Does a quartet of European economists have an answer to solving the NBA’s tanking dilemma?
Perhaps. It’s an interesting, if technical, proposal that I’ll explain more below, but first, let’s discuss how we got here.
As the tanking discussion threatened to overwhelm the NBA’s All-Star Weekend — at least until the players saved it on Sunday — NBA commissioner Adam Silver’s annual All-Star Saturday news conference made it clear that the league is focusing on much bolder remedies. Notably, he said the league is examining “every possible remedy” to address the issue and backed it up by telling general managers that the league plans to make rule changes ahead of next season.
Advertisement
In particular, the league faces a thorny question: How do you redistribute talent to the worst teams in a draft without incentivizing teams to out-awful their rivals?
“The fundamental theory behind the draft is to help your worst-performing teams restock and be able to compete,” Silver said. “The issue is, if teams are manipulating their performance in order to get higher draft picks, even in a lottery, then the question becomes … are they really the worst performing teams?
“My sense is talking to GMs and coaches around the league that there’s probably even more parity reflected in our records. And that goes to the incentive issue. … It’s a bit of a conundrum.”
What Silver means by this is that, if you chart team win-loss records by season, you get an interesting phenomenon: a bulge of teams that win 40 to 50 games and a paucity of those who win 30 to 39.
In the four seasons since 2021-22, we’ve had 44 of the former and only 23 of the latter, and the disparity looks set to be even worse this season. The Milwaukee Bucks and Chicago Bulls may be the only teams that end up in the 30s, while 10 or more teams land in the 40s.
Meanwhile, in 2023-24, seven teams lost at least 55 games, and in 2021-22 and 2024-25, six did. For comparison, we haven’t had a season where seven teams won at least 55 games since 2010-11, and we’ve only had six twice in the 11 full seasons since.
That win distribution is strong evidence of a trend line toward tanking: The majority of teams are “trying to win” and end up with 40 wins or more, but those that aren’t quickly pivot to “trying to lose” and plummet into 50-loss territory. (Or they’re the Bulls, headed for a third straight year with a win total in the 30s and roundly criticized for not pivoting to a hard tank.)
Clearly, this is an issue. If eight teams are tanking to at least some extent, it doesn’t just affect those teams. After all, they have opponents, and every game involving any of those eight basically has the words “don’t watch this” stamped next to it. That turns the last six weeks of the NBA regular season — which should be the build-up to the playoffs — into a series of hugely forgettable games that are only interrupted when two playoff-caliber teams meet and remind us how much better this could be.
Advertisement
What can we do about it? Well, the most radical idea is the one proposed by our David Aldridge: abolishing the draft entirely.
I’m partial to this idea, but I think the league would be extremely reluctant to pursue it for a few reasons. First of all, the draft is an economic engine that generates huge amounts of interest in the offseason, especially for fans of teams that either miss the playoffs or get bounced early, which is the vast majority of them.
Also, the fact that you can trade future draft picks allows for a huge universe of transactions that would become much more difficult without their presence, and the transaction cycle is another major interest driver for the NBA.
Finally, there is the realpolitik of the National Basketball Players Association and the collective bargaining agreement. Abolishing the draft would seemingly transfer more salary money from current players to future ones. Given that the future players can’t vote and the current ones can, the NBPA would seem reluctant to agree to such a transfer.
So, let’s back up, start small and build our way up. Over the years, I’ve heard several bad ideas to limit or stop tanking where the cure is far worse than the disease. But I think I’ve finally come across a few good ones.
Idea No. 1: The easiest fix
Let’s pick the lowest hanging fruit, sitting there just waiting for Silver to pluck it: Eliminate traded draft pick protections between Nos. 4 and 14.
Nearly all of the most egregious tank situations in NBA annals — from Mark Madsen chucking 3s for Minnesota to Golden State moonwalking the standings into the Harrison Barnes pick in 2012 to the Mavericks’ late 2024 swoon that saw them fined $750,000 to the Jazz and others this season — have been the result of a protected pick between Nos. 4 and 14, which incentivized a lottery team to lose exactly enough to preserve the pick.
Advertisement
I’ve heard arguments against this route from execs. One argument is that it’s a restraint on free trade, so to speak, but c’mon now, the entire NBA is a restraint on free trade. More substantially, this change would reduce the value of teams’ tradeable picks a bit and eliminates the potential for creative deals such as the one struck this February between Indiana and the LA Clippers.
It also discriminates among teams, because not all draft picks are created equal. You would value a future first-rounder from the Oklahoma City Thunder very differently from one from the Sacramento Kings, and protections are a way to adjust the sliders on those values.
However, the big picture has to win out here. Preserving the value of the league’s product is orders of magnitude more important than those considerations above. Regardless of what else the league does, this one change seems like a no-brainer.
Idea No. 2: The radical rethink
Since everything is on the table, let’s talk about everything being on the table.
One of the most interesting ideas I’ve seen actually came from my brother-in-law, Joe Lenski, who doesn’t watch much basketball but consumes a lot of European soccer.
He noted that a few soccer leagues with relegation, such as those in Scotland and Switzerland, play “split seasons,” and the same concept can work for draft lottery odds. Basically, the first two-thirds of the season would play out with a normal schedule, and after that point, the league bifurcates: The top 20 teams only play one another the rest of the way; the same goes for the bottom 10 (or bottom 12 after expansion).
