Unless you’ve been living under a rock this week, you know that the NBA is now considering a draft lottery system that smooths out the odds for all teams participating in the lottery and favors teams who finish in the middle of the lottery order instead of the teams with the best and worst records in the process. We’ve talked about this plenty. The initial announcement is here. A discussion on how this could affect the Portland Trail Blazers is here. Upon reflection, though, there’s one more thing
to be said. And it’s important.
For those who missed it, the nutshell version of the new proposal goes like this:
- 16 teams participate in the lottery: the 10 who do not make the postseason in any form, the 9th and 10th teams from each conference who participate in the Play-In Tournament, and the loser of the Play-In 7th vs. 8th game.
- Teams get either 3, 2, or 1 ball(s) in the drawing, based on finish. The 7th-8th loser gets 1, the 9th and 10th teams plus the worst three teams in the league by record get 2, and all teams who finish in the middle of those extremes get 3.
- All 16 lottery picks are drawn for, not just the Top 4 as currently exists. Protections are in place for the worst three teams to fall no lower than 12th.
- Rules prevent teams from earning the first-overall pick in consecutive years and/or Top 5 picks three seasons straight.
I agree with the potential for this new system. It’s actually a small variation on a scheme that we suggested twice, both times back in February when other potential “fixes” were being floated. It acknowledges tanking and tries to mute its effects without completely getting rid of the idea that the draft is meant to provide help to struggling teams. It takes away the incentive to lose as much as possible by slightly decoupling losing and reward. You don’t get any more benefit from finishing fourth-worst in the standings than you get finishing eighth-worst. You get less benefit if you’re bottom-three. This could actually lead to horrible teams trying hard to win games towards the end of the year, bringing intrigue to matchups that otherwise would have been ridiculous tank-fests.
You can read more about the rationales for the approach in those links to earlier posts just above, so I won’t rehearse them here. I will say this, though. The NBA is making one, big mistake that threatens to subvert the whole system.
Two Tools
To understand the problem, we need to look at how a couple of the mechanisms work.
The NBA has two big tools to combat tanking within the lottery system:
- They can change the odds for each spot. Differentiating the odds (giving worse teams a better chance at promotion) incentivizes tanking. Smoothing out the odds—making them relatively equal for all participants—takes away that incentive. The more you smooth, the less tanking matters.
- They can change the number of picks drawn for in the lottery process. This gives each team more chances at getting promoted even though their individual odds are the same.
The smoother the odds and more chances at promotion I spread around the field, the less important getting a bad record is. Anti-tanking! That’s the good part.
Here’s the bad part. The more I spread those odds and chances across the field, the more randomness I introduce into the process. And randomness brings the chance of sub-optimal results like the San Antonio Spurs winning all of their ultra-high picks or the Dallas Mavericks getting Cooper Flagg #1 overall in 2025, leapfrogging a lot of teams that needed help more.
Making the Change
You can see why the NBA would want to alter the current system if combating tanking is their priority. The odds are weighted towards the lousy teams. Franchises with the three worst records have a 14% chance of winning each drawing. A team in the 13th or 14th spot has a 1% or 0.5% chance, respectively. That’s 14-28 times the odds of success for the tankers. Also, the league only draws for the Top 4 picks, then seeds in reverse order of record after. The three worst teams in the league have 20 times the chance of earning a Top 4 pick over the team with the best record in the lottery and can’t fall below 5th-7th, no matter what happens. Weighted odds + Only 4 draws = Incentive to lose.
Obviously the NBA is trying to change this. Make the losing go away!
In this new, 3-2-1 proposal they entered the operating room with a scalpel (smoothing odds) and some forceps (number of spots drawn for), ready to do some surgery. Then they dropped them both in favor of operating with a sledgehammer. The league could have used either tool to disincentivize tanking. Instead they used both, and pretty close to the max, full throttle. Yikes.
