Last summer, we played with a little metric we called Core Score. The idea was this:
- A large portion of an NBA team’s wins come from the value produced by a small number of players: The team’s “core.”
- In the postseason, that core is even more important than normal. They play a greater share of minutes. They still play the highest-leverage possessions, but those possessions are higher-leverage than they are from October through March.
- By putting a number on the value of each team’s core, something we call Core Score, we can identify which teams have a core good enough to seriously contend for a title.
- We can also use Core Score to indirectly measure the strength of all the parts of a team that aren’t its core: The supporting cast. The coach. The culture. The core is the cornerstone. The core is essential. But there is a lot of value outside of the core as well, and by measuring the core, we get a number for these parts.
- We call this Core Score counterpart “Everything Else,” and it’s the difference between 1) a team’s regular season wins and 2) the combination of Core Score and baseline wins. This means it’s built to average to zero. This is extremely inexact. Its most obvious flaw is that it doesn’t account for injuries at all. But it’s interesting in its own right.
To originally measure Core Score, we used last year’s year-end WAR from RAPTOR, a FiveThirtyEight system (RIP), and we used The Ringer’s NBA player rankings. A few updates as we measure it today:
- RAPTOR is not currently being updated. It sounds like it might come back eventually, with Nate Silver retaining the rights to it, but Disney gutted FiveThirtyEight, and RAPTOR was one of the many casualties.
- Going off of this: We’re sticking with the WAR curve we built off of last year’s RAPTOR. It’s a piecewise function that’s largely linear, but we think this is just as good as anything else we’d build in a reasonable amount of time. Its inflection points come where natural divides exist in the roles of NBA players—around the top ten, around the top 25, around the top 90 (where the best starting fives end), etc. It makes some intuitive sense.
- The Ringer reduced their rankings from top-125 to top-100. The only space this impacts us is that last year, we had 21 wins as our rough Core Score threshold for title contention. Losing these last 25 players makes 20.5 the new threshold if we handle this perfectly consistently. Instead of doing that, we’ll call the threshold 20, because it’s already inexact and 20 is a nice round number.
- Summing this up: Teams need to have a core worth about 20 wins to be serious title contenders. With an average Everything Else, a team with a 20-win core would be expected to win 46 or 47 games in the regular season. However. Teams with good Core Scores tend to have a good Everything Else score as well. It’s not a very strong correlation—the r-squared between the two is only about 0.3—but it’s a clear positive correlation. We’ll get into potential reasons behind this below.
Core Scores, Everything Else
The scores, team by team. Again, this is roughly how much WAR is coming from each team’s top-100 players when everyone is healthy:
| Team | Core Score |
| Nuggets | 31.7 |
| Celtics | 27.8 |
| Timberwolves | 23.8 |
| Clippers | 22.5 |
| Bucks | 22.4 |
| Thunder | 19.6 |
| Suns | 18.9 |
| Pelicans | 18.9 |
| Lakers | 18.6 |
| Mavericks | 17.2 |
| 76ers | 17.0 |
| Heat | 16.8 |
| Cavaliers | 15.7 |
| Knicks | 15.0 |
| Grizzlies | 14.5 |
| Warriors | 13.5 |
| Kings | 13.2 |
| Pacers | 13.2 |
| Rockets | 11.9 |
| Magic | 10.2 |
| Hawks | 10.1 |
| Bulls | 9.6 |
| Spurs | 9.0 |
| Raptors | 7.8 |
| Jazz | 7.3 |
| Blazers | 5.7 |
| Nets | 4.1 |
| Hornets | 3.8 |
| Pistons | 2.6 |
| Wizards | 1.6 |
Here, on the other hand, is Everything Else. As an example of how we calculate this: The Timberwolves have a Core Score of 23.8 wins. 26.9 was the baseline wins number this year, the number of games a team with a Core Score of 0.0 would be expected to win with an average Everything Else. The Timberwolves’ expected wins, then, are 50.7, but since they won 56 games, their Everything Else score is 5.3. Last year, we incorporated a conference adjustment in here, but this year, setting that number to zero gave Core Score the greatest predictive accuracy, so we ignored it.
