Julius Randle experienced a career turnaround unlike many seen in NBA history this season… actually, scratch that past tense. Randle is still evolving, and adding new ways to score the ball from deep as this season goes on.

Sigh… When will we learn not to put limits on Julius Randle’s game; not to put a smug fingertip on a far-from-static ceiling; not to prematurely close the book on basketball can and basketball can’t.

We foolishly thought this version of Julius Randle — the consensus frontrunner for the Most Improved Player award this season — had finished getting better. We thought he’d put in his summer of work, fueled by his previous season’s failures, loud and public and damning failures, and cashed in his chips with a neatly packaged career year. We thought Terminator Julius was boxed up and on the shelves, the final product, Randle 2.0, with upgraded All-NBA gadgetry.

We thought he was the most improved: past tense.

We were wrong. 2020-21 Randle doesn’t exist. Julius in December isn’t Julius in February, who isn’t Julius in May. The magic of the offseason work didn’t stop back on Dec. 23, when this fairytale season started. Just like the New York Knicks team he’s the nucleus of, this 26-year-old bowling ball in ballet shoes is somehow still improving in real and remarkable ways. Month by month, he’s explored and mapped the untamed fringes of his offensive world, and no single skill demonstrates this continual development better than the most in vogue single skill in the NBA universe: pull-up 3-point shooting. 

Randle’s pull-up 3-point shooting attempts, per game, by month, this season: 

December: 0.4
January: 0.6
February: 1.3
March: 1.8
April: 1.9
May: 3.1

Only 30 players in the league average more than 3.1 pull-up 3-point attempts per game this season, and they are a who’s who of elite NBA guards. Almost exclusively a list of point and shooting guards, besides the category-shatterers — LeBron James, Luka Doncic — and a few pre-programmed bucket-getting forwards — Jayson Tatum, Kevin Porter Jr., Paul George — it’s all pocket-sized ball dominant sorcerers. This prestigious guest list of volume pull-up threats shines a light on, and is testament to, just how impressive Randle’s in-season development of the near-toughest shot in basketball is: for a player historically classified as a big, suddenly playing like a wing, and evolving into a bona fide category-shattering positional everything.

He didn’t have this shot in January. Now, in May’s nine games, he’s launching more self-created threes per game than Bradley Beal has on the season.

He’s hitting them at a pretty decent clip too: 40% on 110 total pull-up attempts. Only eight other players are shooting more than 40% on more than 110 attempts, and they’re worth listing:

Steph Curry, Mike Conley, Paul George, Khris Middleton, Joe Ingles, Jrue Holiday, Shai-Gilgeous Alexander, and Evan Fournier.

Not bad company for a guy who took just 41 pull-up threes last season and shot an arctic 14.6% on them. Now, one miraculous metamorphosis later, he’s letting them fly: launching 29 in April, and 31 in only nine May games. 

As if this wasn’t enough — acquiring an off-the-dribble long ball and knocking it down at an elite clip in the middle of the damn season — Randle has leveled up within his level up: adding step-back threes, too, to his burgeoning offensive bag. Step-backs are to pull-up shooting as NASA’s best and brightest are to your high school physics professor. A premium NBA skill, a statement of rarefied belonging, a marker of superstardom: and Julius is making them look easy.

He’s hit 16-40 (40%) step-back triples this year. Again, this was a gradual, in season, drip-by-drip development: zero attempts in December, one attempt in January, seven in February, another seven in March, 11 in April, and then 14 in May’s handful of games. 

Drip. Drip. Drip. Splash.

This is a shot reserved for true NBA giants, and it’s notable that Randle has attempted 28 of his 40 step-backs since making his first career All-Star game in early March. Maybe he got a memo that weekend: a license to launch, the confidence to cook, and a sudden notion to keep pace with some of his compatriots from the exhibition of twinkly things. Here are your 2020-21 Step-Back All-Stars:

Bradley Beal 16-52 (30.8%); Kyrie Irving 10-27 (37%); James Harden 77-199 (38.7%); Zach LaVine 28-62 (45.2%); Jayson Tatum 41-107 (38.3%); Steph Curry 60-124 (48.4%); Luka Doncic 92-258 (35.7%); LeBron James 17-43 (39.5%); Kawhi Leonard 12-28 (42.9%); Devin Booker 13-38 (34.2%); Damian Lillard 63-170 (37.1%); Donovan Mitchell 25-67 (37.3%); Chris Paul 15-37 (40.5%); Mike Conley 23-52 (44.2%); Paul George 24-48 (50%); Jaylen Brown 6-18 (33.3%); and, of course, as just mentioned, Julius Randle 16-40 (40%)

Needless to say, last season’s Randle didn’t have this shot, with seven total step-back attempts in 64 games.

