Death from Above: Relying on Three-point Shooting

Live by the three, die by the three — that maxim has permeated the basketball culture to the point where there’s not much of a thought process behind it. It’s a quick way to inscribe a negative connotation onto a team or explain away a wonky result. But we don’t often analyze this phrase and what it could possibly mean.

As the NBA’s reliance on three-point shooting grows, so too does the league’s unpredictability. Three-point percentage has one of the lowest year-to-year correlations out of any stat, probably because it is a stat that takes a while to stabilize. Some of the effects are subtle, but it’s important that we accurately gauge the power of the fluctuation of three-point percentages. If a team fails because its outside shooting goes cold, we all tend to search for other explanations and confirmation bias creeps in. See? A team this small can’t win. And the inverse is true as well, because some streaky hot shooting will typically lead to a few written pieces about how we’re under-estimating team X or why player Y is better than everyone thought. We just have to get better at conceptualizing outside shooting and who deserves the credit.

For instance, the Cleveland Cavaliers had a few hiccups against the Raptors after steamrolling their way to the conference finals. Through the first four games, the team shot 33.3% from behind the arc. About 37% of those shots were defined as wide open, per SportVU, with 30.9% and 28.5% open and tightly defended, respectively. In the regular season, the Cavaliers shot 36.3% from behind the three-point line. They were not just going cold, however, because more of their shots were being “tightly” defended — in the regular season only 17% of those shots were tightly defended with 39.4% open. Using their regular season rates, they were only a couple three-pointers short of their expected total, which is significant but it’s not a major deficit either — there were other factors at play.

Compare those numbers to their two series-clinching wins. Cleveland made 27 three-pointers, but based on how tightly defended they were an expected total would be roughly 19. One can adjust for the players — who’s taking the shots — but the percentages were still much higher than “expected” and even LeBron James, in recent seasons a very iffy shooter from behind the arc, was hitting. As soon as a pattern emerged and they looked “cold” the Cavaliers unleashed havoc from the three-point line, and it was too much for the Raptors.

On the other side of the country, the Warriors were facing a monumental upset from the Oklahoma City Thunder. They were down 3-2 playing on the road, and looked beatable. And, of course, three-pointers were a primary cause. Again, even considering how closely defended the shots the Warriors were taking and their regular season percentages, Golden State was been underachieving. The Dubs hit only 52 of their three-pointers compared to the 60 or so that would be expected through the first five games given those shots. Conversely OKC hit 45 compared to the “expected” 42. All told, that disparity represented a swing of nearly 33 points in the Thunder’s favor. Even factoring in a few some adjustments[1. A missed three-pointer doesn’t equate to zero points because of offensive rebounding. A quick rule of thumb is that a team can expect between a quarter to a third of a point from a missed three-pointer, depending on their offensive rebound rate and efficiency.] it’s still a series-swinging difference. The Thunder had only outscored the Warriors by 22 points, and they couldn’t depend on Andre Roberson hitting 57.1% of those shots or Klay Thompson slumping to 29.5%.

So what happened in that now fabled Game Six? The Warriors hit 46.7% of their deep shots, while the Thunder only managed a pitiful 13.0%. Thompson, specifically, was a terror, breaking the NBA playoff record with 11 made three-pointers. The Warriors needed a come-back late in the game, but the pendulum had swung the other way — they don’t look cold anymore. There are some narratives you can roll with, but the truth is closer to random chance than we’d like to admit. It’s hard to predict when a lot of those shots are going to fall in.

There are no easy solutions here. There are some frameworks in grading how well a team is defending outside shots[2. The best frameworks require non-public data, sadly, making their accuracy is difficult to ascertain.] — the best thing to do, of course, is to limit the number of attempts. But even after the use of some fancy SportVU stats, there’s still a lot of uncertainty in the numbers. The Cavaliers reversed their trend hard, and so did the Warriors even though they’re dealing with a mortal Stephen Curry recovering from a couple injuries.

How will Golden State perform tonight? And how well will the Cavaliers shoot in the finals? Anyone answering with certainty is blind or selling something, and that nature of the NBA, a league increasingly reliant on these shots, is going to confound the media more and more. We just can’t be swayed by quick conclusions and biases because the long stretch of a team’s or player’s history is more informative. This “live-and-die” notion of three-point shooting, as if it’s a double-edged sword, is misleading and misses crucial details. You lose when your strengths aren’t being applied and your weaknesses are exposed — that happens to every stat. Teams often fall because their opponents are “living” by the three-pointer at an “unsustainable” rate. There is no moral failing when a team is unable to hit their open shots from behind the arc. People are creating stories from randomness and missing the storm.

One storm ends tonight, and it’s up to us to interpret the signs and catalog what really happened.