If you have ever looked at the UFC rankings and thought that cannot be right, you were closer to the truth than you knew. For thirteen years those numbers were not a calculation at all. They were an average of ballots submitted by journalists.
That changed in June 2026, when the UFC introduced the Meta UFC Rankings — an Elo rating model built with Meta that replaced the media vote with arithmetic. It is the biggest change to how fighters are ranked since rankings were introduced in the first place.
Here is how UFC rankings actually work now, what the model measures, what it deliberately ignores — and the one thing that did not change at all.
What the Rankings Are For
Start with the thing most fans get wrong: the rankings have never been a qualification system.
In most sports, standings determine access. Win enough and you are in the playoffs. The UFC does not work that way. It is a promotion, not a league, and it books the fights it wants to sell. Rankings are a published opinion about merit, sitting alongside a matchmaking process that is free to ignore them.
That distinction explains most ranking controversies before you even reach the methodology. A fighter can be ranked fourth and never get near a belt. Another can arrive from outside the top ten and headline. Neither outcome is a malfunction — the rankings were never wired to the outcome.
What they do influence is real, if indirect: negotiating position, public perception, media narrative, and in some contracts the bonuses attached to beating a ranked opponent.
How the Media Panel Worked (2013–2026)
Rankings arrived in February 2013. The mechanism was straightforward and, for its era, reasonable.
A panel of media members — drawn from 22 publications — submitted a top 15 for every division each week, plus a pound-for-pound list. The ballots were compiled and averaged, and the average became the official rankings on UFC.com.
The rules of eligibility were narrow and remain informative:
- A fighter had to be on active UFC status to receive votes.
- Champions and interim champions were treated as occupying the top positions of their divisions and so were not voted on within their weight class — though they were eligible for pound-for-pound.
- A fighter competing in two divisions could be ranked in both.
The system's strength was that it could think. A voter watching a fighter dismantle a top contender could react immediately, and could weigh context a spreadsheet would miss — a bad weight cut, an injury mid-fight, a division in transition.
Its weakness was the same thing. Averaged human opinion cannot be audited. When a fighter dropped three places after a win, nobody could show you the arithmetic, because there wasn't any. Reputation, name recognition and momentum all counted, and there was no way to separate them from merit.
What Changed in June 2026
The Meta UFC Rankings went live on Monday 23 June 2026, after development that began in February 2025. The system is an Elo model — the same family of rating algorithms used in chess, where each competitor holds a number and beating a stronger opponent moves that number more than beating a weaker one.
The UFC has been unusually careful about what the system is not. Its own explanation states that the rankings "are calculated by a mathematical model — not determined by AI or a chatbot," and that they are "driven entirely by fight data, not opinion."
That precision matters. An Elo model has no judgement and no memory beyond its inputs. It cannot be lobbied, and it cannot be impressed.
Three things drive it: who a fighter faced, how the fight was won, and when it took place. A UFC representative told CBS Sports that whether you won and who you beat effectively account for around 95% of why a fighter sits where they do.
The Meta board also dropped something the media board had: there is no pound-for-pound ranking on it. Comparing a flyweight to a heavyweight requires exactly the kind of cross-divisional judgement an Elo model cannot make, because the two have never shared an opponent.
Both boards currently run in parallel while the transition completes, with the media rankings expected to be phased out.
What the Model Counts — and What It Ignores
The inputs are narrower than most fans assume, and the exclusions are the interesting part.
Time is weighted heavily. A win begins losing value after roughly five years and becomes effectively obsolete after ten. Inactivity is penalised once a fighter has been out for around 18 months, and longer layoffs are punished progressively. Frequent competition is rewarded.
Finishes earn only a small bonus. A decision win earns none. And strikes and takedowns are not inputs at all — the model does not look inside the fight.
That last exclusion is deliberate and defensible. Round-by-round statistics are noisy and measure activity rather than effectiveness, which is precisely the trap that makes UFC scoring so contested in the first place. But it means the model cannot distinguish a fighter who survived three rounds from one who dominated them, so long as both won.
The philosophy behind the decay rules is the clearest statement of intent: as the UFC put it, the point is that people should not be able to coast on reputation.
What Happened When the Two Boards Disagreed
Running both systems side by side produced an unusually honest experiment: two rankings of the same fighters, one human and one mathematical, published at the same time.
The gaps were large.
Jan Błachowicz was the starkest case. Ranked fourth by the media panel, he landed around fifteenth on the Meta board. The reason is structural rather than disrespectful: a former champion's title run is a vivid memory to a voter and an old, decayed result to the model.
