The NBA draft lottery is tomorrow. Thus, before I get to my annual thoughts on NBA draft prospects, I wanted to address a related issue: the NBA’s tanking epidemic. I define tanking to occur when a team makes roster or lineup decisions based on a desire to lose rather than a desire to win. To some degree, tanking exists in every professional sport with a draft. It’s the natural and probable consequence of awarding teams with worse records higher draft picks. However, the relative importance of top draft picks in basketball compared to other major sports makes the NBA particularly susceptible to the sort of drawn-out, pervasive tanking we saw this year, when a quarter of the league sought to lose, at least from an organizational perspective, for much of the season.
Back in February, the bottom 8 teams in the standings—each within 2 games of the league’s worst record—had a combined 44-game losing streak. That did not simply happen by chance. Mark Cuban had recently admitted that losing was his team’s preferred outcome, and he certainly was not alone among NBA owners and GMs in that sentiment. The structure of the draft incentivizes bad teams to lose as much as possible, as is also true for the NFL, MLB, and NHL. Yet only the NBA sees teams spend half the season making a concerted effort not to win. Over the final 3 months, Phoenix, Memphis, and Dallas—the teams that now have the best expected draft picks—went a combined 23-98 (19.0% winning percentage), after going 44-81 (35.2%) to start the year. The Grizzlies lost 19 straight at one point. The Suns lost 30 of 32, with those 2 wins coming against the Grizzlies and Mavs. Each of these teams routinely rested or benched some of their better players down the stretch to make losing more likely, as did several of the teams “chasing” them for the league’s worst record.
Tanking on such a large scale is bad for the league, in part because being a fan of a tanking team is a frustrating experience. You’re left with only unpleasant alternatives. You can continue rooting for your team to win, despite knowing that winning is against its interests and objectives, or you can root for your team to lose, which is harder than it sounds. There’s a psychological disconnect in cheering against your favorite team that’s difficult to overcome. Most likely you just won’t watch your team very much. As a fan of the tanking Bulls, I’m speaking from experience. Tanking is also bad for the league because it generally lowers the quality of play and reduces the likelihood of a competitive game, which can lessen consumer demand and cause financial harm to non-tanking teams. When the Sixers candidly aimed to be terrible for multiple years, other owners were upset not out of concern for the Sixers but rather because their home dates against the Sixers became unmarketable.
So what should the league do about this? The existing draft lottery is itself a mechanism designed to discourage tanking—absent in other major sports leagues—that denies the team with the league’s worst record the #1 overall pick as a matter of right. Instead, all non-playoff teams enter a lottery for the #1 pick, with odds weighted to favor teams with worse records. That lottery currently extends to the first 3 picks in the draft, with subsequent picks assigned according to team record (worst to first). Here’s a table reflecting the 2018 draft lottery setup:
Any changes to the NBA draft lottery system or draft order require 75% approval of the NBA Board of Governors, which includes a representative from each of the 30 team ownership groups. Therefore, drastic proposals that would eliminate any benefit from tanking but also eliminate or significantly limit the competitive balance aspect of the draft are likely to be non-starters. Suggestions to eliminate the draft altogether or to assign teams predetermined draft picks untied to team record, known as the draft “wheel” for the way teams would cycle through draft slots, have not gained traction.
As you may know, the NBA Board of Governors agreed to make some changes to the draft lottery after this year to lessen the incentive to tank. In sum, the league is both flattening the lottery odds somewhat and extending the lottery to 4 picks rather than 3, so the team with the NBA’s worst record could pick as low as #5 as opposed to #4. The team with the worst overall record also is most affected by the flattened lottery odds; its chances at the #1 pick will drop from 25% to 14%. Here’s a summary of the differences between 2018 and 2019 lottery odds:
I am skeptical that these changes will have much effect, for a simple reason. In terms of maximizing draft outcomes, losing as many games as possible is still the dominant strategy for every team.
Tanking teams find losing preferable to winning in order to achieve the dual outcomes of (1) increasing their odds of landing a top pick in the lottery and (2) improving their draft position otherwise. While on the surface the league is addressing both primary motivators by reducing the top lottery odds and extending the lottery an extra pick, the changes do not meaningfully reduce the incentive to tank; the expected payoff from tanking will still be higher than the expected payoff from any other strategy. No team will have better lottery odds than the team with the league’s worst record, and that team will still have the first pick in the non-lottery-awarded draft order. As a result, losses are still strictly preferable to wins in determining a team’s expected draft position.
The absence of an effective deterrent is further demonstrated by the graph below:
Although the new format will result in lesser expected draft outcomes for each of the bottom-5 teams, it is still clear that having a worse record always leads to a better expected pick. Whoever finishes with the league’s worst record in 2019 may be disappointed that its expected pick is down to 3.66 from 2.64 in prior years, but that’s still better than being in any other spot.
