How Do Fringe NBA Teams Make the Playoffs?

It’s a maxim in the NBA that your typical 8 seed gets into the playoffs by taking care of business against bad teams and then winning just enough against other teams in their neighborhood to get a little separation when the wins are counted after 82 games.

And sure, because it’s a long season, there will be some surprising wins in there, beating Houston or Golden State on the road, and there will be some cringeworthy losses, dropping a game to a tanking team at home, but for the most part, you get to that 42-win range (East) or 48-win range (West) by beating teams worse than you, losing to teams better than you, and splitting the difference with teams on your level.

The “Ideal Scenario”

Let’s take, in a perfect vacuum, a hypothetical fringe playoff team in the Western Conference in a typical year.

They get 30 games against the East, 2 each against all 15 teams.

You can expect them to win, say, 20 of these—the 8th-best team in the West is often good enough to be the 5 seed in the East, so they beat the 6 through 15 teams and lose to the actual 1 through 5 teams. (We’re leaving home/road splits out of this just to keep it simple, but in aggregate it ultimately balances out as they beat “better teams” at home and lose to “worse teams” on the road.)

They’re at the exact median in the West, so they win 26 games against 7 better teams and lose 26 teams against 7 worse ones. Again, perfect vacuum.

So your ideal Western Conference team makes the playoffs by going 46-36 in our hypothetical example, and indeed, last year the cut line was 47-35 (Denver, at 46-36, fell just short.)

Meanwhile, the East is weirder because it’s such a dumpster fire from 10 on down most years, the six worst East teams as bad or worse than just the 14 and 15 seed in the West and rarely anyone else.

So our 8 seed East team goes 26-26 against their own conference, would probably be ninth or tenth in the West, and as such goes (let’s split the difference on that 9/10 conversation) three games under .500 against the West, finishes 39-43 overall, and leads to a bunch of think pieces about the NBA getting rid of conference-based seeding as that team like the ’18 Nuggets (or even the Clippers, who went 42-40 in 2017-18) gets seen as being wrongfully kept out of the playoffs.

That’s our baseline. That’s our absolutely perfect-world scenario where teams never suffer upsets and where injuries and home/road weirdness and season-long streaks never conspire to throw monkey wrenches into the data. It’s basically a world governed by something like FiveThirtyEight’s ELO rankings or Vegas point spreads.

The Real World

Let’s build a little sample out of the 2017 season to run some numbers with what will amount to a 492-game overall sample. We’ll take the 7, 8, and 9 seeds in the East and West that year and see how they did against teams they “should beat” and “should beat them.”

We’ll sort these by Better, Same, and Worse and see what we come up with:

Pacers: Better 13-18 (.419), Same 1-3 (.250), Worse 28-19 (.596)
Bulls: Better 19-22 (.463), Same 3-2 (.600), Worse 19-17 (.528)
Heat: Better 20-21 (.488), Same 1-4 (.200), Worse 20-16 (.556)
Grizzlies: Better 14-18 (.438), Same 1-1 (.500), Worse 28-20 (.583)
Blazers: Better 14-25 (,359), Same 3-1 (.750), Worse 24-15 (.615)
Nuggets: Better: 17-30 (.362), Same 0-0 (-), Worse 23-12 (.657)

2017 Totals:
Better 97-134 (.420)
Same 9-11 (.450)
Worse 142-99 (.589)

Right here, we can see two interesting points. One, this data is incredibly noisy. Two, just because you beat up on worse teams doesn’t mean you get carte blanche to lose out against the best teams—part of the reason the Nuggets missed the 2017 playoffs was because they couldn’t buy a win against any of the top-tier squads.

But a third point—teams tend to do better against teams worse than them—at least holds across a bigger sample. Fringe playoff teams (the six teams in our sample were a combined 248-244) lose to top-tier teams and beat lottery teams.

Oddly, the team that did the best against good teams—Miami, and if you’ll recall, 2016-17 was that weird season they started 11-30 and finished 30-11 in the first and second half of the season—didn’t make the playoffs. And three teams with pretty big splits—the Pacers, Grizzlies, and Blazers—all did.

All of this is fairly obvious, of course, And while it’s just one season—and one where there was relative parity between East and West fringe playoff teams, where both conferences’ 7-9 seeds went a combined 84-82—the broad pattern, where teams in the middle of the pack have about a 20 percent deviation off their overall won/lost record against better/worse teams, suggests a conclusion where while there’s a huge spread between the best team and the worst team that establishes itself with a 50-game spread or more in 82 games, the NBA’s middle class has enough parity that tuning into a game between teams close in record, especially if the slightly worse one is at home, will all but guarantee, on average, plenty of competitive basketball.

Conclusion

But it also tells fans something.

Namely, if you root for a fringe playoff team, no one game is likely to be the end of the world. If your team is truly a playoff squad, they will over the course of a season get some big wins and take some bad losses. You can’t lose too many against the good teams or you’ll be left like the Nuggets were for two straight years, not good enough against good teams to make up for being great against bad teams.

Nor can you give away too many “easy ones”—the 2017 Bulls didn’t clinch until the last day of the season even though they went .488 against teams better than they were, because they took too many losses in winnable games.

But on the whole, it’s probably best to take it one day at a time.