I believe there have only been five truly ethical NBA champions in the last 25 years: teams that don’t need an asterisk by their title.

This is a chart of every NBA champion since 2001, ranked by how ethical their title was. (Don’t worry about the numbers yet, we’ll get there).

Standing on their own are the 2015 Warriors, by far the least ethical champion of the century by my measure.

They got lucky, facing multiple opponents missing important players. But not all roads to the championship are easy…

To make the case, let’s start back in 2001. A simpler time, when AI just meant Haley Joel Osment playing a robot learning to be a real boy.

Back then, players had grit and barely missed games. Lately, though, the playoffs have become a war of attrition. Availability, as they say, is the best ability. But a title decided by who’s healthy is a title decided in some part by luck.

Here is how many games above-average players (defined shortly) missed each playoff year.

But let’s be real: a missing 40 year-old Karl Malone on his last legs is not the same as a prime Kevin Durant.

Plenty of all-in-one stats have been built to quantify a player’s impact. We’ll use DARKO. Without getting too in the weeds, it estimates how much a player helps their team win, measured in points added per 100 possessions versus a league-average player.

Most importantly, DARKO is a forward-looking metric that blends things like box-scores, play-by-play, and age to answer the question “how good is this player right now?” Or in our case, “how much impact is this team missing?”

Roughly speaking, +6 is MVP tier, +3 is a star, and +1 is a solid starter. 0 is a league-average player, replaceable.

With DARKO, we can quantify the impact that is missing from each playoffs by year, using the end-of-regular season DARKO scores (DPM). Now we can go from simply counting missed games to measuring the missing impact in each one.

I’ve chosen to limit to players of at least +1 DPM, which on average are the top ~80 players each year. Key starter-level players. Here is the average DPM missed per game each playoff year.

We need one more ingredient to make things feel right: the Haliburton effect, if you will.

According to the data he played in game seven, so he doesn’t count as a DNP. But he got injured just seven minutes into the first quarter, and the void he left for the rest of the game was undeniable. So we’ll award some partial missing DPM values for these unusual circumstances.*

Okay, it didn’t add a ton, but I can sleep at night knowing unusually low-minute situations are accounted for.

Now let’s focus on NBA champions’ opponents. Charts are fine for general trends, but a Game-of-Thrones-esque pile of heads might help put the volume—and the humanity—in context. Each head is a game missed by the 2015 Warriors’ opponents, approximately sized by their missing DPM value.

Jrue Holiday, Mike Conley, Tony Allen, Anderson Varejao, Kevin Love, and Kyrie Irving all missed games in three of the four rounds. Our champs got to walk through the playoffs avoiding these players, making the path undeniably easier. I’m not saying it was easy, but don’t tell me they didn’t get a little lucky not facing a full-strength Cavaliers roster.

For fun, here is every game missed by the opponents of each year’s champ, as a pile of heads.

If we reduce these piles into a single score, we can plot them cleanly and compare them.

This shows us the numerical advantage each champ got from missing opponents. But that’s only half of the equation. Because the champs might have had their own missing players. That makes things harder, and can balance things out.

People harp on the 2025 (and beyond) Thunder for their questionable bucket-getting approaches. They might not get ethical buckets, but they did get an ethical championship.

In 2025 the Thunder caught a couple breaks by dodging Ja Morant and Brandon Clarke on the Grizzlies, John Minott on the Timberwolves, and the infamous game 7 from Tyrese Haliburton.

But the Thunder had their own setbacks, missing Kenrich Williams for nearly the whole run, Aaron Wiggins for a game, and Isaiah Joe for some pivotal Finals games. While not superstar level players, they were still important pieces.

Subtract what they were missing from what their opponents were missing, and it yields the most analytically ethical championship in our data.

When we apply this to the remaining teams, we see the full, net-impact-of-missing-players picture.

While the 2025 Thunder come out just ahead numerically, the volume of missed games (on both sides) is doing a lot of work. (Insert bad joke about the Thunder taking advantage of the rules). I’ll be the first to admit that no data-driven argument like this is without its imperfections. A clean score doesn’t tell you whether the breaks were decisive. A dominant enough champion wins regardless of who’s missing. This measures the luck, not whether the luck mattered.

Of the five that are quantitatively ethical, two others jump out that appear more worthy of the ethical crown.

Hot-off-the-press champions, the 2026 Knicks, had a historically ethical run. They almost had a near-perfect healthy opposition, missing a single game from Joel Embiid. On the flip side, they played two games without OG Anunoby, and one without Mitchell Robinson. Second place.

But for me, the purest of the ethical champions has to be one of the teams that faced unblemished rosters, leaving no doubt that they outperformed full-strength competition.

The analytical winner is the 2006 Miami Heat. The Heat played a couple games in the first round without Alonzo Mourning. There was no single full DNP game from opponents, and just a single low-minute game from Tyson Chandler on the Bulls in the first round. The smallest of asterisks.

Appendix

Data and Methods

Data from darko.app and basketball-reference.com.

Since DPM is updated daily, I used a snapshot of their last regular-season game. This is DARKO’s best guess of how good they were heading into the playoffs.

To qualify for “low minutes” and partial DPM scores, players had to average at least 20 minutes per game during the regular season, Low minutes was defined as less than a third of their regular season minutes per game. Their amount of DPM was multiplied by the proportion of minutes they missed to arrive at a missing DPM value. For example, if a player missed 75% of their normal minutes and had a DPM of 2.0, then their missing DPM would be 1.5, not 2.0.

I have previously attempted to quantify the asterisks associated with NBA championships. Tom Haberstroh recently published a similar thought experiment.