Here’s the twist: Those bottom teams improve their draft position by accumulating wins rather than losses, as teams such as Washington or Brooklyn couldn’t have the best draft pick or lottery odds without beating some of their tanking brethren in actual games. It’s like a soccer relegation derby, except without relegation. (I’ll note that teams such as New Orleans or Milwaukee that already traded their pick might have a bit less motivation, but they’re also not motivated to tank these games.)
Advertisement
The softer version of this would carry over teams’ records from the first two-thirds of the season and then add losses from the first half to wins from the second half; that still might encourage some Process-level, full-season tanking, but it would prevent middling teams from missing the Play-In Tournament on purpose and then crushing the field in the second half to get a top pick. There are several ways to finesse this general concept to address other micro-concerns of that ilk.
I like the kernel of the idea, however, because it solves two problems at once: Not only would the games played by the bottom-10 teams matter and contain genuine interest for the fans, but the last third of the season becomes awesome at the upper end because of all the killer matchups. The worst part of the NBA’s season would instantly become the best part.
Figuring out where to split the season is a minor issue, but the All-Star break is a natural stopping point and also allows the league some time to set up a second-half schedule.
Alas, it also might be a bit too radical, even in the league’s current open-minded state. So let’s take it down half a notch.
Idea No. 3: The great tweak
Onto our European economists and a third idea. A paper that will be published imminently by four authors — Kenneth Colombe of the University of Bonn, Mats Duys and Elias Tsakas of Maastricht University and Tim Pawlowski of the University of Tubingen — outlines a novel methodology to address tanking. (Here’s a link to the executive summary of the research; it’s the second link down. I’m told the full paper, entitled “A Performance-Weighted Allocation Mechanism for Repeated Contests,” will be released within weeks.)
The formula is a bit, um, math-y, but the gist is that it rewards the worst teams with high picks while still encouraging teams to win as many games as possible. It requires no changes to the schedule or to the lottery or really to anything other than how we assign the lottery odds.
The genius of this, mathematically, is that it simultaneously rewards wins but also a team’s position in the “negative standings.”
The basic equation that determines everything is:
P = W − k(R − c)2
That may look like gibberish to you, but it’s the E=mc2 of tanking research.
Every team gets lottery points (P) based on how many games it wins (W) minus a quadratic penalty for its ordinal rank among the 30 teams (R). Those lottery points then get summed up across all teams and converted into percentages for the draft lottery. If, say, the Wizards have 30 “lottery points” and the league’s bottom 14 teams as a whole have 300, their odds are 10 percent.
Advertisement
What “k” and “c” do here are act as slider dials for the league to toggle, based on how much it wants to reward competing versus how much it wants to still target the worst teams for the best lottery odds.
Increasing “k” incrementally increases the odds for the worst teams but also increases their reward for tanking. Changing “c” has a similar effect but is mainly there to ensure no lottery teams end up with negative points (possible with, say, a 44-win team that has the 14th-worst record).
The paper simulated 1,000 NBA seasons using a “k” of 0.25 and a “c” of 1 and found, on average, that this produced a relatively flat lottery-odds distribution between the second- and 10th-worst records but somewhat worse odds for the team with the worst record.
To get a better idea of how this works, I talked to Colombe (a former Pacers intern!) about a few specific recent NBA seasons and how this as-yet-unnamed process would work.
For instance, setting “k” to 0.25 and “c” to 1 would give the 10-72 “Process” Sixers of 2015-16 the 12th-best odds! On the other hand, the relatively tank-free season of 2021-22 would end with virtually flat odds among the worst eight teams, although Houston’s league-worst 20-62 record would land it eighth.
In contrast, for the tank-free 1985-86 season — where no teams lost 60 games and all but three won 30 — Cleveland (29 wins), Chicago and Golden State (30 each) and Indiana (26 wins) had the four best odds, in order. The worst team, the 23-win New York Knicks, was not in the top four.
Tweaking the value of “k” can tilt the odds more toward the worst teams. Let’s use those Process Sixers as an example; with “k” at 0.5, the Sixers move from 12th to eighth; with “k” at 1.0, they go to fifth.
I’ll note that some elements of this system remain slightly gameable, particularly for teams that have the fourth- or fifth-worst record, and that increases as you increase the value of “k.” However, so much of it depends on a team’s standing relative to other teams that it’s difficult to work out in real time.
Advertisement
One added element could make it even harder: You don’t have to use the entire season for the “W” in the equation above. The league could, for instance, decide to only count the last two meetings between every team, thus putting more emphasis on the tail end of the schedule and making it even harder to game the system. Such a maneuver would particularly punish the shameless second-half “pivot to tank” maneuvers we’ve seen in recent years; sadly, my 2017-18 Grizzlies’ 4-29 blitz to the finish line would likely be among the victims.
I’m told the league is aware of this paper, although I have no idea at this point how seriously it is considering the proposals contained within.
I think the NBA should take it seriously. The complex formula may make it hard for casual fans to understand, and the league should carefully means-test it against past seasons to learn more about what values of “c” and “k” would produce the most desirable outcomes.
But in general, I think this would be a big step forward. While you can’t fully incentivize winning for the league’s worst teams without some form of relegation, which is a non-starter in the NBA, you sure as heck can disincentivize losing. This proposal does so as elegantly and painlessly as any I’ve seen.
