Mark Cuban is about as statistically informed and committed as any owner in the NBA. His willingness to incorporate more detailed analytics into his team-building approach has drawn attention to valuable sources of basketball data, and Cuban’s endorsement of events such as the Sloan Sports Analytics Conference only helps to bolster the reputation and awareness of the utility of advanced statistics in basketball.
That’s why when Cuban drops the “last rites” for building teams using quantitative analysis on his blog, it’s headline-worthy. Or at least, it should be.
Part of me wants to say that the title of Cuban’s blog post overstates his intention. However, Cubes is rather explicit in some of his rhetoric, so much so that the meaning is unmistakable (emphasis mine):
Sure many of us across the NBA have spent a boatload of money on “sabremetrics”. It has not been valueless. We are better able to coach the teams we have because of all the information. But the reality is, we can not put together teams based on stats. You can take all the PER, WP, WP48, Adv plus/minus and the rest and when you add them together the day before the season starts you still know nothing more than the minute before you added them all together. They are meaningless when it comes to putting together a team.
However, as Cuban continues, I think his point becomes a bit more clear:
Until you can quantify coaching and chemistry, you can not use the numbers to build a team. Period end of story. You can use them as partial input along with scouting and other elements, but there ain’t no Moneyball solution or the NBA and I don’t see Bill James walking through that door with a solution. Stats will continue to play a role in lineups, matchups and trends, but teambuilding, not so much.
In a sense, I agree. Basketball is far more complicated than a left-handed pitcher vs. right-handed batter, or even a ball vs. a defender. There are 10 players moving and interacting on the court at a given time. They’re reacting, responding, and adapting. Their individual success depends on the locations and actions of the other nine players on the court, so much so that reducing the game’s entire action to a mere number is a misguided evaluative method.
It’s from there that Cuban’s opinion and my own begin to diverge. Using one number for the purpose of an all-encompassing evaluation is flawed in a number of ways, but if we take that one number, that one source of data, and combine it with many others, the uses are both obvious and beneficial. Just as Cuban says, advanced metrics can be used as a single data source, a “partial input” to accompany comprehensive video scouting and every other useful source of data available. Why would such a thing only be useful for coaches, but not for the managers that assemble the roster of players that those coaches draw from?
Is knowing a target acquisition’s WARP really not useful in determining how they may affect a lineup? Is it not helpful to see a free agent prospect’s PER allowed when playing at various positions in order to evaluate which potential role would be the best fit? How about seeing Brandon Roy’s free throw rate? Steph Curry’s turnover rate? Josh Howard’s usage? Brendan Haywood’s points allowed per possession on pick-and-rolls? Or perhaps the Cuban-endorsed on/off defensive efficiency differential? Maybe Cuban’s definition of quantitative analysis could be very different from my own, but I see no reason why those measures — and many, many more — aren’t contributing to the wealth of basketball knowledge that informs personnel decisions, as well as coaching.
There may not ever be a Moneyball-style discovery that breaks the game open, but the proliferation of more detailed statistics can only endow NBA decision-makers with more, valuable information. Pure scouting won’t catch everything, just as a pure statistical approach will miss plenty of relevant points. The true value of both camps is in finding ways to reconcile the qualitative and the quantitative, and utilize as much of the available data as possible to make sound decisions. No approach is infallible (not even Moneyball, I’d add), but having smart decision-makers guided by as much useful information as possible is probably the closest we’ll get.