UFC Value

JavaScript is required

This research interface needs JavaScript to load forecasts and verify their supporting data.

Read the trust center

Loading interface…

Reload interface

Loading page views Odds

Display settings

Large gaps Gap cutoff (percentage points)

Marks the largest absolute difference between the official model and both current Polymarket references. Missing or stale prices are never marked. A gap is disagreement, not a betting recommendation. Saved only in this browser.

Event schedule

Upcoming & recent cards

Choose event

Upcoming card

Loading UFC schedule

Reading local prediction artifact

Forecasts —

Gaps ≥ 5 pp —

Largest model–market gap —

Moneyline comparison

Fight comparisons

Times shown in your local timezone.

Selected matchup

Matchup

Data quality · —

Fighter A

—

Fighter B

—

Model win chance —

No-vig market estimate —

Gap vs. market —

Hypothetical net result / $100 —

Largest model–market gap Checking favorites

Official model leans more toward Market comparison unavailable Waiting for a valid two-sided quote.

—

Price & data quality

Break-even —

Offered price — Quote status loading

Edge vs. break-even —

Policy input quality —

Quote Quote unavailable Source and time loading

Market comparison —

Model-implied return—

Theoretical value at market consensus—

More predictions

Prediction

Missing predictions stay blank until a tested model supports the exact outcome. Fight averages do not give over/under probabilities; winning a round is different from finishing in that round.

Method cells are whole-fight chances of that fighter winning by that method, conditional on a decisive result (draws and no contests excluded). Other / unresolved combines both fighters’ supplied Other mass, not draw or no-contest probability.

No method estimate is connected for this matchup.

Experimental estimates use a separately authorized saved model for research; accuracy is unconfirmed. Synthetic estimates are local demonstrations, not real forecasts. Neither changes winner forecasts, market comparisons, stake sizing or settlement.

Live round research

Historical strikingNot a forecast

Loading matchup signal

Round winnerNot modeled

—

Win methodResearch only

—

Winning round / endingResearch only

—

Other model views · comparison only

Forecast context

These inputs are descriptive signals, not causal explanations.

Selected-fighter probability comparison Seed spread

Research policy

Research decision

Evaluate side Sportsbook price American odds

Paper-policy size —

Enter a valid price to evaluate the policy.

Awaiting price

A real sportsbook price is required before the paper policy can be evaluated.

Historical model replay

Events by year

Choose a year and event.

Fighter database

Explore UFC fighters

Search by fighter name or weight class

Search or select a fighter.

Model–market evidence

Official model vs. market

Natural prospective freezes for the current official-model and named-book reference cohort.

Cumulative log loss Loading identical model and market observations

Loading model performance.

Evidence details

Research and governance

Model status & evidence

Governed research

Current decision

Career-sequence Transformer is the official forecast model

Edge not validated

Loading model evidence.

The interface could not load

Fights

Upcoming & recent cards

Loading UFC schedule

After round —

Fight comparisons

Matchup

Polymarket

Events by year

Explore UFC fighters

Official model vs. market

Actual bets

Paper bets

Model status & evidence

Career-sequence Transformer is the official forecast model

Model evaluation

Active research lanes

Conservative by construction

Record bet

Record result

Update command window

The interface could not load

Upcoming & recent cards

Loading UFC schedule

After round —

Fight comparisons

Matchup

Polymarket

Events by year

Explore UFC fighters

Official model vs. market

Actual and paper records

Actual bets

Paper bets

Model status & evidence

Career-sequence Transformer is the official forecast model

Model evaluation

Active research lanes

Conservative by construction