Nylon Questions: How can an analytic approach be applied to player psychology?
Sports analytics is a constantly evolving field and keeping up can be a challenge, especially with so much work being divided between the public and private spheres. As we head into the 2018-19 season, Nylon Calculus wanted to take stock of where we are and of what’s coming next.
This project is a throwback to work Keith Woolner did for Baseball Prospectus nearly two decades ago, and an update to Kevin Pelton’s basketball-specific version from five years ago. Our staff compiled 10 questions whose answering will likely guide the next few years of public analytic work. Not knowing what has been accomplished in private by NBA teams and consulting firms, we focused on questions that could be worked on in the public sphere, wouldn’t have to be answered with existing datasets (we don’t want to imagine the data we have now is all we’ll ever have) and things that would theoretically have an effect on how teams operate on and off the court.
Hopefully, these questions will help spark, refocus, and recalibrate conversations and lead to collaborative progress here at Nylon and everywhere else sports analytic work is being done.
3. How can an analytic approach be applied to player psychology?
The NBA analytics world has largely dominated the outward physical space in the most mechanical sense. Analysts take points, rebounds and assists data to formulate some of the most complex observable statistics to measure optimum performance. At this point, one could ask if we’ve reached the full potential on observable statistics. What more could NBA teams want analytically to get an edge in basketball?
Well as any other company could tell you, the next stop for analytics is the internal, conscionable space. As you can imagine, it’s an extremely tough space to navigate
The Known
The good thing is that a ton of smart people know a lot about sports psychology, and while much of the analytic approach to player psychology is in its relative infancy, there’s a ton of information to pore over in the field.
Young athletes, interviewing, quantifiables
The simplest form of getting into a player’s mind is the interview processes conducted by NBA teams. It’s not too complicated, really. A collection of executives or coaches gather data through a carefully constructed cohort of questions adhering to strict psychological processes and guidelines. It’s not totally clear as to how effective this process is, and is wholly dependent on the interviewer to be asking effective questions and analyzing the answers effectively. This process is pretty deregulated and sits extremely low on the totem pole in terms of science.
If we’re looking for something a bit more standardized, then look no further than basketball’s sports neighbor, football. The Wonderlic test has gained predominance in the NFL as a significant means of understanding player’s minds. While the debate rages on about the validity of the test itself, the test does produce a number. True, the number isn’t totally useful, but it’s a start. Some have even found some significance with the score as well.
The NBA doesn’t use the Wonderlic from what I could tell, and if any team uses it, I’d chalk up that use to be about as science-y as a convoluted interview described above.
There are some promising developments in the NBA based on psychological precedence. Scott Goldman of Athletic Intelligence Measures, Inc. has shown to implement intuitive psychological theory as a foundation for their battery of tests. Read Ben Dowsett’s piece on Goldman and his work, as it does a fantastic job providing insight to Athletic Intelligence Measures inner working and development of the Athletic Intelligence Quotient (AIQ). Dowsett mentions that for the most part, teams don’t get rid of Goldman’s services. He has a high retention rate of 93 percent, meaning that not only are his methods reasonable, but the theory behind AIQ testing is sound.
Their claims behind the implementation of the Cattell-Horn-Carroll (CHC) Theory of Intelligence are verifiable, in that it “is the most comprehensive and empirically supported psychometric theory of the structure of cognitive and academic abilities to date.” Per the linked research paper, most of the world is basing new cognitive testing on the theory so, at the very least, it’s reproducing sound results.
What’s great about Scott Goldman is that he’s honest:
"“If someone scores high on our test, that’s great,” Goldman said. “If someone scores low, that’s great. It’s about helping coaches and players to understand an athlete’s strengths and developing ways to negotiate the deficits.”"
Analytics is all about starting points, rehashing, and making sure variables aren’t missed. We are dealing with people here at the end of the day, so the goal should be to get team chemistry right. One negative component is that this data isn’t publically available, which sucks because it’d be amazing to know how somebody like Nick Young, Spencer Hawes, or Dirk Nowitzki score on one of these tests. I’d also be curious to see what someone like Stan Van Gundy or Gregg Popovich would do with their team’s scores in their hands.
Dowsett’s article also introduces Eric Weiss of DraftExpress Chemistry whose method, in essence, is similar to Goldman’s AIQ in that it provides reports based on role or chemistry. Weiss goes into more detail here on Troy Brown before the 2018 draft, and dives in a little bit to his personality type and which NBA player to possibly compare him to in the future. Weiss stresses that a good environment can do wonders for certain personality types and that NBA should follow.
Social media and player moods
Social media is already known to have immense power, and it can possibly provide great insight into player mood. We’re already primed to understand it, by means of the term “team chemistry.” Player morale is just a part of chemistry, and at this point, you should understand that chemistry is the ultimate goal for analytics to uncover and dissect. Let’s say Scott Goldman’s work isn’t necessarily ready yet for the day-to-day deconstruction of an NBA athlete’s thoughts or moods. You could argue that the AIQ is better suited to the broader scope of player psychology. Twitter is, however, storing minute-by-minute information from athletes’ brains directly.
