What are the Meta UFC Rankings?

The Meta UFC Rankings are the UFC's official fighter rankings since 22 June 2026, produced by an Elo-based mathematical model built with Meta instead of by a panel of media voters. Every fighter in a division receives a rating from their fight results, and the ratings are sorted to produce the top 15. They recalculate automatically every Monday after a UFC event, with no human intervention.

Do the Meta UFC Rankings count significant strikes and takedowns?

No. Volume statistics are not inputs. A dominant 25-minute shutout and a razor-thin split decision over the same opponent are worth the same, because both are simply a win. Only the result, the opponent, the method, the date and the division matter.

Are the Meta UFC Rankings more accurate than the media rankings?

By the one measure the UFC has published, yes. When two top-15 fighters met, the higher-ranked fighter won 53.4 percent of the time under the media rankings. During the Meta trial period that figure was 63.5 percent — a ten-point improvement in predicting the winner.

How often do the Meta UFC Rankings update?

Automatically on the Monday following every official UFC event. There is no vote, no meeting and no appeal process.

Meta UFC Rankings Explained — How the New System Works vs Media Rankings

Q: What is the difference between the Meta rankings and the media rankings?

The media rankings, used from 2013 to 2026, were a vote by a panel of MMA journalists, so reputation, narrative and voter inconsistency all fed into them. The Meta rankings use only measurable fight data — who won, who they beat, by what method, how recently and in which division. Nothing else. The practical effect is that inactive ex-champions fall much faster and active fighters on winning streaks climb much faster.

Q: How do the Meta UFC Rankings work?

It is an Elo system. Each fighter carries a rating; beating someone rated above you gains more than beating someone below you, and the size of the gain scales with how unlikely the win was. A finish adds a small bonus. Recent fights count for more, wins older than five years lose value and are worthless after ten, and inactivity beyond 18 months is penalised on an accelerating curve. The UFC has said that who you beat accounts for roughly 95 percent of a fighter's placement.

Q: Why did Raquel Pennington drop to number 15?

Pennington was ranked #2 at women's bantamweight on the media board and #15 on the Meta board. She had not fought since October 2024, so the inactivity decay applied, but most of the fall came from her wins ageing out: her victories were over Mayra Bueno Silva, who has since gone winless in six, and Ketlen Vieira and Aspen Ladd, who are no longer on the UFC roster. The media panel kept crediting her for being a former champion; the algorithm does not recognise the concept.

On 22 June 2026 the UFC retired the media-panel rankings it had used since 2013 and replaced them with the Meta UFC Rankings — an Elo-based model built with Meta that recalculates every Monday with no human involvement. This page is what it counts, what it deliberately ignores, and why the board looked so different the morning it switched on.

The current Meta board

Pulled from our fighter database, which refreshes from the official rankings every Monday morning.

Why the UFC changed the rankings

The media rankings launched in 2013 and worked the way most sports polls work: a panel of accredited MMA journalists submitted ballots, and the ballots were averaged. It was transparent in principle and messy in practice. Voters disagreed on methodology, some voted rarely, and the board had enormous inertia — a former champion could drift near the top for two years on reputation alone while an active contender stacked wins below them.

Dana White had been publicly unhappy with it for years before the change.

"I've been unhappy with the rankings and always believed there had to be a better way."

Development with Meta, the UFC's official fan technology partner, started in February 2025 and ran through a trial period before the June 2026 launch. The stated goal was not to rebuild the same board with maths, but to produce something measurably better at the thing a ranking is supposed to do: tell you who would win.

How the Meta UFC Rankings actually work

Underneath the branding it is an Elo rating system — the same family of maths used in chess and in most sports prediction models. Every fighter in a division holds a numeric rating. When two fighters meet, the model already has an implied probability of each winning based on the gap between their ratings. The winner takes points from the loser, and the size of the transfer depends on how surprising the result was.

What moves your rating

The opponent. This is the dominant term by a distance. Beating the #2 contender is worth many times beating the #14, and beating an unranked opponent is worth close to nothing if you were already ranked.
How unlikely the win was. An underdog win transfers far more rating than a favourite confirming the form book. This is why upsets cause violent movement.
The method. A knockout or submission carries a bonus — but the UFC has been explicit that it is a small one. There is no difference at all between a unanimous and a split decision.
Recency. Recent fights carry more weight. A win starts losing value after about five years and is worth nothing after ten.
Activity. Fighting often is rewarded with a velocity bonus, independently of results.
The division. Each weight class is modelled separately with its own sensitivities, and a fighter who changes division is treated as a new entity with an adjusted score carried across.

What is deliberately ignored

Significant strikes, takedowns, control time. None of it is an input. A 150-strike shutout and a lucky flash knockdown are the same thing to the model: a win.
Being a former champion. There is no legacy credit of any kind. The belt buys you nothing once you stop fighting.
Marketability, callouts, narrative. The model does not know who is popular.
Context. It cannot know a fighter was injured, that a decision was a robbery, or that a division is unusually shallow.

The head-to-head guarantee

One rule sits above the maths: a fighter is never ranked below someone they have beaten. Under a plain Elo model a #12 who upsets a #3 would still be rated below them the next morning, because one result cannot close that gap. The Meta system overrides that, which is why the boards move so dramatically after upsets compared with the old panel — where a voter might reasonably have said "one win doesn't change my view of these two."

The 18-month cliff

Nothing happens to a fighter's rating for the first 18 months of inactivity. Past that, the penalty applies and compounds — longer layoffs are penalised exponentially, not linearly. A fighter who sits out two years slides a few places. One who sits out four is gone from the top 15 entirely, belt or no belt.

