20 Second Timeout

Efficiency Versus High Variance

"Stat gurus" outsmart themselves when they value offensive efficiency over every other factor. They have determined that three pointers and free throws are the most efficient NBA shots. While that may be true mathematically, it is not true in a relevant way in the real world; there is value in trying to improve offensive efficiency, but there is also value in improving proficiency in other areas, including defense and rebounding.

The league's collective three point percentage this season was .358; last season, the league's collective three point percentage was .355. Only seven qualifying players shot better than .420 from three point range this season, while in 2018-19 there were 12 qualifying players who shot better than .420 from three point range. Defense is better in the playoffs than the regular season, so if a team can shoot .380 or even .360 from three point range in the postseason that is quite efficient--and yet achieving such a percentage, even on a high volume of shots, by no means guarantees success. The Houston Rockets prove this annually in the playoffs, yet the "stat gurus" do not seem to be willing or able to draw proper conclusions from the evidence.

Offensive efficiency is not all that matters when trying to construct a championship team: the three point shot produces much more variance statistically than shots taken closer to the basket, meaning that a team that shoots .380 from three point range may shoot .250 from three point range in one game and then .510 from three point range the next game. If a team shoots a high volume of three pointers then it will almost certainly lose when shooting .250 from three point range, but that team is not assured of victory even when shooting .510 from three point range: a team that relies too heavily on three point shooting does not have a backup plan to use when those shots are not falling, but that team's opponents may be able to overcome giving up .510 three point shooting by forcing turnovers, winning the rebound battle, and attacking the paint for high percentage shots.

I made a similar point years ago when I analyzed Gilbert Arenas' playing style: "...if Arenas shoots 6-9 from three point range in one playoff game and 1-9 in the next then the Wizards will go 1-1 at best in those games despite the fact that his three point percentage would be .389. Having your point guard jacking up eight or nine three pointers a game--particularly on a team that is not good defensively anyway and has poor court balance--is not a formula for postseason success."

"Stat gurus" think that if they put together a team of five Gilbert Arenas-type players shooting a collective .389 from three point range then this offensive efficiency will result in winning a championship--but this did not work out well for Arenas or his Wizards, and it will not work out well for any team that prioritizes three point shooting over building a well-balanced team.

"Stat gurus" and the media members who lavish praise on them act as if their critics do not understand the basic math that three is more than two. Everyone understands that a three point shot is worth more than a two point shot*; unfortunately for "stat gurus" like Houston's Daryl Morey, basketball games are won and lost on the court, not on spreadsheets--and on the court you have to account for more factors than just shooting percentages. Three point shooters will wear down and get fatigued in a seven game series after battling against larger players in the paint. Also, when Houston went all-in with small ball the Rockets accepted losing the rebound battle by a wide margin on a consistent basis, which puts tremendous pressure on them to not only shoot a high three point percentage but also to minimize turnovers--and players who are fatigued not only shoot worse but they tend to commit more turnovers. We see this happen every year in the playoffs with Houston in general and James Harden specifically, yet the "stat gurus" and their media sycophants never learn. Last year, Harden's numbers went down across the board during the playoffs and he was unable to lead his Rockets past the Golden State Warriors sans Kevin Durant--but that was a continuation of a pattern for Harden and the Rockets.

Here are the Rockets' three point field goal percentages and James Harden's three point field goal percentages for the last five series that they have lost:

2019: Western Conference Second Round, 4-2 loss to Golden State: .367 (Houston); .351 (Harden)
2018: Western Conference Finals, 4-3 loss to Golden State: .314 (Houston); .244 (Harden)
2017: Western Conference Second Round, 4-2 loss to San Antonio: .366 (Houston); .308 (Harden)
2016: Western Conference First Round, 4-1 loss to Golden State: .268 (Houston); .310 (Harden)
2015: Western Conference Finals, 4-1 loss to Golden State: .333 (Houston); .429 (Harden)

Here are the Rockets' three point field goal percentages and James Harden's three point field goal percentages for those five regular seasons:

2019: .356 (Houston); .368 (Harden)
2018: .362 (Houston); .367 (Harden)
2017: .357 (Houston); .347 (Harden)
2016: .347 (Houston); .359 (Harden)
2015: .348 (Houston); .375 (Harden)

In four of the five series listed above, Harden shot worse from three point range than he did during the regular season that year, and in three of the five series listed above the Rockets shot worse from three point range than they did that year. The larger point that must be mentioned is that even when Houston's three point percentages are not bad, the variance from game to game makes it very difficult to win a playoff series against a good team.