It makes sense that a player now being talked about as a fringe MVP candidate, after last season’s fart of a campaign, would start to see video game Randle as a tangible possibility: “stars take these shots; I am a star, so I can take these shots.”

Step-backs are the reigning cheat code in the NBA, but they’re a relatively new phenomenon. They are the analytically rebooted offspring of the fadeaway — the long-time king shot of superstar isolation scoring — painstakingly mastered, indiscriminately weaponized, and ruthlessly wielded first by Michael Jordan and then by his spiritual successor Kobe Bryant.

This season, all season, Randle has run amok in the midrange: especially with fade-aways, a shot he’s leant on more than any other player in the league, going 79-176 (44.9%) on the year.

From this campaign’s crop of All-Stars, only six other players have taken more than 100 fadeaways: LeBron James (46-103, 44.7%), Jayson Tatum (47-151, 31.1%), Luka Doncic (63-110, 57.3%), Kawhi Leonard (52-124, 41.9%), Devin Booker (51-107, 47.7%), and Joel Embiid (67-137, 48.9%).

The primacy of the 3-point line has certainly contributed to the relative decline of the fadeaway, but part of its dip in relative superstar usage is simply a function of it being a more difficult shot for smaller players to get to. Only a handful of players have the bulk and size to make the fadeaway a tenable go-to move, and have the ball skills and range to hit step-back threes at an efficient clip. Julius Randle, on this trajectory, in the middle of the season, is becoming one of those players.

If there is a skepticism about Julius’ frontrunner candidacy for Most Improved Player, from the very few reservations floating around, it’s that his spectacular season is vulnerable to the analytics boogeyman: regression. It’s an argument rooted in the possibility that maybe this isn’t real, that he’s just hitting shots he used to miss, and that his improvement is more hot-streak than sustainable new skillset. 

The problem with this argument is that he’s not just hitting more shots, but he’s hitting shots he never used to take. More than this: he’s now taking a diet of shots — volume fade-aways, and more recently pull-up, and step-back threes — of almost unfathomable difficulty, considering last year’s reference point as not just a bad, but truly awful catch-and-shoot 3-point threat. It’s made all the more flummoxing when you consider that standstill triples are a considerably easier shot. In 2019-20, Randle took 186 catch-and-shoot threes, and of the 97 players in the league with 186 or more attempts, he was 96th in shooting percentage, at 31.2%. Knowing this, disbelief has to be the default reaction. A reaction that seems to automatically express itself as doubt, given the side-splitting absurdity of Randle’s evolution. 

To watch Julius Randle this season has been to witness potentially unprecedented developmental vertigo: leaps in high-end ability this extreme do not happen.

Of course, this critique and counter-critique is limited to his scoring, when really a bird’s eye view of his candidacy would see so much more: improved playmaking, improved defense, improved leadership. Offensive improvements generally and scoring improvements specifically will always get more attention, but we shouldn’t forget just how total his breakout campaign has been.

There I go again with the past tense. 

Correction: how total his improvement is. As in now, mid-leap, airborne, destination unknown. Yes, Julius Randle is doing things this season he wasn’t doing last season, but he’s also doing things now in mid-May he wasn’t doing two months ago. His improvement isn’t complete: it’s floating through NBA space, it’s caked in possibility, it’s cruising far above the obliterated smithereens of a prematurely assumed ceiling.

Much has been made of the sustainability of a supposedly precarious Knick offense that leans so heavily on Randle’s Broadway-wide shoulders: how far can a team built around his game go? Which would be a reasonable question, if we knew what the endgame of “his game” is. Which — if we’d just sit down and shut up and humbly gawp at this allegedly human man’s developmentally batshit brilliance — we don’t.