How to End NBA Tanking

If you don’t watch the NBA, you’re probably completely unaware of how bad tanking is getting. (Tanking, if you don’t know, is losing on purpose to get a high draft pick in the draft lottery in the spring.) Tanking is not only getting worse but it’s getting worse despite the draft odds being flattened - i.e. the worst record in the league gives you less good odds of getting the number one pick than it used to.

Some teams are currently resting their starters in the fourth quarter. That is, in addition to claiming their starters need rest and not starting them, or diagnosing phantom injuries, or having them get mid-season surgeries for things that could be dealt with in the off-season, some teams are letting them play three quarters and then sitting them during competitive games. This began in early February, i.e. 2 months before the end of the season, about as early in the season as anyone can remember tanking this egregious. (Unfortunately, it was the Raptors who at least partly pioneered this approach last season.)

Tanking is bad because fans (most fans?) want to see their teams win. And it’s especially bad for people who pay for tickets. It’s one thing if you never see a game to say “I don’t care, I just want a high draft pick.” But why would anyone pay for tickets if they know the team is trying to lose? Bill Simmons made the point on his podcast last week that this is the time that the NBA should be trying to show it’s best competitive basketball, because the NFL just ended, and there a ton of sports fans who may or may not watch the NBA who no longer have NFL games to watch. It’s in the best interests of the league for the February games to be good.

So tanking needs to be fixed. How to fix it?

The Real Way to Fix Tanking

There is a solution that will fix tanking, it is just extremely unpalatable to many NBA owners.

That solution is to detach the draft from the standings, so that a team’s record in a season will no longer impact where they draft in the summer.

I was first introduced to “The Wheel” in 2013 by Zach Lowe when he was still at Grantland. To me, it is the clear solution to this problem. The idea is that each team gets each draft position once in a 30 year cycle. So one year your team drafts 13th and the next it drafts 21st and the next it drafts 2nd until every team has received every pick once. These picks are all determined and known in advance. You can read the full details at the link above but this would end tanking.

However, “small market” teams in particular don’t like this proposal because they don’t think they can attract free agents and they worry that if they don’t have any good picks for quite a while, they will no longer be able to get better. Plus, as Zach notes, to like this proposal, you have to be okay with the best team in the league being scheduled to get the first overall pick. Many people are not comfortable with that.

I love it still and I think it solves the problem but I’m under no illusions and I think it will never be adopted.

Zach Lowe introduced a different, less radical idea on his podcast last week: the “conclave.” This is the idea that NBA owners (or management) vote on who should get the picks each season. So, all 30 teams would meet and they would rank all 30 teams by draft order, with the stipulation they cannot put themselves first. (They’d all put themselves second, then, right?) The format that has 50%+1 of the vote is the draft order that year and if they don’t have a proposal then they’d have to keep voting. (They could eliminate the least popular ideas.)

I don’t love this because I think it' would be determined by NBA ownership/management politics. My worry is that teams would vote for the teams/owners they like least for the 30th spot and if there was enough of a consensus then it’s possible that a “small market” team would consistently get that spot despite sucking and needing better picks. (I’m thinking of Vivek currently but there have been plenty of NBA owners in the past who would have fit this bill, such as the Clippers’ former owner. This system would have killed the Clippers for decades though you could argue he didn’t need any help.)

Again, I think this proposal is unlikely to be adopted, though it is way more likely to be acceptable than The Wheel.

If we have to keep the draft lottery and keep it tied to the standings, I think there are at least three ways we can make it very hard to tank effectively. And if tanking becomes incredibly, obviously ineffective, it will mostly disappear.

1. End All Draft Pick Protections

If you don’t follow the NBA, you likely do not know about the insanity of draft pick protections. Currently, when trading draft picks, teams are allowed to “protect” where the pick they are giving up falls in a future lottery. They can do this is all sorts of ways that make figuring out who is drafting where quite complicated. So, for example, they can protect it if it’s in the Top 4 but they can and do also protect it if it’s, say, in the bottom 10 (i.e. 21-30). Why? Because teams have some idea of where they will likely draft in the future and so they want to create protections beneficial to them. They do this more and more.

On its face, it’s gotten absurd. It might have made sense once but it’s getting more and more complicated. But it’s actually encouraging tanking and the reason is that it’s encouraging teams who aren’t bad enough to get the #1 pick in the current “flattened” draft odds lottery to tank. Multiple teams in the last few years have had picks that were, for example, protected Top 8, meaning they keep the pick if it lands in the Top 8 but lose it if it’s 9 or worse. What this results in is mediocre teams who are not bad enough to win the lottery tanking for the 8th worst record in the league, in this example.

Some people want to just limit protections to some fixed number going forward, such as Top 3 or lottery-protected (i.e. Top 14). I don’t think that’s going far enough. Draft protections make trades confusing for fans and any limits on protections will still be gamed by teams just like the current rules. The only solution is to get rid of them entirely.

2. Fines Must Have Teeth

Currently, whenever the NBA is embarrassed enough by tanking, they fine teams a few hundred thousand dollars. The problem with this is that the owners of these teams are billionaires. So it’s the equivalent of fining one of us a few dollars or something. It has no effect.