Odds and (Bad) Ends
In the new system, only 2 teams out of the 16 lottery participants have to settle for one ball: the two that lose the 7-8 Play-In matchup in each conference. Everyone else has two balls or three. Those aren’t completely smooth odds, but it’s close: the difference between an 8% chance of getting promoted and 5.5%. You can see how tanking gets much less profitable, especially since most of the field is at the 8% level.
But drawing for all 16 picks creates a huge number of trials in which that randomness comes into play. That opens the door for some truly wacky, and tragic, results.
The entire system as laid out involves 37 balls in that lottery hopper. 13 of those 37 will belong to the teams in the 10th-16th positions…the best teams in the field, by record. Theoretically you’d want promotions for those teams to be possible, but rare-ish.
13 is about 35% of 37, though. That means there’s a 35% chance (1 in 3 or so) that a team between 10-16 in the order will get the #1 overall pick. That Dallas Mavericks/Cooper Flagg thing is going to start happening far more often.
That’s not even the bad part. Let’s say one of the bad-record teams does win the #1 pick. Their balls come out of the hopper. That means our 10th-16th-place teams now own 13 balls out of 34 or 35. Now we’re up to a 38% chance of promotion. In the next draw, their chances at getting the third overall pick go up as high as 42%. And we keep going, and going, and going.
Disasters Ahead
This is going to create chaos. You are literally going to see weeping and gnashing of teeth. Suffering teams are going to wonder why these “less-deserving” franchises get ultra-high picks, leapfrogging a dozen teams to get there. Playoffs teams who aren’t involved in the lottery at all are going to ask why a franchise barely below their level gets gifted a superstar or a super-high pick, providing an ability to compete that they didn’t earn.
We need some randomness in the process! That’s the whole “anti-tanking” part. But introducing too much is going to make a farce of the system and imbalance the league in odd ways.
I understand why the NBA is choosing to smooth the odds. That’s an appropriate variable to mess with. I don’t know if they have the smooth-balance right in this 3-2-1 concept, but they’re trying.
I do not understand why they took the extra step of drawing for all 16 positions. When every participant has the same (internal) odds of ending up with the first pick, the last, or anything in between, wacky results are GOING to happen. It makes no sense.
Easy Fix
The easy solution is to draw for fewer spots. Maybe four is too few. I get that. But what about six or eight? You’ve already disincentivized the plunge for the worst records in the league by giving the worst teams in the lottery fewer balls. You could draw for spots 1-8 then seed the rest of the teams from 9-16 in reverse order of record, just like the current system does. I’m pretty comfortable saying nobody is going to lose droves of games for a guarantee of the 7th or 9th position in the draft order.
This would have the following benefits:
- Reduce the chaos and weird results
- Get more help for the truly bad teams without overtly encouraging them to be bad
- Take away some of the benefit of being a good-record team in the lottery under the new system. The way it’s proposed, finishing 9th or 10th in the conference is exactly as beneficial as finishing last. There’s incentive not to get to that 8th or 7th spot, maybe even to not make the playoffs. As we said in an earlier post, you haven’t taken away the incentive to lose entirely, you’ve just moved it to the Play-In/Low-Playoffs tier instead of the bottom of the league. Reverse order by record for the second half of the lottery field lessens that issue.
Still Time
Seriously, I don’t understand why the NBA took a good idea and threw it into overdrive. They’re opening the door for it not working in spectacular and ridiculous fashion. It reminds me of teaching my kid to drive, telling him that he needs to be a bit more prompt and aggressive turning onto a main street from a side one, then watching him fishtail onto the avenue with smoke billowing from his tires because he floored it. DUDE. There is a smart, safe middle ground. Take it!
The Board of Governors doesn’t meet until May 28th. There’s still time to make this relatively-easy tweak to the system before it’s even adopted. If they don’t, I’m going to be watching future lottery drawings with my hands held over my eyes.
Draw from 1-8 or 1-6, then reverse seed. You’re going to have enough randomness, odds-smoothing, and loss-penalizing to nerf tanking without making a mockery of the drawing and watching pundits make fun of your system—and your league—for weeks on end.
