Research from T.J. (Timothy) Highley, Ph.D., chair of La Salle’s Math and Computer Science Department, and computer science majors Tannah Duncan, ‘27, and Ilia Volkov, ‘27, proposes a way to get rid of tanking, solving a longtime problem for the NBA draft.
Timothy Highley, Jr., Ph.D., chair of La Salle’s Math and Computer Science Department and associate professor, and computer science majors Tannah Duncan, ‘27, and Ilia Volkov, ‘27, started their work as a summer 2025 research project, that has culminated in the preprint Carry-Over Lottery Allocation: Practical Incentive-Compatible Drafts.
A three-person research team from La Salle University has found a potential way to solve a problem that the NBA has been facing for decades: teams tanking games to get a better position in the draft.
Timothy Highley, Jr., Ph.D., chair of La Salle’s Math and Computer Science Department and associate professor, and computer science majors Tannah Duncan, ‘27, and Ilia Volkov, ‘27, started their work as a summer 2025 research project, that has culminated in the preprint Carry-Over Lottery Allocation: Practical Incentive-Compatible Drafts.
“When new players enter the league, they do so through something called a draft where, generally speaking, the worst teams get first pick,” Highley explained, adding that this is done on purpose to promote competitive balance. “But that creates the incentive for teams to lose on purpose.”
The NBA draft system currently enters all non-playoff teams into a lottery for a chance to win the first four draft positions. The worse a team’s record for the season is, the higher odds they have of winning the lottery. Teams that don’t win one of the top draft picks are then put in order per their record for the season, with the worst remaining team being slotted in at number five. This creates an atmosphere where teams may strategically aim to lose games to improve both their chance of getting one of the top slots, and if not, raising their place in the overall draft. This is known in the league as tanking.
Tanking, or when teams either shut down or sit players for games, is something that the NBA has been trying to prevent for a long time. Most recently, in Dec. 2025, the league began gathering input from key figures to figure out ways to stop it from happening.
“That’s the problem we want to solve,” Highley said. “We want to help the bad teams for competitive balance without inducing teams to try to be bad.”
The way to do this, they propose, is to use the Carry-Over Lottery Allocation (COLA) Draft Mechanism. This method, identified by the authors in the paper, creates a draft environment where teams won’t benefit from losing games.
“We are, to my knowledge, the only proposal that will allow you to give advantages to the worst teams without inducing tanking,” Highley said. “That has been the tradeoff that people have been trying to choose one or the other for decades, and we can do both.”
To do this, they said, there needs to be a paradigm shift in how teams are classified as unsuccessful. Instead of the current process of looking at the most recent season, they consider playoff results over several years.
“We look at playoff track records over the previous several years, where if you get to the playoffs, that means you’re good, if you don’t make the playoffs, that means you’re bad,” Highley said. “But anybody who doesn’t make the playoffs is equally bad. You don’t get any bonus for having a worse record than somebody who just missed the playoffs, and that means that when somebody misses the playoffs, there’s no incentive to lose more because it doesn’t help you.”
COLA works by using a draft lottery where every non-playoff team receives the same number of lottery tickets.
“COLA achieves incentive compatibility by giving all non-playoff teams the same number of lottery tickets each year, eliminating any advantages from additional losses,” the authors said in the preprint. “COLA rewards the weakest teams through carry-over: lottery tickets that do not win a top draft pick are retained for future lotteries, while playoff success or winning a top pick diminishes a team’s accumulated tickets.”
Their mechanism fits three important draft criteria, they say. First, it is practical and can be implemented in the real world. Second, it promotes anti-tanking. Third, it favors the worst teams.
“The core of our system is putting together two things that previously were impossible,” Highley said. “That is, you want to advantage the bad teams, and you also want to not encourage people to be bad by tanking, and that is actually impossible if you’re looking only at end of year standings.”
This impossibility comes from research presented in 2020 at the MIT Sloan Sports Analytics Conference by two Stanford University Ph.D. students, Evan Munro and Martino Banchio.
To make it possible, Highley, Volkov, and Duncan changed the definition of worst to how COLA presents it, meaning that both can be done.
The team applied COLA to a number of different possible scenarios, such as strong or weak draft classes, to show how their mechanism would still work. Some of these scenarios do, however, use the element of the mechanism that Highley believes people could object to the most: a media survey that will help identify when and how the line that defines which teams will be included in the lottery needs to be moved, this scenario is also addressed in the preprint.
“COLA demonstrates that the tension between rewarding weak teams and eliminating tanking incentives is resolvable,” the preprint says. “The COLA Draft Mechanism provides a solution to a problem that has plagued professional basketball for decades.”
For the student authors, Duncan and Volkov, being part of the research process has been a great hands-on learning experience.
“I think for me it’s been a really good learning opportunity. As we said, it was meant to be a summer thing, but I feel like we learned so much, and we’ve made a lot of progress during the summer, so it was exciting to continue,” Duncan said. “Now that we finally have something out there and something concrete, it’s rewarding.”
Both students were involved with all parts of the research, but had a particular focus on the coding and simulations involved in validating the COLA mechanism.
“It was exciting for me the past week because I was building everything for all the graphs and statistics,” Volkov said, noting that the team built the project from scratch. “It was interesting to see the results and all those graphs, to see that my simulation actually works and gives good results.”
-Naomi Thomas
The post La Salle University research identifies potential new mechanism for the NBA draft appeared first on La Salle University.
0 Commentaires