The NBA can save its ratings by splitting the season in two. Here’s how

Everyone considers the NBA to be in crisis even though revenue and culture importance has grown every year since the Great Recession. It would appear the NBA agrees with the view to some extent, given how commissioner Adam Silver is embracing the concept of certain reforms. Among the perceived problems everyone is trying to fix: the regular season has become superfluous and feels really long, load management is annoying fans and hurting ratings, conference imbalance has long been unfair to the best West and worst East teams. (I would argue parity has driven lower top-line national ratings, but the league is obsessed with competitive balance, so it’s really not worth arguing in the margins because no one will admit having a few star-laden powerhouses is better than having 10 very good teams.)

The league is planning to introduce an in-season cup-style tournament to spice things up and replace some regular-season games. This will marginally decrease the overall calendar for most teams. This is not enough: stars will still rest, bad teams will still spend four months aimlessly losing games, good teams will still spend months optimizing their teams for long playoff runs (not gunning for top seeds). If you really think the NBA has a big problem and you really want to fix some things related to the schedule and the regular season ennui, you’ve got to think bigger.

What about cutting the regular season in half?

Not as in going from the current 82 games to 41 games in an entire season: no one would go for that because of the massive gate, concessions and broadcast revenue cut it would require. But literally cutting the regular season into two splits: a winter split and a spring split.

This happens in esports and some other smaller professional leagues. You could marginally decrease the overall number of games to something like what the NBA is currently planning: say, 76 total games per team with 38 in each split. The 38 games could be split up like this: 15 games against teams in the other conference (one each), eight games against your division rivals (home-and-homes), 10 games against one other division in your conference (home-and-homes), five games against the other division in your conference (one each). In the second split, the in-conference division you’d play twice would flip, and you’d get home games for the teams you only played on the road in the first split (and vice versa).

Standings are segregated within the splits and by conference. If you finish in the top three in your conference in either split, you are guaranteed a spot in the NBA playoffs. If all of the top three teams in each conference in each split are different, that’s 12 teams. Odds are this will not be the case: you’ll have some decent overlap among the top teams in each split. (For instance, you’d expect a team like the Bucks to be in the top three in each split.) This will open up additional playoff slots that will go to the top remaining teams. The winner of the midseason cup tournament gets an automatic bid into the playoffs. If this team otherwise makes the playoffs by being a top-3 seed in a split, this spot goes to another remaining team.

You seed the split top-3 teams first, put the tournament cup champion in the next available seed (if they aren’t already accounted for) and then the remaining teams to get up to 20 total teams based on their best split record. For instance, if the Mavericks have a 19-19 winter split and a 28-10 spring split but didn’t finish in the top-3 in the West in either split, they are seeded after the top 3s based on that 28-10 split. The 19-19 first half of the season essentially doesn’t count. Why do this? So that teams running away with a top-3 finish in the first split don’t take their foot off of the gas pedal given that teams in the second split could do better and get a higher seed, and so teams with one awful split (like the Warriors) still has reason to compete hard in the second split. If you average or combine the split records, a team like Golden State would just tank out the spring split instead of trying to win and make the playoffs.

If you want to keep this all within the conference structure and just reseed the final four teams (as the NBA is currently proposing), that’s fine. It’s not a make-or-break for this system. The teams seeded Nos. 17-20 play single-elimination play-in games (similar to what the NBA is currently proposing going forward): No. 17 vs. No. 20, No. 18 vs. No. 19. If you keep conferences in the playoffs, it’s the No. 7 vs. No. 10 and No. 8 vs. No. 9. Then you run the best-of-seven series from there, reseeding in the top four if you’re keeping conferences.

Let’s offer an example to help put it all together. Here’s a scenario: in the winter split, the top three East teams are the Bucks (33-5), Heat (28-10), and Celtics (27-11). The top three West teams are the Lakers (32-6), Clippers (30-8), and Rockets (28-10). In the spring split, the top three East teams are the Bucks (30-8), Sixers (29-9), and Raptors (26-12). The top three West teams are the Clippers (31-7), Lakers (30-8), and Nuggets (26-12). The Mavericks win the in-season tournament.

We seed all these teams based on each team’s best split:

1. Bucks
2. Lakers
3. Clippers
4. Sixers
5. Heat
6. Rockets (we’ll assume the Heat win the tiebreaker with the Rockets since the NBA hates the Rockets)
7. Celtics
8. Raptors
9. Nuggets (we’ll assume the Raptors win the tiebreaker with the Nuggets)
10. Mavericks

Due to split top-3 overlap, we end up with nine split seeds and the tournament champ. That leaves 10 spots for the next 10 best teams in either split.

If you do it by conferences, you’d have the Bucks, Sixers, Heat, Celtics, and Raptors in the East in that order with five additional East teams filling out the bracket, and the Lakers, Clippers, Rockets, Nuggets, and Mavericks in the West with five additional West teams filling out the bracket.

What about the NBA draft? Seed the lottery just as you would the playoffs, but in reverse: each team’s worst split counts. This removes the incentive to tank both splits. A team might be awful in both splits, but there’s no draft advantage to 76 games of misery, so you might as well compete for a play-in slot in the spring!

All this complexity is a feature, not a bug: it keeps fans of teams super engaged in the minutia of the playoff race, with every game counting more and “lost seasons” being chopped in half. On the broader national stage, there would likely be fewer meaningless games for TNT and ESPN (although teams might cut bait on a bad split sooner — you’re still not going to get much value out of all these nationally televised Pelicans games while Zion Williamson is out). The cup tournament and play-in games should be T.V. winners, and late-split scoreboard watching should be really interesting.

There are surely some disadvantages to moving in this direction, and some tweaks would need to be made as we see how it all actually plays out.