whopping five miles per hour harder than it was last season.

Across the board, Green’s pitches grade out as identical to those of last season, approximating the break, spin, and whiff rate of last season despite having resulted in a radically different batted ball profile. For those perplexed by Green’s handful of contradictory metrics, the odds are good that he’ll continue to be as steady as he was between 2017 and 2019 — i.e. one of the best relievers in baseball by production and volume — even if he’s not one of the game’s absolute best like he was in 2020.