UFC rankings lag the market, and Saturday proved it twice

Twice on Saturday night, the No. 12 fighter in a UFC division beat the No. 8 fighter in that division. Twice, the No. 12 fighter walked in as the betting favorite. Nobody who watched the odds move was surprised by either result, and that should bother the promotion more than it apparently does.

Quillan Salkilld was ranked 12th at lightweight and closed at 60 percent to beat No. 8 Mateusz Gamrot. He submitted him in the first round. Alexia Thainara was ranked 12th at women's strawweight and closed at 73 percent against No. 8 Amanda Lemos. She won every round and held control for 7:43 of a fifteen minute fight.

Two divisions, same gap, same direction. The board had the fight right and the ranking had it backwards, by four places, in both cases.

Photo: RapidSpin33 (talk) (Zach Able) / CC BY-SA 3.0

I want to be fair about what a ranking is for. The UFC's list is voted by a media panel, not calculated, and it was never designed to be a prediction. It is a record of who has done what, and by that standard Lemos at No. 8 was defensible on résumé alone. She has beaten Mackenzie Dern, who is the champion now.

The problem is that résumé decays and the list does not price the decay. Lemos had lost three straight fights before Saturday, to Thainara, to Gillian Robertson and to Tatiana Suarez, and she was still ranked four spots above the woman who most recently beat her. She is 39. The market knew all of that and moved. The panel had her eighth.

Salkilld is the sharper case, because the ranking was not slow, it was inert. Before Saturday he had won six straight in the UFC, five of them by first round finish, including a knockout of Beneil Dariush in May. That is the single best active run at lightweight and it bought him 12th, one spot below Rafael Fiziev and three below Dan Hooker. The books had him favored over the No. 8 man. The panel had him behind four fighters he would have been favored against.

Some will say this is exactly what rankings are supposed to do: reward accumulated wins over hot streaks, so that one good year does not vault a prospect over a decade of work. I agree with that in principle, and I would defend a system that lagged by one or two places on those grounds. Four places is not lag. Four places is a list that has stopped tracking the thing it claims to track, in a sport where the promotion uses that list to sell fights and to justify who does and does not get a title shot.

And the cost of the lag is not paid by the panel.

The third fight on that card was Darren Elkins, 42 years old, 31 UFC appearances, getting knocked out in 35 seconds by Yadier del Valle. Elkins landed nothing. He attempted two significant strikes, took a knockdown, and it was over. The market had him at 13 percent, the widest line on the card, and the market was correct within one round.

Nobody needs a ranking to tell them that matchup was one way. What the ranking does control is who gets fed to whom on the way down, and the sport's habit of booking a beloved veteran against a 30 year old finisher with an 11-1 record is not a ranking problem, it is a matchmaking one. But the two are the same machine. We published research two days before this card finding that age combined with accumulated knockdowns, not total fights or total minutes, is what predicts the fall. Elkins fit that profile before the walkout, and it took 35 seconds to confirm it.

Here is the fix, and it is not complicated. Publish the panel's ballots. The UFC already publishes the board; it does not publish who voted for what or when a voter last updated. Ballot transparency is standard in every other sport that runs a media poll, it costs nothing, and it would let anyone see whether a fighter sits at No. 8 because voters believe it or because nobody bothered to move him.

Failing that, at least stop pretending the number means what the promotion says it means when it sells you the fight. On Saturday the market beat the ranking twice, by four places, in two divisions, in the same night.

The board is not smarter than the panel. It is just paying attention.