We are seeing some of the same high variance patterns in Houston's first round series versus Oklahoma City. In Oklahoma City's 117-114 game four win, the Rockets opened the third quarter by making eight straight three pointers. This is the first time in the last 20 postseasons that a team opened a quarter by hitting eight consecutive three pointers. Then, the Rockets shot 5-26 from three point range during the rest of the second half, turning a 15 point lead into a three point loss.

The Rockets set a single game playoff record for three point shots attempted in game two versus the Thunder, and they broke that mark in game four. Their overall three point percentage during the first four games of this series is not bad--and yet they are two losses away from being eliminated in the first round for the fourth time in eight years.

It is too soon to say what will happen in this series; the Rockets won the first two games, and the Thunder countered by winning the next two games. Russell Westbrook has yet to play, and if he returns to action at anywhere close to his normal form then that should be enough to help the Rockets to eliminate the Thunder--but it is important to recognize that the same themes we see above are recurring for the Rockets, and whether the Rockets lose in the first round or the second round they will meet their demise the same way that they have for the past several years: pursuing an "efficient" strategy until they perish due to the high variance that is an inherent part of relying too much on three point shooting. These slumps that afflict Houston at the worst possible times are not flukes; they are the natural byproduct of a flawed, high-variance strategy, and the natural result of human beings becoming fatigued over the course of a seven game series--fatigue that is exacerbated when undersized players have to battle against larger players.

There is a false narrative that the Golden State Warriors proved that a three point shooting team can win an NBA title. The Warriors proved nothing of the sort. From 2015-2019, the Warriors ranked first, third, first, third, and third in defensive field goal percentage; not only did they have a backup plan if they suffered from cold three point shooting: they had a primary plan, namely make it very difficult for their opponents to score. Also, the Warriors did not become a championship dynasty until they added Kevin Durant, who added deadly midrange scoring (and, to be sure, another three point shooter) to the three point marksmanship provided by Stephen Curry and Klay Thompson.

In contrast, from 2015-2019 the Rockets ranked 12th, 19th, 23rd, 16th, and 19th in defensive field goal percentage; even when the Rockets make a lot of three pointers, they are still vulnerable against good teams--and the Rockets are dead against just about anyone if they do not make a lot of three pointers, because they cannot control the boards or limit their opponents' scoring.

Daryl Morey has been running this "stat guru" experiment in Houston since 2007. The Rockets have missed the playoffs three times, lost in the first round four times, lost in the second round three times, and lost in the Western Conference Finals two times. That adds up to being eliminated in the first round or earlier seven times in 12 years, with no NBA Finals appearances. Has there ever been a general manager or team president who held on to his position longer with more hype and less championship success than Morey?

---

*I can say from years of personal experience that in pickup games played to 12 points with three pointers counting as two and regular field goals counting as one a post player has no legitimate complaint about not receiving the ball unless (1) he makes at least 60% of his shots and (2) his perimeter players shoot worse than 30% from three point range: in such games, a 3-10 (.300) three point shooter generates six points, while a 6-10 (.600) two point shooter also generates six points (as a reliable three point shooter--particularly from the relatively easy 19-9 three point line--I had many such discussions with teammates who made less than half of their inside shots but could not understand why I preferred to shoot threes instead of passing to them so that they could brick layups). So, yes, I understand the value of the three point shot, and I understand under which circumstances three pointers are preferable to two point shots--and I also understand that a strategy based on shooting as many three pointers as possible while hoping for the best in terms of rebounding, defense, and all other factors is not a recipe for championship success at the NBA level.

Labels: Daryl Morey, Houston Rockets, James Harden, Oklahoma City Thunder, Russell Westbrook

posted by David Friedman @ 11:21 PM

5 comments