Meta rankings vs media rankings: the differences that matter

	Media rankings (2013–2026)	Meta rankings (2026–)
Produced by	Panel of media voters	Elo model, no human input
Updates	Ad hoc after events	Automatic, every Monday
Ex-champion credit	Yes, informally	None
Inactivity punished	Slowly, at voter discretion	Hard, after 18 months
Finish bonus	Whatever the voter felt	Small and fixed
Volume stats used	Informally, yes	No
Winner above loser	Not guaranteed	Guaranteed
Higher rank wins	53.4%	63.5%
Pound-for-pound	Yes	Not yet

That last row of numbers is the strongest argument for the change. In top-15 versus top-15 fights, the higher-ranked fighter won 53.4 percent of the time under the media panel — barely better than a coin flip. During the Meta trial period the figure was 63.5 percent. In a sport as chaotic as MMA, a ten-point gain in predictive power is a lot.

Who won and lost in the switchover

Raquel Pennington: #2 → #15

The single most-discussed movement of the changeover. The former women's bantamweight champion sat at #2 on the media board and landed at #15 on the Meta board, clinging on. She had not competed since losing the belt to Julianna Peña in October 2024, which triggers the decay — but the layoff alone does not explain a 13-place fall. The bigger factor is that the wins that got her there had stopped being worth anything: her victories came over Mayra Bueno Silva, since winless in six, plus Ketlen Vieira and Aspen Ladd, neither of whom is on the roster any more. The panel was still crediting a former champion. The model has no such concept.

Pat Sabatini: unranked → #7

The mirror image, and the case the UFC leans on. Sabatini was outside the featherweight top 15 under the media panel despite a four-fight winning streak against real opposition, with only one finish in the run. The Meta model does not need finishes or highlight reels — it needs wins over rated opponents, and he had them. He debuted at #7. Whether the matchmaking follows and hands him a top-six opponent is the open question that decides whether any of this means anything.

The honest criticisms

It cannot see context. Injuries, robberies, short-notice fights and shallow divisions are invisible to it.
Garbage in, garbage out. If a division has few fights between its top contenders, the model has little to work with and the order is close to arbitrary.
Nobody can audit it. "Objective" and "transparent" are not the same thing. The factors are published; the formula, the weights and the ratings themselves are not, so a fan who disagrees with a placement has nothing to check it against.
Rankings still may not drive matchmaking. The system is only meaningful if the UFC books according to it. If Sabatini never gets a top-six opponent, the board is a scoreboard with nothing riding on it.

Try it yourself

Because the factors are public even though the formula is not, you can build a model that behaves the same way — which is what our Meta UFC Rankings Simulator does. Pick a division, book any fight, choose the winner and the method, and see the board that would result. Stack up a whole card, or sit a contender out for three years and watch the decay do its work. It is an approximation and it says so on the page, but the direction and rough size of every move follows the rules above.

FAQ

Are the Meta rankings the official UFC rankings now?

Yes. Since 22 June 2026 they are the default board at ufc.com/rankings. The old media rankings are still viewable there behind a toggle, but Meta is what you see first and what the UFC treats as official.

What is the difference between the Meta rankings and the media rankings?

The media rankings were a vote by a panel of MMA journalists, so reputation, narrative and voter inconsistency all fed into them. The Meta rankings use only measurable fight data — who won, who they beat, by what method, how recently, in which division. Nothing else. In practice inactive ex-champions fall much faster and active fighters on streaks climb much faster.

How do the Meta UFC Rankings work?

It is an Elo system. Every fighter carries a rating; beating someone rated above you gains more than beating someone below you, and the gain scales with how unlikely the win was. A finish adds a small bonus, recent fights count for more, wins older than five years lose value and are worthless after ten, and inactivity past 18 months is penalised on an accelerating curve.

Is it artificial intelligence?

Mostly not, despite the framing. The core is a conventional Elo rating system with decay terms. The UFC describes machine learning and its own domain expertise as part of the build, but the day-to-day mechanism is maths you could write down on paper.

Do the Meta rankings count significant strikes and takedowns?

No. Volume statistics are not inputs at all. A 25-minute shutout and a razor-thin split decision over the same opponent are worth exactly the same, because both are simply a win. Only the result, the opponent, the method, the date and the division matter.

Why did Raquel Pennington drop to number 15?

She was #2 on the media board and #15 on the Meta board. She had not fought since October 2024, which triggers the inactivity decay — but most of the fall came from her wins ageing out. Her victories were over Mayra Bueno Silva, since winless in six, plus Ketlen Vieira and Aspen Ladd, neither of whom is on the roster any more. The panel was still crediting a former champion; the model has no such concept.

Who benefited most from the switch?

Active fighters on streaks against real opposition. Pat Sabatini is the clearest case — unranked at featherweight under the media panel, he came in at #7 on the Meta board on the strength of a four-fight winning streak, with only one finish in the run.

Can a fighter appeal their Meta ranking?

No. There is no human in the loop by design — removing the panel was the entire point. There is no vote, no meeting and no appeal process.

Are there Meta pound-for-pound rankings?

No. Meta ratings are only comparable within a single division, because each division is modelled separately with its own sensitivities. The UFC has said a pound-for-pound board would need a different model, and none has been announced.

More UFC tools

Meta rankings simulator · UFC Matchmaker Simulator · UFC Card Builder · MMA Career Simulator · UFC weight classes · Fighter database · All MMA games

Sources: UFC press release, 22 June 2026; MMA Junkie, 22 June 2026; CBS Sports briefing on the Meta rankings methodology. MMAFightSim is not affiliated with, endorsed by or connected to UFC, TKO Group Holdings, Zuffa or Meta. "Meta UFC Rankings" is referenced descriptively.

Meta UFC Rankings Explained