| Team | Everything Else |
| Thunder | 10.5 |
| Magic | 10.0 |
| Celtics | 9.3 |
| Knicks | 8.2 |
| Pacers | 7.0 |
| Kings | 5.9 |
| Mavericks | 5.9 |
| Warriors | 5.7 |
| Cavaliers | 5.5 |
| Timberwolves | 5.3 |
| Pelicans | 3.2 |
| Suns | 3.2 |
| 76ers | 3.1 |
| Bulls | 2.5 |
| Heat | 2.3 |
| Rockets | 2.2 |
| Clippers | 1.6 |
| Lakers | 1.5 |
| Nets | 1.0 |
| Bucks | -0.2 |
| Hawks | -0.9 |
| Nuggets | -1.5 |
| Jazz | -3.1 |
| Hornets | -9.7 |
| Raptors | -9.7 |
| Blazers | -11.5 |
| Wizards | -13.5 |
| Spurs | -13.9 |
| Grizzlies | -14.4 |
| Pistons | -15.5 |
A few things that jump out here, and some thoughts:
- Two young teams find themselves at the top. It’s possible The Ringer is undervaluing guys like Franz Wagner, who had a tremendous year. It’s also possible our WAR curve is undervaluing Shai Gilgeous-Alexander, who likewise had a tremendous year.
- Good teams are almost all in the top half. Bad teams are near the bottom. The seven teams with the worst Everything Else scores were the seven worst in the NBA. This sort of thing will always be the case—it’s a function of how we calculate Everything Else—and some of it comes from injuries, but I think the real thing it does illustrate is that in the NBA, teams with bad cores have a strong incentive to put together the worst supporting cast possible, while good role players have a strong incentive to join teams with good cores.
- What we called this last year was “over/underperformance,” and while we’ve realized upon reflection that this isn’t what it is, there’s still an element of that present. Betting markets demonstrate the prevalent thought that the Thunder aren’t ready to contend for a title despite having the best record in the Western Conference in the regular season. The logical corollary there is that the Thunder overperformed from October until now.
Contenders, Non-Contenders
Looking at title contenders, then, and a few notable teams who’ve already been eliminated:
| Team | Core Score | Everything Else |
| Nuggets | 31.7 | -1.5 |
| Celtics | 27.8 | 9.3 |
| Timberwolves | 23.8 | 5.3 |
| Clippers | 22.5 | 1.6 |
| Bucks | 22.4 | -0.2 |
| Thunder | 19.6 | 10.5 |
| Suns | 18.9 | 3.2 |
| Pelicans | 18.9 | 3.2 |
| Lakers | 18.6 | 1.5 |
| Mavericks | 17.2 | 5.9 |
| 76ers | 17.0 | 3.1 |
| Cavaliers | 15.7 | 5.5 |
| Knicks | 15.0 | 8.2 |
| Warriors | 13.5 | 5.7 |
| Pacers | 13.2 | 7.0 |
| Magic | 10.2 | 10.0 |
We’ll go through these in groups.
The Favorites: Nuggets, Celtics
Between the Nuggets and the Celtics, I think it’s fair to call Core Score meaningless, but in a way that’s bad news for the Celtics. Nikola Jokić is so insanely valuable that it’s hard to measure just how valuable he is. Our curve says he’s about 50% more valuable than the next-best player, but RAPTOR’s WAR last year had him twice as good. Core Score, like so many of these systems, breaks down at its extreme end, which in this case is the top. Assigning value to Jokić is hard. We’ll come back to this concept when we get to the Mavericks, below.
The Nuggets are also unusual in that they have the worst Everything Else score of any playoff team. What is this? Some of it might be that they had an understandable and harmless championship hangover during the regular season. They could afford to take it easier, and they did. Some of it, though, is their depth problem. They really miss Bruce Brown.
You could spin the Core Score and Everything Else gaps as the Celtics really drilling how to win. You could also spin it as the Nuggets not playing their best ball yet this year.
Contenders (if Healthy): Timberwolves, Clippers, Bucks
Our other contenders, using that 20-win Core Score threshold, are the Timberwolves, Clippers, and Bucks. The Clippers and Bucks are fairly simple: Kawhi Leonard and Giannis Antetokounmpo are two of the best basketball players alive, and they each have some serious firepower next to them. The issue for those two is health.
The Timberwolves are more interesting, because even with The Ringer still having Anthony Edwards as only the 14th-best player in the league (maybe he’s better, maybe he’s worse, but he’s getting attention like a top-five guy), their total rankings are very high on Minnesota. The T-Wolves have six players in the top 100. They’re the only team in the league with that many. On top of that, they have a solid Everything Else score. Their core is deep, and so is their depth.
