Imagine a hypothetical NBA player. He brims with talent and commands the ball, leading to per-game averages of 27.1 points, 10.1 rebounds, and 8.2 assists. When he’s on the court, his team outscores the opposition by an average of 10 points per game. His true shooting percentage is higher than the career mark for sharpshooters like Klay Thompson, Damian Lillard, and Kyrie Irving.
This hypothetical player would be an All-Star, an All-NBA honoree, a strong MVP candidate. He’d generate columns and podcast discussions galore. Oh, and this player isn’t actually hypothetical. It’s Luka Doncic—in the second-year Maverick’s worst games this season.
Through a quarter of the 2019-20 season, four MVP candidates have sprinted ahead of the pack. LeBron James, Giannis Antetokounmpo, and James Harden have all won before (James four times), while Doncic is putting forth unprecedented, unfathomable numbers for a player his age.
Best Box Plus/Minus Seasons, Age 20 and Under
|Points Above Average per 100 Possessions
|Points Above Average per 100 Possessions
A new way to frame the MVP debate and appreciate their tremendous starts is to look only at their worst games, as with Doncic above. It’s one thing, especially in this free-flowing, high-scoring era, to post ludicrous stat lines against overmatched defenses; it’s another still to contribute an MVP-worthy box score on an off night.
In Michael Jordan’s 1995-96 season, for instance, when he earned 109 of 113 first-place MVP votes after leading the league in scoring and the Bulls to a then-record 72 wins, Jordan scored 24.4 points per game across his 41 worst games. Back in the low-scoring days of the mid-’90s, that would have ranked fifth among all players and first among guards—so Jordan was still the NBA’s best scoring guard even when he was at his worst.
An even better example might be Steph Curry’s 2015-16 season, when his Warriors broke the Bulls’ win record and he became the league’s first unanimous MVP. In the worst half of Curry’s games that season, he still averaged 22.8 points, 6.6 assists, 4.8 rebounds, and 2.3 steals per game, plus he still made 3.4 3-pointers per game while shooting 36 percent from beyond the arc. At that point in NBA history, only Ray Allen, Klay Thompson, and Curry himself had ever reached those shooting benchmarks in a season—and Curry matched them in his off games. Or, for further comparison, 36 percent is Harden’s career 3-point accuracy—meaning Curry at his worst still shot as well as Harden overall. Splitting his games this way only emphasizes Curry’s nonpareil dominance as a shooter.
For the leading 2019-20 MVP contenders, we find similar lessons from their worst games, as found by game score. This John Hollinger–developed stat compresses a player’s entire line in the box score into one representative number of his play that night. So if we sort all the games in a player’s season by game score and divide at the midway point, we can separate the numbers from his best games and worst games for closer analysis.
Averages from worst games: 26.2 points, 12.2 rebounds, 5.6 assists, 1.5 steals, 1.4 blocks, .548 true shooting, plus-9.6 on-court differential
Compare those worst-game stats from this season to his overall per-game stats from last season, when he won the MVP:
27.7 points, 12.5 rebounds, 5.9 assists, 1.3 steals, 1.5 blocks, .644 true shooting, plus-9.1 on-court differential
Aside from a decline in shooting accuracy, the numbers are basically all the same. Antetokounmpo opened this season playing better than ever before—a boost the player himself could have predicted, after he said over the summer he’d been at only “60 percent of [his] potential”—and now he’s doing just as well in his worst performances as he did overall last season. Indeed, the only players to average 25-12-5 over a season are Elgin Baylor, Oscar Robertson, Wilt Chamberlain, Kareem Abdul-Jabbar, Charles Barkley, and Giannis himself. In 2019-20, Giannis is reaching those numbers again overall, and he’s also doing so in his off nights alone—while also shooting better than 50 percent from the field in his worst games, because why not.
If there’s an area in which Giannis struggles, it’s other shooting categories: He’s at 26 percent from 3 in his worst games and a ghastly 51 percent from the free throw line. But those inaccuracies haven’t prevented his overall performance from shining as brightly as ever. The relatively bad version of Giannis this season is as good as MVP Giannis from before.
