With a minute left at UFC 278 in Salt Lake City, Kamaru Usman was about to keep his welterweight title. He had won 15 straight in the UFC, he was ahead on the scorecards, and he had already beaten Leon Edwards once, back in 2015. Then Edwards landed a left head kick and it was over, 4:04 into the fifth round.
We come back to this fight more than any other when we explain how the Index works, because the numbers and the result point in opposite directions.
On the usual fight stats, Usman won comfortably. He landed 83 significant strikes to Edwards' 55, took him down five times to one, and spent 636 seconds in control to Edwards' 176. That is more than ten minutes on top in a fight that lasted 24. A rating built only from those numbers would have given Usman credit for a fight he lost.
We score it 88–12 to Edwards.
The reason is the order things are weighed in. Every fight starts with how it ended. A knockout or submission is worth the most, then a unanimous decision, then a majority, then a split. The stats (strikes, control, knockdowns, how early the finish came) move the score around inside that. They can make a decision win look wide or narrow, but they can't turn a knockout into a loss, and they can't drag a finish below 85.
So Edwards gets full credit for the finish, a little extra for the knockdown, and nothing for the 24 minutes before it, because he lost most of them.
Is that right? We think so, but it is a choice, and this is the fight that tests it hardest. The alternative is a system that partly rewards Usman for being ahead when the fight was stopped. Judges don't do that, and nobody who watched it would either. A finish settles the question of who won, and a rating that kept arguing with it would be measuring something else.
What the stats do affect is how much the loss cost him. Usman went into that night at the highest rating he ever had. The knockout took 53 points off it, a real drop but not a collapse, because the system had seven years of evidence that he was very good and one fight that said otherwise. The rematch at UFC 286 in London the following March, which Edwards won on a majority decision, took another 33.
Four years on, Usman is still on the welterweight board at No. 6, though he hasn't fought there since June 2025 and nobody else in the top ten carries as much uncertainty in their rating. Edwards is No. 9.
The part we still turn over is the word "dominant". Most people would say Usman dominated that fight for 24 minutes. By our definition the only dominant performance that night was Edwards', and it took one kick. Both statements are true. The site only measures one of them.
The Dominance Index is built and run by Hidden Labs. Spot a number you think is wrong? Tell us and we’ll check it.