NBA Schedule and Seeding Reform Proposals
One small benefit of The Bubble Season is that it has given people within the NBA and media a new construct in which to discuss ways to improve the league's design. Group phases, play-in tournaments, and relegation have all become part of the universe of imaginable changes that the league could try to adopt.
A month too late, here are my proposals for post-COVID reforms with two simple goals in mind:
Improve the quality of play by incentivizing winning games as often as possible
Only propose ideas with some hope of being approved by the league's governors and the NBPA
Expand to 32 Teams
The league hasn't expanded in 16 years; the longest it had ever gone before this is eight years. Basketball has continued to develop as a worldwide sport, growing the talent pool. Sports medicine has also improved, extending the length of many careers. There's more available talent than ever before and yet the growth of the size of the league has slowed. Adding 2-way contracts helped some with holding more talent in the NBA's sphere but it's not enough.
The league collects expansion fees for adding teams. Profits are taking a significant hit due to COVID and so the governors should be happy for a new infusion of cash. Two new teams add 30 more full NBA contracts so the NBPA wouldn't have any reason to fight it, either.
Going to 32 teams will also make a few other changes easier.
Cut to 76 Games
Many proposals like this go for things like a 60 game season, arguing that the improved quality of play will drive popularity so high that it overcomes the revenue loss from fewer games. This might be true for the national TV deal, where only a relatively small number of games are involved to begin with, but I'm dubious that it would be the case for the local TV deals. In-person attendance is near capacity in most cities already so there isn't any way to recover losses there unless ticket prices sharply increase and I'm simply not going to advocate for that.
Having 32 teams makes for an easy schedule design. The two conferences split into two divisions each. The East is made up of the Atlantic and Central; the West is the Sun Belt and Pacific. Each division is eight teams. Every team plays the other seven teams in their division four times and the other 24 teams twice each.
The schedule is still imbalanced but you get rid of the weird "some conference teams play each other only three times" quirk of the current schedule.
Losing six games is still a potential hit to revenue but it's only three home games. There should still be some uptick in quality of play as "load management" would decrease some and we're also offsetting short-term loses with the expansion fees and other changes.
Keep the 8th Seed Play-in
This is an easy one; just keep the rules from The Bubble. The 8-seed has to win one game against the 9-seed, if there is a 9th seed within four games in the final standings, before the 9-seed seed wins twice.
One question here for a non Bubble league is where the games get played. Ideally I think a neutral venue hosting both play-in pairings (if there are two to be played) would be best. It would give the league, ad partners, media, etc. one more gathering before everyone splits off to the different early round playoff locations. You could even fold in things like G-League Playoff Finals.
In reality, playing both games at the home of the 8-seed is probably what would happen, though I think that ends up being too big of an advantage for 8th.
Top Seeds Pick Their Playoff Opponents
I'm not going 1-16 playoff seeding because I want to build up rivalries as much as possible and the best way to do that is to have teams run into each other in the playoffs over multiple years. Seeding by Conference helps that. It also cuts down on travel, though I'm not as concerned with that benefit.
You know what else builds rivalries while also giving more teams something to play for? Letting the top-3 seeds in each conference pick their first round opponents from the (post play-in) bottom four seeds! Nothing will build a rivalry as quickly as knowing your opponent chose you as the team they think would be easiest to beat, especially if there's an upset.
There are scheduling challenges here, but they're minor. In round two you just re-seed by end-of-season placement and go on from there.
With these changes, we now have all these goals to play for:
Best record gets home court throughout
1-seed gets first opponent choice
2 and 3-seeds get opponent choice
4-seed avoids being in the pool that the higher seeds can pick
5, 6, 7 in the playoffs and either chasing top-4 or fending off 8
8-seed get the play-in advantage
9-seed gets a chance via the play-in
Lottery by Elimination Wins
I'm not sure who originally came up with this idea but it's been kicking around for years. The idea is that lottery combinations are apportioned based on how many wins a team achieves after hitting a certain badness threshold.
The plan usually calls for "wins after a team has been eliminated from playoff contention" but that is always a bit complicated to track and becomes much harder with the play-in. Instead, we'll simplify to these rules for assigning the 1,000 available lottery combinations:
"Lottery eligible teams" consists of all teams that missed the playoffs plus any teams that qualified for the playoffs via the play-in
Every lottery eligible team is assigned five lottery combinations
The remaining 910-920 combinations are apportioned based on how many games a team wins after suffering their 38th loss (half our 76-game schedule), as a percent of league-wide elimination wins
No team can accumulate more than 200 combinations
In short, a team wins lottery combinations for themselves after it has become obvious that they aren't very good and are going to need draft help. You can't tank your way into a good lottery slot by a season of losing; it's ultimately wins that get you there.
A worse team will have more opportunity to gain lottery combinations by way of getting to 38 losses sooner than a team that was chasing the playoffs and just missed out. If you're truly awful and literally can't win any games for the whole season, too bad!
There are other lottery reform proposals that disincentivize losing but I like this solution because it incentivizes winning. It's not perfect, and there are obvious points where there may still be a little tactical tanking (a bad team sitting at 13-37 probably rests their best players for a game, for example), and a particularly well times injury and recovery can warp things, but it's much harder to game than the current system.
The actual biggest problem here is that a team that has traded away their pick has no reason to try to win and might even want to lose to make the outcome look less bad. There are more complicated solutions that might solve that but I think they all create other more significant problems. This at least isn't much of a change from the current situation.
This system doesn't exist in any other sports and so would be the hardest to get approved. However, I think there's obvious reasons that it could grow revenues for bad teams and that would help win some markets over.
With this system, we now adds a reason for franchises all the way to the bottom of the standings to be trying to win games through to the end of the season. Most teams are not going to throw away their beginning of the season because they'll believe they have some chance to make the playoffs. Envisioning a team that strategically tanks by being awful up to 38 losses but who can also pivot to winning games in the second half of the season is difficult, too.
These changes come together to give every team something to play for through almost all of the season. The only "dead spots" are teams that get locked into a seed near the end of the season, possibly the 5-7 range, the 8-seed if there's a big gap to 9th, and bad teams when they get close to 38 loses. Everyone else is playing to win all the time, so we've created the benefits of a shortened season without cutting many games.