-- Will Russell Westbrook average a triple-double this season? And what are his chances of doing so?
History can only do so much to answer this question. After all, Westbrook is already in triple-double territory later than anyone since Oscar Robertson some five-plus decades ago.
So instead of looking to the past, let's look to the future with a simulation that can help us estimate Westbrook's chances of joining Robertson as the second player in NBA history to average a triple-double.
Projecting Westbrook's performance
What's an appropriate baseline for what we can expect from Westbrook going forward? Given that he's never averaged more than last year's 7.8 rebounds per game over a full season, the 10.9 he's averaging so far this season seems unsustainably high.
To answer this question, I took a look at defensive rebounding percentage and other key stats for players through the end of last November as compared to the rest of the season. Strong starts tell us more about some areas than others. Shooting tends to regress heavily to what players have done in the past, while stats like usage rate that are largely a function of a player's role tend to stabilize early in the season.
Rebounding falls in the latter category. By this point, Westbrook's current defensive rebound rate (25.9 percent) makes up a higher percentage of his projection the rest of the season than his preseason estimate from my SCHOENE projection system (17.2 percent).
(Rebound rate is an estimate of the number of available rebounds a player grabbed while he was on the floor; Westbrook, for instance, is getting approximately 25.9 percent of the defensive rebounds.)
Based on that weighted projection and his current playing time, my best guess is Westbrook will average about 9.4 rebounds per game the remainder of the season, along with 28.4 points (he's currently at 31.1) and 10.7 assists (currently 11.0).
When it comes to assessing Westbrook's chances of finishing with a triple-double, rebounding is naturally the most important category. He should comfortably finish with double-figures assists, having averaged 10.4 per game last season, and obviously the scoring category isn't an issue.
Assuming Westbrook plays 79 games and averages 9.4 rebounds the rest of the way, he'd finish at 9.8 per game -- putting Westbrook the closest anyone has gotten to a triple-double since the Big O but just shy of joining him.
However, looking at Westbrook's overall projection doesn't really tell us his chances of averaging a triple-double. For that, we need a simulation.
Simulating 1,000 Westbrooks
Using the standard deviation of Westbrook's points, rebounds and assists per game so far this season, I put together a function that randomly generates totals in each category for every remaining game to provide a realistic estimate of game-to-game variation.
The next step in the simulation is determining how many games Westbrook will play based on the past distribution of games lost to injury. A small handful of times, our simulated Westbrook misses most of the remainder of the season due to a serious injury. More often, he misses only a handful of games.
After determining how many games to count, the simulation totals Westbrook's points, rebounds and assists and adds them to his current totals to determine a final stat line. And while the inputs are always the same, the averages can vary. In seasons where he played all 82 games, our simulated Westbrook averaged as many as 10.8 rebounds per game or as few as 8.6.
So how often does simulated Westbrook finish with a triple-double? Looking only at seasons where he plays the minimum 58 games required to qualify for the league leaderboard, Westbrook averages a true triple-double -- more than 10 rebounds and assists per game, no rounding required -- 33.8 percent of the time. Rounding up in situations where Westbrook averages at least 9.95 rebounds per game brings this percentage to 40.4 percent.
How many games Westbrook plays could have a big impact on his triple-double chances. Because of his strong start, the fewer games Westbrook plays above and beyond the minimum of 58, the better because that means his current averages will make up more of his final total.
When simulated Westbrook plays at least 80 games, he finishes with a triple-double just 26.8 percent of the time. If he plays between 58 and 69 games, his chances improve to better than 50-50 -- though obviously at a cost to the Thunder's playoff chances and Westbrook's MVP hopes.
Conversely, considering Westbrook's triple-double chances via simulation reinforces how much strong rebounding games can help him. Not only do they pad his current total, they also improve his projection going forward.?His chances have improved dramatically in the couple of weeks since I first built the simulation.
Simulated Westbrook is a lot less likely to match Robertson's triple-double total. Just once in 1,000 simulations does Westbrook reach Robertson's total of 41 in 1961-62.
Why Westbrook might beat the simulation, or vice versa
As useful as this exercise is, Westbrook only seems like a machine. As a living, breathing human being, Westbrook might fatigue over the course of the season, falling short of even projections that are regressed to his pre-2016-17 mean.
At the same time, there are reasons to believe the real Westbrook might outperform his simulated counterpart. Perhaps his improvement on the defensive glass will prove permanent. Or now that he's averaging a triple double, maybe he will be more focused on rebounding and making history -- something challenging to build into a simulation.
Whether Westbrook's chances of averaging a triple-double are really better or worse than the simulated 33.8 (or 40.4) percent, one thing is clear from this analysis: making triple-double history is now a real possibility.