Imagine an NBA team that:
- Hired a Director of Basketball Operations who’d gotten an MBA from Duke and risen through the league thanks to his focus on analytics
- Drafted a player in the first round who was rated so highly by a statistical formula, not even the formula’s creator believed the player should have been selected there
- Received a projection of 49-33 and fifth in its conference from one statistical formula and was expected to finish 46-36 and fourth in the conference by another, independent statistical formula
However, this team fell nine games below .500 early in the season.
The head coach had repeated run-ins with the team’s highest-paid player. The first-round pick showed solid potential, but he struggled after the team vaulted him into the starting lining up because it failed to acquire viable alternatives at his position. The top three players had no chemistry together.
Wouldn’t that seem like a team too reliant on analytics?
As I’m sure you’ve figured out by now, I’m talking about the Pistons.
For a team that supposedly pays minimal attention to analytics, the Pistons sure put together a team that just happened to rate well in preseason statistical prognostications.
ESPN’s SCHOENE projection predicted 49 wins, and a projection based on Wins Produced yielded 46.4 victories. Kevin Pelton’s WARP projections placed Caldwell-Pope No. 4 overall, but Pelton ranked Caldwell-Pope behind Trey Burke and C.J. McCollum in his own subjective rankings.
And the Pistons hired Ken Catanella three years ago as their Director of Basketball Operations – a move The Palace president Dennis Mannion said was not forced on Joe Dumars.
This is largely a top-down approach and the Pistons have a couple of old-school guards in the top chairs. I’ve never heard Joe Dumars utter the word “analytics,” much less apply one. Maurice Cheeks, asked about advanced statistics earlier this year, responded with something about 3-point percentage. They govern by eyeballs.
Lawrence Frank was attuned to analytics and who/what succeeded/failed under specific conditions or stimuli. He had a coaching staff that was attuned to it, too. Cheeks’ coaches are, too. Cheeks has admitted his coaches would inundate him with such statistical analysis if he showed interest in it, which he admits he generally doesn’t. Cheeks’ stated approach is that he’ll listen to anything interesting that’s brought to him. If you bring him good information, he’s more likely to listen. That’s the analytics department here.
Perhaps, Mayo knows enough about how the Pistons truly run to make that assessment despite the circumstantial evidence to the contrary I presented above. But, if he does, he doesn’t credibly make his case here.
Dumars has never applied an analytic? NBA teams are especially secretive about their use of advanced stats, and the Pistons are secretive about everything, anyway. I don’t expect Dumars to run around talking about his use of analytics.
He doesn’t need to use the word “analytics” to apply one, anyway.
When Dumars says of Brandon Jennings, “We like his ability to score off the bounce,” that might have been determined through analytics. Maybe Dumars just watched game film without taking any notes, but I’d say it’s more likely he at least complemented that method with Jennings shooting-off-the-dribble statistics. You don’t have to talk in numbers to have numbers influence what you say.
As far as Maurice Cheeks, I’ve never seen him as a beacon of statistical understanding, but I don’t know how bringing up 3-point percentage disparages him – as if 3-point percentage is unworthy of the higher minds of basketball analytics. Perhaps, Cheeks embarrassed himself with his answer on the topic – it wouldn’t be the first time – but there’s nothing inherently wrong with mentioning 3-point percentage. If Cheeks’ answer lacked insight, it needs to be explained more thoroughly than Mayo did.
Regardless, this goes back to the first point. If another member of the staff crunches numbers and determines something about a player and then tells Cheeks in plain English, the analytics are working. There’s a value in translating numbers to digestible form.
On a related note, from the outside, it’s impossible to tell exactly how much the Pistons – from Dumars down – use the information Catanella and other statistically inclined members of the organization give them.
But it sure seems Catanella is the type of statistical analyst Dumars would listen to.
Catanella, who used to work on Wall Street, spoke at the 2009 New England Symposium on Statistics in Sports. He called communicating with co-workers who aren’t necessarily statistically inclined “by far, the most important part of the job.”
Catanella, years before the Pistons hired him, essentially described how he’d make an impact on a team run by Tom Gores and Joe Dumars:
Implementing that in the sports world is even more important, especially if this a new position to the team you’re joining, which it was when I was in New Jersey. … You’re developing a trust level with the organization, with the coaches, with the GM.
And that takes time, but if you’re open to telling them when you don’t know the answer and telling them why, I think they’ll trust you more. As opposed to, ‘I know the answer every time, and this is right, and this is what you have to do.’
Also, choosing your battles wisely. When it’s a really important decision in your mind, it might not be a really important decision to the general manager or to the coach. Be sure to make that clear.
And then, at the end of the day, if you actually are getting those things implemented, then you know you’re being successful. You could be the best statistical analyst in the world, and you know you’re right 100 percent of the time, and you knock the ball out of the park every time you do an analysis. But it only gets implemented one time out of 10, the guy’s who’s a little bit, not less aggressive, but not as accurate, but gets things implemented five times as often is probably the better analyst, in my mind.