Averages from worst games: 27.1 points, 10.1 rebounds, 8.2 assists, 1.7 steals, 0.1 blocks, .578 true shooting, plus-9.6 on-court differential
As discussed above, Doncic is nearly averaging a triple-double in the worst half of his games. Only Oscar Robertson and Russell Westbrook have ever averaged a 25-10-8 line before, as Doncic is during his off nights.
Like Antetokounmpo, Doncic struggles from 3 in his worst games. (That’s a common theme across the board—the easiest way to lose game score points is to miss a lot of shots.) But he compensates by getting to the line 9.4 times per game, which allows his true shooting percentage—the best measure of a player’s per-shot scoring ability—to flourish regardless. Bad Doncic has a TS% of .578 this season; for context, Thompson, Lillard, and Irving are all lower in the .570s in their careers.
So Doncic at his worst still scores as efficiently as All-NBA guards and nearly records a triple-double every night. He’s been so impressive on a nightly basis that his 31-point, 13-rebound, 15-assist explosion against the Lakers last month qualifies as a worst game for him. He’s finished below 20 points in just one game all season.
Averages from worst games: 33.0 points, 5.8 rebounds, 7.3 assists, 1.1 steals, 0.4 blocks, .575 true shooting, minus-0.6 on-court differential
Harden’s top line is simple: He would still be leading the league in scoring even if only his worst half of games counted. Let that fact sink in.
Beyond points, Harden would also still be the league leader in free throw attempts and makes in his worst games, and, unlike Giannis, he shoots just as well from the line regardless of other context (87.6 percent in his best half of games, 87.5 percent in his worst). Harden’s not just a scoring machine, either, as his other averages suggest, and his true shooting percentage is nearly the equivalent of Doncic’s thanks to all those free throws. The lesson, as with all of these multidimensional MVP candidates, is a constant ability to contribute in various ways every game.
There is one key difference between Harden and the other three players on this list, though: Harden has the widest gap between his best games and worst. Moreover, the Rockets are plus-15.1 per game when Harden’s on the court at his best, and minus-0.6 per game when he’s on the court at his worst—again the largest difference; this might suggest that the Rockets’ performance is more dependent on Harden’s playing well than other teams with MVP candidates are on their stars. The Rockets are 8-2 in Harden’s 10 best games and 5-5 in his 10 worst.
Averages from worst games: 22.2 points, 7.6 rebounds, 9.8 assists, 0.8 steals, 0.5 blocks, .508 true shooting, plus-3.7 on-court differential
On the surface, James’s worst game stats lag a bit beyond those of his competitors. Bad LeBron is averaging “only” 22.2 points per game, and his shooting percentage plummets on his off nights. His true shooting percentage in those games—weighed down in part by a mere 60 percent mark on free throws—is in the same range as Rajon Rondo’s career mark, not Thompson’s or Lillard’s.
Still, just as Harden would be leading the league in points if only his worst games counted, James would be leading the league in assists. And he’s stuffing the box score regardless, as only five players have ever averaged a 20-7-9 statline before this season: Robertson, Westbrook, Harden, Magic Johnson, and James himself. In most other seasons, frankly, that line plus an elite team record (the Lakers are 8-2 on James’s off nights) would be enough for MVP favorite status.
On a broader level, it’s somewhat challenging to interpret such absurdly great statistics as this quartet has accumulated. The league’s combination of efficiency and pace has made it much easier to amass high point totals (both individually and on a team level), which in turn increases assist totals and allows creative, ball-dominant players to record the sort of across-the-board numbers previously reserved for single-game outbursts, not season-long constants.
But even that contextual caveat can’t much dim the splendor of these statistics, as each of these four players produces at an MVP level every night, even when their shots aren’t falling. Only a quarter of the season has passed, but they’re also positioning themselves for a historically great awards race. The debate will surely grow contentious—if the Giannis-vs.-Harden discourse was rough last season, imagine what will happen when LeBron and the league’s best young player enter the fray—but it’s far more productive, and entertaining, to enjoy each and every one of their performances. Even at their worst, they’re still the best players in the league.
Stats through Tuesday’s games.