Which of the Yankees’ biggest surprises is least likely to regress?

The Yankees arrived ahead of schedule in 2017. It was hardly expected they would contend for a championship, and to do so, they needed surprises; significant leaps forward in performance from players across the board. For the most part, that's what the Yankees got in order to win over 90 games for the first time in five years and nearly reach the World Series.

There were plenty of small surprises that were important, among them Jordan Montgomery and CC Sabathia stepping up as solid starters, or Starlin Castro and Chase Headley playing the role of quality veteran players. Yet a few leaps in performance stand out as the primary revelations.

Those leaps belonged to Aaron Hicks, Luis Severino, and of course, Aaron Judge. Hicks rebounded from a below replacement level season in 2016 to play like an All-Star. Severino harnessed his obvious talents in a breakout season, earning him a top-three Cy Young finish. Judge shook off an uneven 2016 debut and made good on his prodigious power in putting forth an MVP-caliber rookie campaign.

Each of these surprises was in itself a delight; a reason to turn on the TV every night when the Yankees were playing. Going forward, though, it's uncertain which big leaps in performance will stick. Some players that made strides will stride further. Others will regress a little, while still others may regress a lot.

The Yankees' three biggest surprises, Hicks, Severino, and Judge, are also their primary candidates for regression. It's hard to expect players that made such massive gains to retain said gains without missing a beat. That being said, of the Yankees' most significant surprises, which looks like the best bet to fight off regression in 2018?

The best candidates for regressions are the ones who hugely outpaced expectations, and needed to outplay their own underlying numbers to do so. Let's start there. Severino posted a shiny 2.98 ERA, but also a sterling 3.08 FIP. He struck out 230 batters compared to 51 walks. Baseball Prospectus pegged Severino's DRA at 3.05. There's no indication that Severino's jump was undeserved.

Judge also backed up his record-setting numbers with similarly eye-popping underlying figures. He led the league in average exit velocity, in exit velocity on balls hit in the air, and percentage of batted balls struck at 95+ mph, per Statcast. Moreover, by Baseball Savant's expected wOBA metric, Judge actually profiled as perhaps a bit unlucky. Factoring in his batted ball quality and even his proclivity for strikeouts, Judge's xwOBA was .446, which is 16 points higher than his actual wOBA.

Hicks' season doesn't hold up quite as strong here. His xwOBA of .335 was well short of his actual, career-high wOBA of .367. He struck batted balls at 95 mph about a third of the time, a rate that ranked middle of the pack. These figures don't give Hicks credit for his strong walk rate or defense, but it seems that he wasn't as good a hitter last year as his surface numbers suggested.

Hicks could regress towards the mean as a hitter and still remain quite valuable thanks to his ability in center field and patience at the plate, but it's clear Severino and Judge have stronger cases at this point to remain at their new levels of performance. Now, beyond underlying numbers, there are other reasons to believe the Yankees' young stars can stay at their peaks.

The best argument for Severino is that he actually got better as the season wore on, suggesting that even as the league got more looks at him, they were unable to come up with a way to beat his tremendous stuff. Severino's average fastball velocity actually peaked in October at 98.1 mph per Brooks Baseball. He posted a 3.54 ERA in the first half and a 2.28 ERA in the second half. He allowed a wOBA of .283 in the first half, and just a .239 wOBA in the second.

Judge's argument is two-fold. First, strange as it may seem, the fact that he underwent shoulder surgery last week may enhance his chances of sustaining his performance. Sure, the Yankees would rather their star not deal with shoulder ailments going forward, but it seemed obvious Judge was playing through shoulder pain as he slumped during the summer. If Judge's midseason swoon can at least partially be explained by injury, then it wouldn't be unreasonable to expect better things from him should he get healthy.

Just as important, though, is that Judge demonstrated the ability to adapt last year. Much was made of Judge's apparent weaknesses to chase at the breaking ball low and away, and at the high fastball. Pitchers targeted those weaknesses as the season progressed, but Judge countered by swinging at fewer of those pitches.

Consider these heatmaps, courtesy of FanGraphs, of his swing rate, first from his six-week post-All Star break slump, and then from his strong September finish:

Over the final month, Judge was starting to lay off when pitchers attacked his weaknesses. Not only that, Judge started hitting better in those weak areas. Here’s his slugging percentage over those same time frames:

Judge still had nothing doing when he did swing at those low breakers, an area for him to work on going forward, but he completely papered over his weakness on those upper pitches. Whereas it seemed pitchers might have a couple options to get Judge out, the down and away breaking pitch may be their best and only option now.

Thus, Judge may have been playing hurt during his slump last year, and in the season’s final stages, he flashed the ability to counter opposing attempts to exploit his weaknesses. Scary as it may seem, Judge really could replicate his performance next year if he continues to adjust and keeps crushing the ball.

Likewise, Severino showed no signs of slippage as he worked his way through the league and set a career-high in innings. It may be a toss-up between the two in terms of who has the best chance to fight off the pull of regression in 2018, which would only mean great things for the Yankees’ chances of fighting off regression as a team.