To actually deter tanking, more substantial changes are needed. I now present 3 more effective options:
(1) Lottery Relegation
The concept of relegation is best known from the English Premier League, where the bottom 3 teams each year lose their EPL membership and are “relegated” to a less prestigious league for the following season. Under my NBA draft lottery analogue, the bottom 3 teams would lose their right to participate in the lottery and be “relegated” to a fixed draft position outside the top few picks. The top lottery odds would then shift down 3 slots record-wise. Using the existing framework as a guide, the lottery odds could look like this:
Beyond the absence of ping-pong balls for the teams with the 3 worst records, I propose that the relegated teams draft in reverse order to preserve the desired anti-tanking incentives. In other words, the league would implement a 3-2-1 draft order as opposed to a 1-2-3 draft order immediately following the draft slots determined by lottery. That way the team with the worst overall record would always draft after the teams with the second- and third-worst records. Given a 3-pick lottery drawing, the league’s worst team would have a fixed draft position of #6 overall, following the second-worst team at #5, and third-worst team at #4. Here’s what the expected payoffs look like under this model:
Now the team with the 4th worst record has the best expected pick. The league’s worst team suddenly finds itself in an inferior position to everyone else among the bottom half of the non-playoff teams. Meanwhile, the teams that barely missed the playoffs are very minimally affected in terms of their expected picks, so they still lack an incentive to tank their way out of a playoff spot.
You may notice that, under the proposed Relegation system, the graph of expected draft outcomes somewhat resembles Nike’s logo, the “Swoosh,” which appears on Ben Simmons’s jersey in the photo at the top. That’s roughly the shape that provides optimal incentives against tanking while still affording an advantage to lesser teams to promote competitive balance. It’s also appropriate given Nike’s billion-dollar apparel contract with the NBA, which I discussed in a previous post on suggested changes to the Collective Bargaining Agreement. [See Section 2 on scheduling and the financial impact of reducing the number of regular season games.]
The Lottery Relegation model very clearly deters teams from trying to secure the league’s worst record, as no team or fan base wants to endure such a season in exchange for a guaranteed payoff of the #6 or #7 pick in the draft. That pressure to avoid finishing at the very bottom of the standings pushes would-be-tanking teams to be better, which should foster a higher baseline level of competition.
Late in the season certain non-playoff teams will still have an incentive to tank, and that’s okay. Limiting tanking to a smaller subset of teams for a shorter period of time would be a major net gain for the league, and such conduct would be offset in a way by teams at the bottom of the standings aggressively trying to win late in the year.
(2) Rainbow Lottery Odds
If lottery relegation is perceived as too punitive, an alternate option is to have escalating rather than declining lottery odds through the bottom portion of the standings. Then a typical non-playoff team will have a better shot at the #1 overall pick than the league’s worst team, so regardless of the non-lottery draft order there’s a downside to tanking. Here’s an illustration:
In this scenario, the league’s worst team has a lower probability of winning the lottery than any other team in the bottom 10. Teams with better than a 10% chance are those in the 4-8 range as opposed to those in the 1-4 range under the current setup. In addition, the rainbow is slightly skewed to dissuade playoff-caliber teams from seeking to tank their way out of the postseason.
(3) Checkmark Non-Lottery Draft Order
As noted earlier, there are 2 major benefits to having the league’s worst record: (1) maximizing the odds of landing a top pick in the lottery, and (2) guaranteeing the top pick in the non-lottery draft order. This final proposal deters tanking by removing that latter guarantee and rather replacing the existing draft order with one much less favorable to those at the bottom of the standings.
The new draft order would zigzag, or radiate out, starting from the 4th or 5th worst team. I’m going to call the net result a checkmark based on the following visual:
Here, the league’s worst team has the #7 slot in a non-lottery draft order that proceeds according to team record as follows: 4, 5, 3, 6, 2, 7, 1, 8, 9, 10, 11, 12, 13, 14. This method is the least straightforward and intuitive of the 3 options, though it substantially discourages tanking compared to the current draft order. Paired with reasonably flat lottery odds like the 2019 setup, this would greatly diminish the expected value of the worst team’s pick, such that its expected draft position would be inferior to each of the handful of teams directly above it in the standings.
Any of these proposals can be tweaked to combat tanking more or less aggressively. No matter the precise lottery odds or draft order, reasonable methods exist to incentivize bad teams to prefer winning over losing. But until the league implements such a system, tanking will remain the dominant strategy for any team looking to make the most out of the draft.
Leave a Reply