What We Talk About When We Talk About Analytics

Jan 17, 2015; Houston, TX, USA; Golden State Warriors guard Andre Iguodala (9) shoots during the third quarter against the Houston Rockets at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports

[Ed. Note: This piece first appeared at our sister site, Hardwood Paroxysm]

With 5.8 seconds left in the 4th quarter and the Golden State Warriors losing to the Chicago Bulls, Andre Iguodala attempted a three-pointer to win the game. The ball bounced off the rim, but thanks to a mix of good positioning and luck, Draymond Green was there to tap the ball in and send the game to overtime.

It was one of the best games of the season, with the Bulls, led by an incredibly bizarre night from Derrick Rose, squeaking by the Warriors in overtime.

Chris Webber, who was calling the game for NBATV, wasn’t a fan of Iguodala’s shot, saying that while the fashionable analytics would say Iguodala should take that three, he’d rather see him take it to the rim and draw a foul.

Iguodala’s shot wasn’t good — Webber’s right. Where he’s wrong is in assuming that analytics would declare it a good shot. Actually, they’d say the exact opposite. Per the NBA’s tracking data, Iguodala shoots 31% from three in pullup situations. Driving wouldn’t have been much better, as Iguodala’s free throw rate is a career low .263, his free throw percentage is also a career low (number) and he only scores .8 points per game on drives.

But this isn’t about what the numbers do or don’t say, it’s about what we say.

“Analytics” has become a catch-all shibboleth for both the pro-and-con parties. The naysayers mockingly say that “any shot that isn’t a three or a layup is a bad shot,” which isn’t true. It’s the strawman they set ablaze when their own notions of What’s Right are proved correct. This is what happens when you over-simplify a concept — it crosses the line from being easily understood to being buzzy.

The pro-analytics crowd, meanwhile, tend to roll their eyes not only when someone is dismissive of advanced numbers, but when someone cites regular numbers. Some take it too far, fulfilling the awful stereotype of those who forego watching games and instead rely on numbers alone.

Both sides are at fault here. The analytics side has done a poor job of not just explaining the numbers, but allowing for nuance, context and exceptions. Their opponents, meanwhile, which are usually former players or coaches, are ardent in their belief that numbers can’t tell them anything different than what they’ve seen or experienced.

Perhaps both groups feel as if they’re being either pushed out the door or unfairly locked out of the building, when really neither is true.

It’s a bit like religion. Faith is a good thing, even a great thing. When wielded as a weapon, however, it’s a terrible mechanism that only creates problems. Zealotry leads to ostracism, not inclusion. The more one side ignores the other’s argument while simultaneously trying to tear it to the ground with their One Truth, the more people in the middle grow to either ignore or outright scorn them. Which is a shame, because it’s the people in the middle who could benefit the most from hearing both sides.

Webber’s commentary was both right and wrong, but more importantly, it showed just how warped the term “analytics” has become on both sides of the argument. We have more access than ever to advanced numbers, thanks in large part to SportVU, but people won’t seek out the information for fear of being shamed if they don’t immediately understand it, or being accused of not watching the games.

Maybe it’s time we stop arguing about who’s right and who’s wrong — an argument that, throughout history, has rarely led to anything besides greater enmity and conviction on both sides –and just start talking. Sometimes, the discussion is just as important as the answer it produces.