Take a look at this story, at how researchers attempt to show “show how NBA players’ pre-game emotional state, as captured through their tweets, or the messages they post on Twitter before a game can help predict on-court performance in the game.” If you hate reading, the following video by one of the researchers himself explains the process pretty well:
The Twitter development is pretty astounding, in that it’s a preliminary look into the mind of a player with actionable data. Some images of the data on the research can be accessed here, and it concludes that player moods iare “positively associated with their on-court performance.”
Contrast with all of the famous methods that players and coaches deal with their own moods and psychology on a daily basis. Phil Jackson was well known to be a little before his time, while LeBron James takes a slightly more clinical approach to the mind. On a more technical note, the likes of Jaylen Brown, Aaron Gordon, and Skal Labissiere use an app called Lucid, “which promotes a new, on-the-go vision on how to get nervous athletes in the right frame of mind before a big game.” Kinda plays along the same note about Twitter use to denote pregame moods right? The app sounds corny (there’s an app for choking in crucial moments), but the company’s goal is to solve a real problem in big games. Per the linked research article, historically “the threat of severe losses didn’t lead to an elevated level of performance.” The goal is to ultimately achieve perfect team chemistry by optimizing the different characteristics of NBA players. The researchers of the pregame Twitter mood of athletes touched on something important in that players just people. They deal with problems the way we do with money, mental stability and general person-person friction.
We understand that there are a bevy of financial problems from athletes of all backgrounds, and wild gambling practices that provide a direct reflection of mental stability. Could it indeed be possible to curb or hone in on this type of psychological perturbance? It’s just a start, but apps like Lucid offer a way to get over such slumps.
At the team level, the Dallas Mavericks haven’t been totally shy about their approach to basketball psychology, so far as having investing in their athletes’ brains since at least 2011. Most other teams hire outfits armed with the workforce capable of enhancing player mental abilities, they do have their own analytical apparatuses, although nothing quite as encompassing as AIQ or normal statistical equivalent.
The unknown and possible solutions
While we know a lot about player psychology and analytics, it’s not everything, so there’s still a lot we don’t know, either. Mainly, it’s not clear cut on how to use the data. Goldman’s data, or any other research for that matter, is immensely useful, but he isn’t in charge of issuing these tests. Although the likes of Goldman, Weiss, and other institutional research gives analytics and psychology certain stability, the environment is still pretty deregulated. Much like the interview process, the data received is about as good as the collector is as good at processing it.
Because teams pay for these kinds of services, it’s much to be discerned if these tools are getting to the player and what kind of feedback they’re providing on said tools. Much like a team psychologist, a player doesn’t necessarily have to have meetings with them. If an NBA team purchases rights to a tool, how do we know that said tool is working properly?
No doubt that both Weiss’ and Goldman’s approaches provide great help to their NBA clients, but it might be in the best interest of the NBA as a league to form a consortium with the most knowledgeable sports psychology departments in universities. Every other industry has understood the absolutely unique value in direct investment in higher education for research and development, so the NBA should follow suit in the analytical realm in terms of psychological analysis. This of course, is just my personal opinion on how to solve certain problems in the NBA, and shouldn’t be considered as a knock on any private entity looking to disrupt a particular market. Weiss and Goldman surely speak for themselves, as they’re pretty smart people.
Positive and negative implications
This past season gave fans a look into the NBA players mind from the outpouring of some this past season. The likes of Kevin Love and Demar DeRozan gave us their stories on the various stressful hardships they faced on a daily basis, and how difficult it was to deal with such stresses in a tough environment. Should the NBA get serious about psychologically based analytics, a whole bevy of mental health issues could theoretically be solved. Getting a better understanding of player roles could help teams better allocate duties on the court for certain players, giving them certainty in their daily work lives. Maybe a player like Kevin Love doesn’t feel comfortable with a specific shot on the floor, but he doesn’t want to tell his coach no. If Love was consistently taking tests or if the team’s been using one of Goldman’s or Weiss’ methodologies, teams could have built their rosters a bit more accommodating to Love’s strengths. Although, not sure how he’d fair this upcoming season even with every psych test available out there.
Analytics might not be able to save bad team chemistry however. Upon quick review, there’s nothing to suggest from Bobby Portis’ or Nikola Mirotić’s social media accounts that foreshadowed a bad altercation back in October of 2017, but maybe DX Chem or AIQ could speak to their lack of fit?
Next: 25-under-25 -- The best young players in the NBA
One interesting part of this might be that teams could split from some tense on court relationships sooner rather than later. Take the Russell Westbrook and Kevin Durant situation for example. If the OKC Thunder could formulate role characteristics between the two players that could determine a bad fit, could the team have flipped Durant sooner for a player more suitable to Westbrook’s style of play? We could also probably assume that psychological analytics could’ve steered the franchise away from pairing two ball dominant guards in Victor Oladipo and Westbrook as well.
No matter the implementation, I think the pioneers of the brain analytics front holds the human conscious to a high standard. If the motivation is about fit, then positive possibilities are endless for player health care and performance. One area to keep an eye out in the future would be how the development of machine learning amplifies how we understand a basketball player’s mind and decision-making abilities.