Robert Whittaker, Derrick Lewis and Raul Rosas Jr. dropped off the Meta rankings entirely while remaining media-ranked. Brian Ortega and Yair Rodríguez landed well below the positions voters still gave them.
And the model produced at least one result it should not have been able to. Carlos Prates appeared ahead of Ian Machado Garry despite having lost to him — which contradicts the system's own stated principle that you move ahead of anyone you beat. Divisional rating gaps can produce that outcome, but it is exactly the kind of anomaly an auditable system is supposed to eliminate.
This is the honest trade-off, and it is the same one that arrived in football when expected goals came for the eye test. A model is consistent and explainable. It is also literal, and it will confidently produce an answer that everyone watching knows is wrong.
Why Rankings Still Do Not Decide Title Shots
Here is what did not change in June 2026.
The UFC did not adopt the Meta model as a qualification system. Matchmaking remains discretionary — driven by drawing power, timing, availability, injuries, contract status and what makes a card sell. Rankings inform that conversation. They do not govern it.
Dana White has been consistent and blunt about this for years, dismissing the significance of a fighter's ranking number when it conflicted with the fight the promotion wanted to make. Fighter representatives, for their part, have complained that ranking importance is invoked selectively — decisive when it supports the booking, irrelevant when it does not.
So the change is narrower than it first appears. The UFC replaced a subjective input with an objective one, and left the discretionary part of the process exactly where it was. What improved is transparency: you can now, in principle, reconstruct why a fighter sits at number six. What did not improve is consequence.
A Short Timeline
| Date | What happened |
|---|---|
| February 2013 | Rankings introduced; media panel begins voting |
| February 2025 | Development of the Elo model begins with Meta |
| 23 June 2026 | Meta UFC Rankings go live; media board continues in parallel |
| Ongoing | Media rankings expected to be phased out |
The Bottom Line
UFC rankings in 2026 are an Elo rating: a number per fighter, moved mainly by who they beat and when, decayed for age and inactivity, and blind to everything that happens inside the fight. That is a genuine improvement in consistency over thirteen years of averaged ballots, and a genuine loss of context.
But the most useful thing to understand about them has not changed since 2013. They describe merit. They do not distribute opportunity. Keep those two ideas separate and the rankings stop looking broken — because you stop asking them to do a job they were never wired to do.
For more on how fights are actually judged and how data is reshaping sport, see our UFC coverage hub and our explainer on how AI is changing football.
Frequently Asked Questions
Who decides the UFC rankings now?
Since 23 June 2026, a mathematical model does. The Meta UFC Rankings apply an Elo rating built with Meta, driven by fight results rather than votes. The older media panel, in which journalists from 22 publications submitted weekly ballots, is still published in parallel but is expected to be phased out. No human adjusts the Meta numbers after the fact.
Are the UFC rankings decided by AI?
No, and the UFC has been specific about this. Its explanation states the rankings are calculated by a mathematical model and not determined by AI or a chatbot. An Elo system is arithmetic: it updates ratings based on results and the strength of the opponent. It does not watch fights, form opinions, or generate text.
Does being ranked higher get you a title shot?
Not automatically, and this is the most persistent misconception about UFC rankings. The UFC is a promotion rather than a league, and matchmaking accounts for drawing power, timing, availability and contracts. Rankings are a published view of merit that informs those decisions without binding them. Fighters have been passed over while ranked highly, and others have been booked from outside the top ten.
Why do champions not appear in their division's rankings?
Under the media-panel rules, champions and interim champions were treated as already occupying the top positions in their weight class, so they were not voted on there. They remained eligible for the pound-for-pound list. This is why divisional rankings begin at number one with a contender rather than the titleholder.
Why did some well-known fighters drop sharply under the new system?
Time decay. A win starts losing value after about five years and is effectively obsolete after ten, and inactivity is penalised once a fighter has been out roughly 18 months. A former champion's title run stays vivid for a human voter but is a decayed result to the model. Jan Błachowicz moving from fourth on the media board to around fifteenth on the Meta board is the clearest example.
Is there still a pound-for-pound ranking?
The Meta board does not have one. Pound-for-pound requires comparing fighters who have never shared an opponent, which an Elo model cannot do meaningfully — its ratings only carry information within a division's network of results. The media rankings did include a pound-for-pound list, so it remains visible while that board is still published.
Sources
- UFC — Official Rankings and methodology
- CBS Sports — Meta UFC Rankings explained: everything you need to know about the new model
- Yahoo Sports — UFC's new Meta ranking system: how much do these numbers really matter?
- Yahoo Sports — UFC Meta rankings: what changed and why big names dropped
- MMA Mania — Meta UFC ranking system is not A.I.: "It's a mathematical model"