It is trending more and more towards an analytical approach, partially due to the changes in ownership that have occurred in our teams. So, a lot of times, the owners of today are coming from backgrounds where they own businesses or they ran hedge-fund companies, and they’re used to seeing in-depth analytical studies on every major decision that their companies make. And they want to see the same type of work at their sports organizations.
I supposed it’s possible, despite his preparation for and focus on having influence, Catanella hasn’t achieved meaningful results in Detroit. But he spoke about that very subject during last year’s MIT Sloan Sports Analytics Conference.
Here are a few things he said:
I found the most benefit from having divergent backgrounds, especially in the front office, because each person lends a completely valuable different and valuable perspective that, at the end of the day, it probably comes down somewhat of a wisdom-of-the-crowds philosophy, right? And a lot of the times, we’ve seen more literature, that the best decisions are often made from completely different opinions.
In our organization, it’s exactly that way, and I think it’s structured beautifully to have someone that’s focused purely on, perhaps, the background element of a player or someone that’s focused purely on the coaching elements. We have former coaches that are scouts. We have former players that are scouts that get into the personal side of people. The analytical piece. And perhaps somebody that’s more experienced. And at the end of the day, the chief decision maker has much more information to make that final decision.
I think it gets back to the human element, because, if you think about somebody coming into your office and grandstanding and putting on a show and saying, “I’m 100 percent right, there’s no way that you could possibly be right, there’s no way you could possibly have something to add to this discussion” or that it couldn’t get better – doesn’t that make it eminently hard for you to accept the idea and then not only internalize it, but then make it an actionable item? That’s what we’re really trying to do. That’s what analysts are trying to do. They’re trying to make an impact on an organization by transferring their knowledge to somebody that will actually make use of it.
Whether it’s trying to get information to a player who’s about to guard somebody that has to be able to internalize it to the point where he doesn’t even have to think about it. He knows exactly the way he wants to guard that individual so that it’s reflex. To the GM that gets faith in the information and the decision that he’s ultimately making from that discussion, and he actually, potentially, partakes in the ultimate conclusion by adding to the discussion and tinkering with the idea, like you said.
A lot of times, our ideas are not entirely fully formed, even. Sometimes, I come into someone’s office looking for guidance. So, I think that just allows other people to gain greater acceptance and embrace your ideas when you have that type of approach.
That was a very challenging one, because you think about, I had played some professional basketball in Europe and worked in the front office over there. But here, there’s a program and a coach that had credentials, infinite credentials and a prestigious program. And how could I add value walking in as a graduate assistant at that time?
I noticed a few things in terms of their pregame prep, and I was doing some video logging of opponent games. And I automated a process that created what now is commonplace, but over a decade ago was a rarity, is a shot-zone chart that had visuals and colors. At that point, I gave it to the coaching staff and thought nothing of it. And the next thing I knew, at the practice later that day, in preparation for the next day’s game, Coach had blown it up to an infinite size, brought it over to the bench where the guys were sitting. Of course, it was a proud moment for a geek like me, but he showed it to the guys and said, ‘This is what we have to do to stop this team if we play this player this way.’
And at that moment, I realized, if you can just find that niche of something that is missing or that you can add an element that can make them better at their job, they’re going to really appreciate you and trust that you have their best interests at heart, like you were saying Alec. And if you can show that you have a passion for the game too and winning and that you’re a person that has a similar mindset and similar goals, those things also have a powerful impact.
Data really has a hard time with context. I can have as much information as I want, but if I don’t – I’ve explained this to you.
A term we use a lot is, ‘Smell the gym.’ Just get in there and feel the game again and feel what the player is seeing and what his interactions are like with his teammates and his coaches. Those are data points.
We just have to think about data in a different way and try to develop as much information around the core base of information that we use to evaluate a player. It might be on the periphery. It might be weighted heavily. It might be weighted lightly. But we definitely want to consider it.
Doesn’t that sound like the exact type of analytical specialist who can persuade Dumars?
Again, it’s still possible Dumars has ignored Catanella.
But would Catanella really choose to sit on a panel about having influence within an organization if that were the case? It’s possible, but that seems unlikely.
Would Gores compliment Catanella by name before while praising Dumars before the season for meshing with Catanella? Gores, via Dave Pemberton of The Oakland Press:
“Early on I said to Joe, ‘We got to make changes.’ I think the thing Joe has done is he’s adjusted along the way,” Gores said. “Just like in my own business, I have to grow, I have to adjust, it’s all about getting better and not getting stuck in the old way. Joe has shown every sign of a person who can grow.
“He’s done amazing things with his own basketball operations. Ken (Catanella) and George (David) and that group, those are great young men that are smart. They compliment Joe’s talents. We had success in the offseason.
I still can’t say with total certainty the Pistons don’t “govern by eyeballs,” but the circumstantial evidence is piling up that they don’t.
Analytics are not a magic bullet that solves every problem. A team can apply analytics and still make bad decisions. After all, the Pistons are competing with other teams that definitely use analytics.
We can have another discussion about whether the Pistons use analytics effectively. That’s not the question I’m addressing.
Do the Pistons use statistics in their decision-making? I definitely believe the answer is much more likely yes than no.