UFC betting & AI glossary
Every term FightVex uses, in plain English — from moneylines and no-vig probability to EV, CLV and Monte Carlo.
- Moneyline
- A bet on who wins the fight. Negative odds (e.g. -150) mark the favourite; positive odds (e.g. +130) mark the underdog and a bigger payout.
- Underdog
- The fighter the market expects to lose, priced at plus odds. FightVex's Upset Radar surfaces underdogs the Vex AI actually favours.
- Implied probability
- The win chance baked into a price. -200 implies about 67%; +150 implies about 40%. It's how you turn odds into a probability.
- No-vig (fair) probability
- Implied probability with the bookmaker's margin (the 'vig') stripped out, so both sides add to 100%. FightVex compares the Vex AI's number to this fair line.
- The vig (juice)
- The bookmaker's built-in margin — why both sides of a fight imply more than 100% combined. It's the house edge casual bettors quietly pay.
- Value / edge
- When your estimated probability is higher than the market's implied probability for the same outcome. Positive edge is the whole point of betting profitably.
- Expected value (EV)
- The average result of a bet if it were repeated many times, given your probability and the odds. Positive EV means a mathematically profitable bet over the long run.
- Closing-line value (CLV)
- Whether you bet at a better price than where the market closed. Consistently beating the close is the strongest evidence of a real edge — FightVex tracks the model's CLV.
- Kelly criterion
- A bankroll formula for how much to stake based on your edge and the odds, balancing growth against risk of ruin. FightVex includes a Kelly calculator.
- Monte Carlo simulation
- Playing a fight out thousands of times with realistic randomness, then counting the outcomes. FightVex runs each bout 50,000 times to produce its win %, method and round.
- Calibration
- How well stated probabilities match reality — when a model says 65%, it should win about 65% of the time. FightVex's predictions are calibrated against thousands of real results.
- Method of victory
- How a fight ends: KO/TKO, submission or decision. The Vex AI estimates the probability of each, not just the winner.
- Strength of schedule (SOS)
- The quality of opponents a fighter has faced. Beating elite competition counts for more than beating journeymen, and the model accounts for it.
- Short notice
- Taking a fight with little time to prepare or cut weight — a real disadvantage the model debuffs when it applies.
- Missed weight
- Failing to make the contracted weight at the official weigh-in, usually meaning a hard, draining cut. FightVex factors it into fight-week adjustments.
- Backtest
- Re-running the model over historical fights — using only data available before each bout — to measure real accuracy without hindsight (FightVex: ~64.6% over 10,000+ bouts).