NBA lottery reform: How to fix tanking? Don't give the bottom three teams the best odds

The NBA has been trying to solve tanking for about as long as I've been alive. Smoothed odds. Weighted odds. The ping-pong machine. Flattened top-four picks. Stern threats about "competitive integrity.” Even fines. None of it has worked, because none of it has addressed the actual disease, which is that the structure of the lottery itself rewards being terrible. The league can tinker with the percentages all they want. As long as the worst teams in the league have the best shot at a franchise-altering pick, somebody is going to figure out how to be the worst team.

A few weeks ago the league leaked three concepts that I criticized for their lack of clarity. This week, during the NBA's general managers meeting, the most interesting thing about the call wasn't which concept the league walked in favoring. It was which concept they walked out with.

The favorite going in was the simplest of the three. “Concept #1.” It looks like this:

• 18 teams in the lottery: bottom 10 that miss the play-in + the 8 play-in teams
• Bottom 10 teams have 8% chance each
• Next 8 teams split 20% in descending order
• All 18 spots drawn as part of lottery

Multiple GMs raised their hands and said some version of the same thing: Under this concept, a team sitting in the fifth or sixth seed in March has a genuinely interesting question to ask itself. If they slip to the seventh or eighth seed, they get two cracks at making the playoffs through the play-in and they get a lottery ticket, since all play-in teams are now in the lottery. So a 44-win team in the sixth seed has a real, defensible reason to lose its last four games and slide into the seventh.

Then there's the cliff at the bottom. Under the 8% flat rate for the bottom 10, if you're the 11th-worst team in the league — a play-in hopeful, in other words — you have a meaningfully worse shot at a franchise-altering pick than the team one game beneath you in the standings. The incentive to slide out of the play-in and into the bottom 10 gets even stronger than it is now.

“This solves nothing,” said a general manager that was on the call. “It could make the problem even worse just like the 14% odds did.”

I would agree with that. The league would effectively just relocate tanking from the bottom of the standings to the middle, which we’ve already seen happening with the flattened 14% odds for the bottom three teams. Now teams are happy to get in the middle and see what happens on lottery night. This would bring that to a greater extreme, which is the inherent issue with 18 teams: it creates an incentive for teams in the top six to fall into the play-in, and an incentive for teams at the bottom of the play-in to fall out of it.

Several general managers on the call expressed that they don’t believe this approach will resolve tanking in the slightest. But league sources expect the specifics of this concept to change, which was what much of the conversation centered around during the GM meeting.

What was brought up by multiple teams is whether 22 teams is actually better than 18 teams. The NBA initially mentioned 22 teams in other concepts, and league sources say it's still certainly on the table. Having 22 teams in the lottery would include the 10 teams that miss the playoffs, the four play-in losers, and eight first-round exits. Doing this would require even flatter odds, which could operate as a greater deterrent to tank at any spot in the standings if the reward for doing so is much smaller.

League sources say a general manager pushed back during the call with a concern that teams could try to lose in the playoffs, or there could even be a perception that a team could try to if they shut down an injured player instead of bringing them back for the remainder of the series.

“We’d be moving the perception of intentionally losing from the regular season to the playoffs,” a front office executive said to me. “I don’t think it’s worth the risk.”

What if the bottom three teams had slightly lower odds than the teams ranked four through 10? (Photo by Jacek Boczarski/Anadolu via Getty Images)

(Anadolu via Getty Images)

But others disagree. They argue that it’s unrealistic that any coach or player would ever intentionally lose in the playoffs. And if a team pushed a player not to return because of a hope of getting lottery odds, doing so during the playoffs could lead to catastrophic results with their relationship with the player.

I tend to agree with that perspective. Playoff games are nationally televised, played in front of sold-out arenas. You'd be asking a head coach to throw away the biggest moment of his year, a star player to take an L on his permanent record, and a front office to explain to its fanbase why the team mailed in a winnable series all in exchange for a marginal bump in lottery odds. If a team so badly wanted to tank for lottery odds, they’d do so during the regular season. Not during the playoffs, especially if odds are flattened.