On the Edge: Thunder
The Thunder are so, so close to the 20-win threshold. If you flip Giannis and SGA in The Ringer’s rankings (they’re third and fourth), the Thunder cross the plane. Either way, they’re right on the edge. SGA has to be an especially good top-3 player to put them comfortably in the contender pack, or Jalen Williams has to play like a great second option instead of a solid one, or Lu Dort and Josh Giddey have to play like serious top-100 players. All those names being so young is very fun for Oklahoma City, but in terms of this year, the Thunder need more from their core to seriously challenge for a championship.
Notably Low: Mavericks, 76ers, Knicks
We mentioned the gap between Jokić and the second-best player, and that’s relevant to the Mavs because the second-best player is currently, on The Ringer’s rankings, Luka Dončić. Our WAR curve doesn’t have a huge gap between 2nd and 3rd, or 2nd and 4th, or even 2nd and 36th. The gap between Nikola Jokić and Luka Dončić, in Core Score, is a little bit bigger than the gap between Luka Dončić and Scottie Barnes. Is Core Score underestimating Dončić? It could be. He’s young. The curve was built last year, when he was younger and not ranked second by The Ringer. He finished cleanly in second place for the MVP in betting markets, even if we don’t yet know the outcome of the real votes. If he’s separating himself from the pack, then the Mavericks have a better chance than Core Score implies.
Then, there’s Kyrie Irving. The Ringer has him 30th, behind guys like Brandon Ingram, Lauri Markkanen, and Domantas Sabonis. Their writeup of Irving begins, “Set aside all the bullshit for a minute…”
What’s happening with Irving is that The Ringer has him devalued because of his off-court volatility, and how it impacts his on-court availability. This is fair, but he’s on the court right now, and during the playoffs, that’s unlikely to change. He’s capable of setting aside all that bullshit for two months.
Still, those are the only two top-100 players for the Mavs. They have one of the shallowest cores in the league. It immediately gets into Everything Else. And while their Everything Else is solid, it’s asking a lot of Daniel Gafford to label this team a serious title contender.
The Knicks are eye-popping in part because their Core Score includes Julius Randle, who’s out right now, and they’re still very low. Take away Randle, and they’re even lower. I wouldn’t personally be surprised if their score improves when The Ringer next updates its rankings (the last update was April 10th), but it’s fair to observe that Donte DiVincenzo and Josh Hart had shockingly good regular seasons. Again, there may have been some overperformance there.
As for the Sixers: Tyrese Maxey’s stock is rising, and Joel Embiid is still very good, but it’s a thinner front line than it was a year ago.
Only Technically Alive: Cavaliers, Pacers, Magic
Not a lot to note here, but since they’re still playing, we thought they should be included.
Weren’t Good Enough: Suns, Lakers, Warriors
These are the interesting ones. I think the Suns are maybe the most interesting of the three. They built this Big Three. They have the best second option in the league. The Ringer gives Grayson Allen a little love, and it has Bradley Beal’s stock down. The Ringer does not seem low on the Sun. The Suns just don’t have a good enough roster. They did all they’ve done, and they’re still not good enough on paper to be a title contender. They flopped, but they weren’t winning the title anyway. They need more firepower.
Similarly, the Lakers have a great second-best player and then it’s down to D’Angelo Russell and Austin Reaves. Darvin Ham might have been *a* problem this year, but he was not *the* problem.
And finally, the Warriors. Again, these rankings account for the off-court stuff, something that has a bigger impact on Ja Morant than Draymond Green. And these rankings have the Grizzlies’ core better than that of Golden State.
Surprisingly High: Pelicans
The other really interesting one is the Pelicans. Some of this is The Ringer really buying into Zion’s last couple months of the year (they have him 20th after having him 31st in February), and some of this is Brandon Ingram looking a little better before this postseason than after it, but the Pelicans have a good core! Health is the huge concern, of course, but even if you get aggressive and assume that one of Williamson and Ingram will always be hurt, this roster could be the third-best in the league if its front office could find a way to add one superstar. That is far easier said than done, but I’m pretty sure they have seven first-round picks over the next four years. One option is to bet on their drafting ability with those and plan on acquiring a superstar the old-fashioned way. Another is to do something crazy, like try to get Devin Booker.
