Is It Getting Harder for NBA Superteams to Win?

Moving beyond anecdote, some recent evidence suggests yes.

James Harden, Kawhi Leonard and Paul George of the LA Clippers sit on the bench during the fourth quarter of the game against the Portland Trail Blazers at the Moda Center on March 22, 2024. (Alika Jenner/Getty Images)

In my recent story about the Philadelphia 76ers signing Paul George — giving them a championship amount of star power, on paper — I speculated that the only thing holding Philly back might be a break in the long-standing historical relationship between said star power and rings, now that the NBA appears to be in an era of uncharacteristic parity.

But that’s all it was: Speculation. I noted some anecdotes around big-name cores that had failed recently, but I didn’t offer much more in the way of evidence that star-led teams were prevailing less in the playoffs. So today, let’s change that — let’s see if star power really is a less potent formula for winning titles than we’re used to it being.

To study this, I will turn back to my friend and former boss

Nate Silver

’s framework for judging the star power of an NBA team, based on how its best players compare with the top stars on a typical championship roster. Specifically, we can hand out “Star Points” to players depending on whether their full-season¹ Estimated RAPTOR ratings overlapped the usual ranges of the best player (3 points), second-best player (2 points) or third-best player (1 point) on the average champ since 1985:²

In this Estimated RAPTOR-fueled update of the research, the primacy of stars over the past four decades of NBA history remains clear, as it pertains to winning titles. If we add up the total Star Points on each roster in that time period, teams with 8 or more points win the championship 36 percent of the time, while teams with 7 points win 24 percent of the time. Then, after that, things take a pretty steep drop-off:

Based on their ratings last season,³ the new-look Sixers would have 6 Star Points between Joel Embiid, George and Tyrese Maxey. That’s better than most of the groupings on the list above, even if it doesn’t quite qualify for “superteam” status by this method. (Star Points judge more harshly than my original “look at whether a team has guys who fit the pattern of Nos. 1-3 on the typical champ” concept.)

But the main point here isn’t to quibble with how super Philly’s new team is. It’s to study whether superteams truly are losing their effectiveness in more recent seasons. And to that end, I settled on looking at teams with at least 7 Star Points at five-year intervals since 2000. (I chose 7 as the cutoff because there just aren’t many teams with more than that in any given decade to use as a sample.) Over the long run, 7+ Star Point teams win the title about 30 percent of the time, including a 28 percent rate before 2000 — but that number has varied a lot since, depending on the half-decade:

In the late 2000s and early 2010s, superteams such as the 2005, ‘07 and ‘14 Spurs, the 2012 and ‘13 Heat and the 2008 Celtics all won championships with 7+ Star Points. (And the 2009 Lakers weren’t far off, either, with 6 Star Points.) Overall, 40 percent of the teams with 7 or more Star Points won the championship during the decade from 2005-2014.

However, that number has fallen in each of the two half-decades that followed — to 27 percent from 2015-2019 and now 22 percent from 2020-2024.

While there have still been plenty of success stories, too, from the historically dominant 2024 Celtics to the comically stacked 2017 Warriors and their 11 (!) Star Points, even more superteams have fallen flat in recent seasons. Just since 2020, we’ve seen seven total failed 7+ Star Point seasons from the Celtics, Bucks, Mavericks, Nuggets, Suns and Nets, with George’s Clippers also failing multiple times at 6 Star Points. In many ways, the batting average for superteams has never been lower than it is over the past five years.

Now, that’s still just a sample of nine teams. (It’s 20 if we also include the preceding half-decade, of which five won titles, or 25 percent.) NBA history is so dominated by star-laden teams that it would silly to cast aside that entire paradigm and draw sweeping conclusions off of the past 5-10 years alone.

But, that said, there are structural reasons to think that this trend isn’t merely small-sample noise. With injuries playing as large a role in the playoffs as ever, it might be exposing a flaw in the model that relies on a few core stars to be healthy. The NBA has also been making a concerted effort to discourage this approach to team-building — how many times do we hear the phrase “second apron” nowadays?? — or, even if you do manage to assemble your stars, make it difficult to surround them with any kind of supporting cast.

And, at the end of the day, only one team can be champion. Last season’s Mavs, Nuggets and Clippers had big star power, too, but the Celtics were too good to be denied. More teams qualified with 7+ Star Points in each of the past two half-decades than in any of the half-decades before them (since 2000); if a greater share of teams are trying the strategy but only one can still win, then of course its success rate will drop.

But that’s also the point. Stars will never not be important for winning in the NBA, but the market had become so saturated with would-be superteams over the past decade that something had to give. We’ll just have to see whether that trend continues to manifest itself over the next half-decade and beyond.

---

Filed under: NBA

1. So, including stats from all teams a player played for when judging his star power for any particular team.

2. The particular cutoffs for each are as follows: 3-star players have a RAPTOR of at least +5.4, 2-star players are between +3.7 and +5.4, and 1-star players are between +1.9 and +3.7. (These cutoffs are slightly — though not significantly — different from the numbers in Nate’s original research, as that story predated RAPTOR by years and included seven fewer champions that had not yet won their titles, which really just makes me feel old.)

3. If we fudge things to include Embiid, who played only 1,309 minutes last year.