What could flattened odds look like? With a 22-team lottery, odds could look something like: 6% for the bottom 10 teams, 5% for the four play-in losers, 3% for the worst four teams that lose in the first round and 2% for the best four teams that lose in the first round. There would be minimal incentive to be worse, and no dramatic cliffs that increase odds a significant amount from one spot to the next.

But with flattened odds at the bottom, there is still the philosophical issue that has dogged every version of this exercise. The worst team in the league is still in the raffle: 6% is not 14%, and 14% is not 25%. Each round of reform has decreased the incentive to bottom out, but teams still find reason to since even in a worst-case scenario on lottery night, their pick could end up fifth or sixth.

According to league sources, one general manager floated a nuclear option: just make the bottom three teams ineligible for the top picks entirely. Not flatter odds. No odds. If you finish 28th, 29th, or 30th, you cannot win the first pick under any circumstances. The league, per multiple sources, found this to be way too extreme.

But then another person on the call offered a softer version of the same concept: What if the bottom three teams just had slightly lower odds than the teams ranked four through 10? Not zero. Just a little less.

Sources on the call say Adam Silver responded enthusiastically to this idea and that the league is taking it into consideration for the official lottery reform proposal.

I think it’s brilliant. Under that structure, with the bottom three teams having slightly worse odds, there is no longer a single point in the standings where losing helps you. Tanking all the way to the bottom hurts you a bit. It’s not quite relegation that you’d see in the Premier League, but it’s the NBA’s own form that would punish being the worst in the league. And much like Premier League teams have entertaining games to prevent relegation, NBA teams would too.

Picture two bad teams in late March, both within a game of the bottom three, both desperate to win. That's a win for the fans. Picture the front office of the Wizards doing the calculus on whether to shut down Trae Young and Anthony Davis and realizing that, actually, no, the vets need to go play, because falling in the standings is a real cost now, not a reward. That's a win for the sport. Picture Sacramento intentionally fouling Seth Curry late in a game, and the conversation around it shifting from "nefarious tanking" to "bad coaching." That's a win for the league.

Taking into consideration everything I’ve learned, I’d hope for a proposal that ends up looking something like this:

• Bottom 3: 5% odds
• Spots 4-10: 6% odds
• Spots 11-14: 5% odds
• Spots 15-16: 4% odds
• Spots 17-19: 3% odds
• Spots 20-22: 2% odds

The dream version of lottery reform removes every reason to want to lose. That's what this aims to accomplish. The worst teams still have better odds, but not so much better that a front office can justify torching a season to chase them. And if you do finish in the bottom three, you're penalized for it. Because the first-round losers are in the mix too, there's no reason for a fringe playoff team to tank into the play-in, and no reason for a play-in team to tank out of it. Every rung on the ladder is worth climbing, or at least not worth jumping off of. The cliffs are gone. The middle is no longer a hiding place. The bottom is no longer a destination.

Any conversations about flattening odds get immediate pushback from teams about how much more difficult it could be to rebuild. Flatten the odds enough and you turn the lottery into a coin flip for franchises that desperately need a sure thing, which means more teams stuck in the 30-win range for longer, which is its own kind of basketball purgatory. The counter is that we're already living in that world: the 14% reform was supposed to discourage tanking and instead it just created a wider band of teams willing to lose, and the rebuilds didn't get noticeably faster or cleaner. Rebuilding is hard because building an NBA roster is hard, not because the lottery isn't generous enough to the losers.

But it's a real concern, and any honest version of this proposal has to admit that some small-market team is going to spend two years in the basement, walk away with the seventh pick one year and the 11th pick the next. And yet, the league has seen Steph Curry and Shai Gilgeous-Alexander drafted in those slots. Franchise changers can be found all over the draft. Kawhi Leonard went 15th. Jimmy Butler went 30th. Nikola Jokić went 41st during a Taco Bell commercial. The teams that found those guys weren't rewarded for being terrible. They were rewarded for being smart, which is supposed to be the whole point of running a basketball team.

If the cost of this reform is that a few franchises spend a couple of extra years figuring it out the hard way, that's a cost I can live with. Every fix the league has tried has sprung a leak somewhere. This one will too. But it'll be a smaller leak, in a less visible place, in a league that's finally stopped rewarding its worst behavior.