Skip to main content
Is The Stephen Curry Revolution Over Already?

There can be no diminishing the accomplishments of LeBron James and the Cleveland Cavaliers, who came back from a 3-1 series deficit to end the Golden State Warriors’ reign as NBA champions. King James deserves every accolade he receives — and I’m sure he’ll have words for the doubters who didn’t think this was possible. Like me.

I’ve been bullish on the Warriors for a while now. Early in the season, I wrote about Stephen Curry’s ability to handle virtually any shooting burden before I even knew that he was a legitimate threat from 30 feet. My operating theory was that Curry should take more and more and more 3-pointers. In the final game of the season, he had 14 attempts (yay!) but made only four (oops).

More Sports

Teams with recent championship experience tend to be money in the playoffs, and this Warriors squad seemed not to be affected much by the strength of their opponents — both of which should have made the team bigger favorites than even our models suggested. So when the winningest team in NBA history needed to win only two of five (and then one of three) games against a team it had beaten all four times they played this season — by an average of 22 points — I felt pretty good about my position. I was wrong.

The Warriors ended up losing as many games in the playoffs as they did in the regular season, finishing with a particularly un-GOATish 7-7 stretch against Oklahoma City and Cleveland. Although a 14-game break-even stretch doesn’t sound horrible, it would be highly improbable for a team that won 90 percent of its games, as the Warriors had with Curry playing. Of course, the Thunder and Cavaliers were stronger than the Warriors’ typical opponents (the Warriors did go 5-0 against them during the regular season, and remember they crushed strong opponents as easily as weak ones throughout the season), though this is partly offset by the Warriors’ home-court advantage.

A result like that 7-7 raises questions: Have the Warriors been solved? Is Curry unable to carry an offense single-handedly after all, or was his run of bad form because of something else, like a lingering injury? From an empirical standpoint, this bizarre end to the season doesn’t tell us as much as we would like, but it does hold a few insights.

Although Curry’s performance in the playoffs clearly regressed from his in the regular season, why that happened is unclear. There are, however, reasons to believe it was more than simply a run of bad shooting luck, but not something as dim as the Golden State model succumbing to playoff basketball. Since his first injury, in Game 1 against the Houston Rockets, Curry hasn’t really been Curry. He’s struggled to turn the corner on defenders on his way to the rim, and he’s had a troubling turnover rate, which may have happened (in part) because he found himself with unfamiliar passing lanes after being unable to create his usual space. Curry’s dropoff was also evident in his shooting weeks ago, and it hasn’t improved. Since his return, Curry has been worse in virtually every significant metric — even compared to his pre-injury performance against the same opponents:

Pre-injury, non-playoff opponents 44.8% 56.5% 24.3 61-7 89.7%
Pre-injury, playoff opponents 49.6 56.6 25.8 11-1 91.7
Post-injury 39.3 48.5 17.1 10-7 58.8
Post-injury Steph played worse, won less


Game Score is an imperfect metric for combining box score stats, but in this case, it gets the job done. Pre-injury, Curry performed better against his playoff opponents this season than he did against a typical team.1 Moreover, he performed similarly against all four squads.

Looking at the bottom line: The Warriors went 10-7 with Curry playing after his first injury. Again, although that doesn’t look dramatic, the likelihood of it happening by chance alone can be quite slim: For a team that wins 90 percent of games, the chances of losing seven of 17 are around 1 in 10,000. At 80 percent, they’re around 1 in 100, and at 70 percent, they’re around 1 in 10.2

Also, when the Warriors lost three games to the Thunder (before going on to win the series), it seemed unremarkable — in part because it was in line with the tendency of teams that are good at winning also being good at winning playoff series. But now that the Warriors have lost four games to the Cavs, those results corroborate each other, suggesting that the Warriors weren’t just running badly, but that there was something systematically awry.3

A 1-in-10 phenomenon is well within the range of stuff that happens in sports every day, and even 1-in-10,000 phenomena still happen. But the question isn’t whether the Warriors’ dreary finish was unlikely, but, given its unlikeliness, what is the most likely explanation. Did the Warriors just get unlucky? Are they — gasp — anti-clutch? Did two teams suddenly figure them out? Or was Curry’s injury a bigger factor than he let on?

As usual, when something crazy happens, there can be many causes. A few Warriors may have performed poorly in the clutch. Teams may have “figured them out” to some degree. And they may have gotten a little unlucky. But those are the sorts of things that all teams have had to deal with historically, and teams as good as the Warriors haven’t broken overnight. More importantly, for the Warriors to pin this on fortune alone would require luck so profoundly bad that they’d be dodging falling pianos. That’s good news for Warriors’ fans. It means that, should they be able to heal what’s ailing them — such as a lingering injury to the league MVP — winning more championships still depends on fairly predictable outcomes rather than cruel turns of chance.

At the very least, revolutionaries have good reasons to be hopeful that next season will continue where this season seemed destined to go rather than where it ended up.

VIDEO: The greatness of LeBron James


  1. “Pre-injury, playoff opponents” includes regular-season games against the Rockets, Blazers, Thunder and Cavaliers and Game 1 of the Warriors’ playoff series against the Rockets. “Post-injury” includes Game 4 against the Rockets and beyond.
  2. Based on a binomial calculation assuming that the Warriors were as good as their record, that their odds of winning were the same in each game, and that each game’s outcome is independent of the others. Note that these would not be great assumptions for making a precise calculation but are fine for a first cut, particularly with a Warriors team that for most of the season — did I mention? — didn’t seem to care much who they were playing.
  3. By Game 7 of the Finals, sportsbooks appeared to have this pegged, with most putting Golden State around -180, which would be absurdly low under normal circumstances.

Benjamin Morris researches and writes about sports and other topics for FiveThirtyEight.