clock menu more-arrow no yes

Filed under:

Lionel Hollins' war on numbers and the NBA's war on concussions.

New, comments

The Hook explores why Lionel Hollins' rejection of basketball analytics is troubling, and why the NBA's concussion policy is leagues ahead of one other popular sport.


Our man Kevin Lipe of Straight Outta Vancouver had the best line about what Lionel Hollins means to the Memphis Grizzlies: "It's hard to imagine this Grizzlies team being anything like what it is today without the presence of Lionel Hollins." That's just spot on. The team is a reflection of its coach, and that's a good thing. That's something a lot of coaches, GMs and fans want and don't get.

But there's no question that his recent comments about the usefulness of analytics and, in particular, line-up data, are pretty troubling. Here's the part that stunned me:

Analytics has a place. It can't be the be-all, end-all. I'm still trying to figure out when the Oakland Athletics won a championship with all the analytics they have. It takes talent. We had a guy a few years ago that was sending me emails about different lineup combinations, and he was saying, 'this lineup should be on the court a lot more because they're the most effective.' So, then you coach that lineup and keep them on the floor for 40 minutes. I'm going to stay with the lineups that I have on the floor. No matter what anyone wants to say, there are players that get it done in the last six minutes, they're players that do it in the first quarter.

Emphasis mine. Of course it's about talent! No one involved in basketball analytics would argue otherwise; in fact, the GM most closely tied to the analytics sect of the NBA, Daryl Morey, loudly states whenever he gets an opportunity that you need a superstar talent to make the playoffs and compete for a title. You use analytics to find the talent in the draft, in free agency, in trades. If you're a team that has trouble picking up signature free agents (like the Grizzlies, traditionally), you use analytics to find undervalued (read: cheap) talent. Analytics aren't a replacement for scouting. They are used in conjunction with scouting in a holistic approach to evaluation. No analytics person in any position of influence has ever advocated on behalf of a specific player without seeing that player perform.

The funny thing -- maybe it's not so funny in Memphis -- is that the knock on Hollins is that his line-up choices are sometimes poor. The Grizzlies have talent, thanks to the good work Chris Wallace and Company have done in pulling in undervalued players. Zach Randolph was practically free via trade. Tony Allen was a bargain. Marc Gasol was one of the league's most cost-effective players for a few years. Mike Conley's deal has turned off to be rather reasonable. And Hollins has done excellent work in developing that talent, particularly Conley. And the Grizzlies play extremely well on balance. But at times, Hollins appears to have the wrong players on the court. This is most evident when Tony Allen is on the bench and the opponent goes on a run. Advanced stats show value to playing T.A., even in the fourth quarter where, perhaps, he's a drain on offense. The metrics have shown that he's still worth playing because of that crazy good defense.

The Grizzlies had the services of Aaron Barzilai, one of the world's foremost experts on line-up data and adjusted plus-minus. (Barzilai is now the Sixers' director of analytics.) I don't know if Hollins is referring to Aaron or to a random fan sending in line-up data, but to ignore the advisories on principle because the freaking Oakland A's never won a title is absurd. Just as absurd as it'd be to believe them wholesale and run the team solely by the stat sheet.

The more interesting issue in the background of all of this is how someone who uses analytics can convince someone as apparently opposed as Hollins is to believe in them a little bit. There's reason to believe that the Grizzlies should keep Hollins. He's been the most successful coach in franchise history, and Hubie Brown coached this team! But can a front office that stars John Hollinger really rely on a coach that thinks the work the analysts are doing is worthless? Is that even workable? This seems like a communication and trust issue more than anything, and perhaps the prevalent Rudy Gay trade rumors are having more of an impact than we'd think. It'll be interesting to watch how Hollins' tenure plays out.


Henry Abbott and Beckley Mason have an excellent, thorough piece on concussions in the NBA. I encourage you to read it in full.

Henry and Beckley talk about a real concern that as concussion tests and precautions become more prevalent, players will begin underreporting symptoms and basically fib about whether they feel any ill effects from blows to the head in order to avoid sideline tests. That's a definite concern, and the pro sports culture of playing through pain at times looks insurmountable. (Dan Le Batard's incredible column on Jason Taylor from this past weekend is a must-read on this topic.)

And coaches, of course, are not medical professionals and are focused on winning games, which is the single biggest determinant in whether they keep their jobs. So when Monty Williams bitches about the league's concussion policy keeping his 19-year-old star out of a game, that's his job security speaking. So the question is how we can convince coaches and players to accurately report head injury symptoms when it would appear to be in their self-interest to hide it.

I think the answer is, in part, found within the NBA's current policy, where there are tests that must be passed before a concussed player can return to action. (These are the tests keeping Pau Gasol off the court right now.) If you bump your head and lie to the trainers about feeling woozy, and you continue to play, you're risking further damage (short- and long-term). Chances are you're going to have more trouble passing those post-game concussion tests, which means a longer absence, which is worse for you and your team than if you'd just been honest and come out of the game when you hit your head and felt woozy. It takes a little forest-viewing, but I think the inherent disincentive for honesty (being pulled out of a game) can be framed by the league and its doctors as an incentive (preventing further damage that would extend the injury absence).

The NBA's program will be improved over time, and it's unlikely you can totally eliminate the risks -- we've seen some horrific blows over the years. But it's a great start and a strong commitment to protecting the players, which is frankly a lot more than one particular league can say.