With the Thunder down 92-82 with six and a half minutes left in Game 5, OKC Coach Scott Brooks decided that Omer Asik needed to get well-acquainted with the Oklahoma City fans. And by well acquainted, I mean he's going to get yelled at every thirty seconds of game-time while at the free-throw line.
Obviously not a fan of Kazaam, Brooks harkened the days Shaq would love to forget and employed the Hack-Asik strategy for the rest of the fourth quarter. In that time, Asik shot 16 free throws, making 11 of them and essentially clinching the game for the Rockets. The Thunder fans were left unamused, and now the series moves to Houston for Game 6.
I personally enjoyed reading the reaction on Twitter. I found the words "garbage", "hero coaching", and "idiotic" strewn about with little care for the ideas behind the strategy. Trust me, I didn't want to argue; it turned an exciting series into an excruciatingly-slow fourth-quarter to watch. I almost turned to the Sharks-Canucks Game 1. (Almost.)
However, the numbers don't lie: Scott Brooks had a specific reason for his gameplan. And in that situation, it's a reasonable and correct one.
Sure, the Thunder's strategy didn't work in Game 5. But if I'm Coach Brooks, and I'm in the same situation in Game 6, I'm not afraid to go back to it. And the reason is an idea that none less than Rockets GM Daryl Morey has been touting all playoffs: increasing the variance.
Increasing the Variance
To say Asik is poor at free-throw shooting is like saying James Harden's beard is kind of cool: you're severely understating the severity of the issue. In 299 free throw attempts during the regular season, Asik held a 56.2 percent free throw rate. In points-per-possession terms, that averages out to 1.124 points per possession expected from the Rockets, with a roughly 20 percent chance that he will miss both free throws.
The chance for an offensive rebound increases the expected offensive efficiency slightly, but how often does that occur? The Rockets finished the regular season 17th in offensive rebounding during the regular season, and leading boards man Asik isn't going to help much from the free throw line. As estimated by John Hollinger, offensive rebounds may add about .02 expected points per possession. While something, in this case, it's not enough to make a large difference.
During the regular season, the Rockets averaged 1.097 points per possession, slightly under what Asik's free throw shooting would be expected to give them. Given a pace of 94.7 possessions per 48 minutes (the average between Houston's NBA-quickest pace and OKC's slightly slower pace), that would be a 2.76 point difference over a whole game, even taking offensive rebound chances. Given Asik's poor free-throw shooting, Oklahoma City is losing a little bit of efficiency by putting him on the line, but not much.
However, the main reasoning actually comes on the Oklahoma City offensive side. You see, the Thunder take a decent amount of three-pointers; their 27.5 attempted threes per playoff game trailed only Houston entering tonight's Game 5. And what happens when you shoot a lot of threes? If you make them, it works very well; just ask the Game 2 Warriors. If you don't, then you'll lose; just ask the Game 4 Knicks. But either way, the variance is increased so that there is a larger pool of possibilities from which to draw.
Considering Oklahoma City was down ten with six minutes left, they needed to create as many possessions as possible to increase the number of chances for these highly-variable three-pointers to fall. Remember, Houston's potential points per possession is capped at two by fouling - something that has use when the Rockets shot a Stephen Curry-esque 40 percent from long-range in tonight's game and 36.6 percent from long-range on the season. OKC, meanwhile, was eligible to hit three points per possession if they made their threes.
And given Oklahoma City's offensive history, there was a solid chance they would have done just that. Oklahoma City averaged 1.124 points per possession during the regular season. Recognize that number? You should - it's exactly the same points-per-possession that Houston averaged through Asik's free-throw shooting. Only difference is that OKC had a more highly-variable outcome because each possession could result in three points instead of two.
That's exactly how comebacks happen, when both teams hold similar averages but one team has a higher potential for greatness than the other. OKC also held a bigger bust potential, but given they were already down 10, that point is moot.
Finding the exact odds of victory for the Thunder with the Hack-Asik strategy versus their normal defense is next to impossible using our metrics. You would need to find a finite way to estimate the number of possessions at the end of the game if the Thunder had played it out, and considering how basketball games move differently towards the end due to changing game situations, that's just not reasonable.
As a Coach, Brooks' job isn't to please the fans or to make things interesting; it's to try everything in his power to win a basketball game. How do you win a basketball game? By limiting the other team's possibility to score while allowing yourself the chance to go on a run.
It might not be pretty, and it may not have worked, but the rationale behind the Thunder's Hack-Asik strategy is sound. It's all about the variance, my friends.