Who is Daryl Morey?

Daryl Morey is what Billy Beane was to the Oakland Athletics, a MIT graduate who believed in taking a mathematical based approach to decision making. His journey to become the General Manager of the Houston Rockets didn’t come without an uncanny set of events where he learned that people’s opinions are laced with biases, heuristics and stereotypes. Morey always wanted to help basketball teams win, but no one really let him do that. While studying in university, he would write letters to convince basketball organizations to hire him for entry level jobs. “I didn’t have, like, any way to penetrate organized sports, 1” he said. “So I decided at that point that I had to be rich. If I was rich I could just buy a team and run it1”. He later enrolled into business school and landed a job at a consulting firm. As luck would have it, the consulting firm Morey worked for secured a new client, the Boston Celtics to help them with data analytics. It was at this time Morey left his consulting job to join the Bostin Celtics, where he was asked to help with the NBA Draft. As any sports fan would unequivocally agree, the draft is perhaps the most difficult task to undertake given its ability to expose the human mind’s weakness to emotion.

Now that Morey joined the Celtics to use his statistically based approach to decision making, he was immediately given the opportunity to draft in the 2003 NBA Draft. The Celtics wanted to experiment with their 56th pick – a late round pick in which players usually turn out to be nothing more than a bench warmer or a dunk contest participant. Morey used his statistical model to pick Brandon Hunter, a power forward from Ohio University. Not only was he the first player in the NBA to be picked by an algorithm, but he ended up becoming a pretty good international player. In 2007, Morey accepted a job offer from the Houston Rockets who were looking for a new general manager that closely resembled Billy Beane. Not surprisingly, the owner and executives faced plenty of criticism for risking Morey with potentially a more experienced, well-rounded general manger. Morey proved all the naysayers why he deserved this position, however. He helped the Houston Rockets achieve the third-best record in the NBA from 2007 to 2018 and made them a playoff contending team by drafting players such as Clint Capela (pick 25, 2014), Chandler Parsons (pick 38, 2011), Aaron Brooks (pick 26, 2007), Patrick Patterson (pick 14, 2010) and Marcus Morris (pick 14, 2011).

Andrea Bargnani’s 2006 NBA Draft

As a religious basketball fan, I’m less fascinated by watching the Warriors and Cavs play in the finals for the fourth consecutive year and more fascinated by watching the NBA Draft. Watching the NBA Draft is like watching the Bachelor, a highly emotional feeling that questions predictability and human decision making. I remember watching the heart-breaking 2006 NBA Draft when the Toronto Raptors selected Andrea Bargnani first overall. “Dirk Nowitzki like shooter, Chris Webber like passer, Paul Gasol like potential” were comparisons that helped ease my pain, though it didn’t take long for that pain to reoccur after his dismal first year coupled with an injury that looked more like a disinterested 7-foot tease. What was going on in Bryan Colangelo’s mind? I suspect the optics of his decision was more influenced by the figment of his own imagination than it was about an accurate interpretation of data analytics. When scouting Bargnani, for example, Colangelo overvalued the story that he was creating in his mind (21 year old with height and skill, a 21 year old who could turn out to be like Dirk Nowitzki, a 21 year old who could bring a different perspective to the Raptors). It seems to me that this is the quintessential example of what not to do, and therefore why Morey created his statistical model. I’m not suggesting one is better than the other because both approaches do have flaws, just like both approaches do have benefits, and it’s for this reason why I think you need to have a combination of human instinct and statistical modeling to draft successfully. The statistical approach should be very carefully interpreted, however.

Interpretation of statistics

Imagine for a second that you played for the Toronto Raptors. In your first game you scored 20 points, 8 rebounds and 4 steals. Nick Nurse wanted to test the water and so he played you for 40 minutes. Everyone immediately believed that the water tested well because those stats look impressive for a player who hadn’t played in the NBA before. A knowledgeable basketball critic, however, would ask, how many points, rebounds, steals, etc., did you get per minute? This critic understands that there’s a substantial difference between scoring 20 points in 40 minutes and scoring 20 points in 20 minutes. In the former, what we know is that you can score 20 points in 40 minutes and still hurt your team. In fact you can be the reason why your team lost (think about JR Smith’s blunder in the finals, let’s assume he scored 20 points in 40 minutes). In the latter, we can make a sensible prediction that you scored a point per minute. To score a point per minute means that your efficiency rating is very high which therefore minimizes the chances of causing harm to your team. This is how Morey interprets stats using his model. He goes as far as collecting data that hadn’t ever been collected before. That is, not just basketball data but data about players’ lives. For example, did it help a player to have two parents in his life? Did a left-handed player have an advantage? Did players with strong college coaches tend to do better? Does it matter how much a player can bench press? While Morey continues to face a lot of criticism for his emphasis on analytics he certainly produces results. His strategy is no different than stock market traders and political organizations that strive to make predictions. I hope the Raptors use Daryl Morey’s statistical model in future drafts considering how poorly they drafted players historically.

By Vinu Selvaratnam
The views expressed in this article are my own and does not represent any institution.

_________________________________________________________________

1 The Undoing Project by Michael Lewis
Ideas and facts taken from The Undoing Project by Michael Lewis

Advertisements