Tom Ziller recently introduced a better fine system that will have real effects: removing lottery combinations. I love this idea but I also think it hasn’t been talked about enough which is why I’m mentioning it here even though I think it’s more of a backup plan to the real solution (below) than an actual solution. Basically, instead of or in addition to fining teams for tanking in dollars, the NBA removes some potential lottery combinations for that team every time they tank. So if you tank once, the league takes away 1234 and next time they take away 5678 etc. (These would be random numbers in reality.) They might have to do many more than one combination per fine but I don’t want to do the math to figure out how.

The league would have to be very willing to levy these fines, though, to make this effective and, given how long it takes to fine tanking teams now, I’m not sure they would be willing to do this enough. But I think it’s a great idea and, in combination with the removal of pick protections above, and true solution below, I think it could help to make tanking incredibly ineffective if enforced properly.

3. Teams Cannot Know Their Draft Odds Mid Season

Bill Simmons introduced his “Entertaining as hell tournament” in 2007. The idea has inspired many other ideas but the gist is that many people agree that one way to fix tanking is to, at some point, make wins count more than losses for teams destined for the draft lottery. I.e., if you’re the worst team in the league, at some point you need to win games not lose them in order to keep the odds of getting the best pick at the end of the season.

I’ve always liked the idea of bad teams having to play for the best picks ever since I first learned about this but with the NBA Cup and the play-in tournament both already happening I think turning this into a tournament would only further complicated the NBA schedule.

More importantly, teams will still game this out. The worst teams will still tank before the tournament or before whichever game is chosen to be the demarcation point between a bad record giving bad draft odds and wins improving those odds. If the league says that teams that are in lottery position as of game 50 need to win games in order to keep their draft lottery position, they will just tank before game 50, which just shuffles the problem to an earlier part of the season.

I think there’s an obvious solution here. I didn’t invent it. I’m pretty sure I heard it on a Simmons podcast years ago but I haven’t heard him talk about it so I’m resurrecting it. (It’s also possible I heard it somewhere else.)

The solution is that nobody knows the demarcation point.

For example: every October, Ernst & Young and a member of the league office randomly generate a number between 21 and 60 and this is put in a box. (The number is generated analogue and is not online so only the couple of people in the room know the number. That is key.) And it has to be absolutely random, no weighting. And it’s also important that it can indeed be the same number as the year before. That helps make it harder to game.

For the first 20 games of the season, the lottery odds stay the same (or we go back to the old odds or whatever): the worst team has (roughly) the best odds to get the number one pick.

For games 61-82, wins count for any team that would fall in the lottery under the current system, so the worst 14 teams in the league (even though making the play-in). The team that wins the most in this period gets the best odds. And the two sets of odds are combined.

But, crucially, between game 21 and game 60, nobody, not even the league, knows which games will count under the traditional “worst team gets the best odds” system and which games will count under the “bad teams need to win to improve their lottery odds” system. The only people who know are the people who were in the room.

The night after the last team the league has played its 60th game, the box is opened and the league announces the demarcation point. So, say the number was 57. As of game 57, then, teams who would be in the lottery at that game need to start winning. And each year the number would be unknown.

What this means is that tanking is basically impossible. With the odds of this demarcation point being totally equal for game 21 through game 60, it could literally be any single one of these games at which point teams need to start winning. Teams can bet that it will be, say, game 41, but they could be wildly wrong, meaning they could have seriously hurt themselves if they choose to tank between game 41 and 57 if the number is actually 57.

Additionally, the league can still fine teams for tanking meaning any team dumb enough to try to tank under this system would still lose lottery combinations and hurt themselves even more. My guess is that this would only take a year or two to make tanking so ineffective it disappears.

It might not work, of course. As long as the draft is tied to standings, teams will find ways to game the system. It’s possible teams would figure out how to game this system too. Maybe it’s even likely. I still think this idea is much, much better than the current system and would work better than any proposal where teams actually know which game they have to start winning from.

Bonus: The Draft Beauty Contest

If this doesn’t fix tanking, there is one more idea I like that would help teams be on their best behaviour, one Simmons suggested last week on his podcast.

Most drafts, there are a few players considered to be potential all-stars or better. Nobody knows for sure, obviously, but usually there is a consensus that it is a “3 player draft” or a “5 player draft” or what have you. Some drafts are deeper and some drafts are shallower.

Let’s say we decide that the average draft has at least 5 players that every team in the league would want for their team. (It has to be an odd number of avoid ties.) The top 5 teams in the lottery (receiving that position through a combination of bad record to start the season and wins to finish the season) list out their 4 preferred players for the draft.

So, last year, that would have likely been Cooper Flagg first, Dylan Harper likely second, and two more guys third and fourth. (There was debate about the next players and when they should go.) So if the vote is Flagg is #1 pick of at least 3 of the 5 teams, he would choose first and if Harper were the #2 pick of at least 3 of the 5 teams, he would go second.

By “go” I mean they would pick their team. Simmons’ idea is that the very best players in the draft each year pick the team they want to go to. So Flagg would have chosen one of the top 4 teams in the lottery and Harper would have had his choice of the remaining 3.

Simmons believes, I think rightly, that this would encourage teams to present themselves as “well run” and as with hope for the future. He thinks this would discourage outright tanking and I tend to agree. But I don’t think it would completely get rid of tanking which is why I think it is only effective as an additional step to adjust the above proposal if that fails to eliminate tanking on its own.

Anyway, there you have it. This could be done next season if the league really wants to.