Markets coveredMatch OddsCorrect ScoreOver / UnderFirst HalfSecond Half

Academy  /  Glossary

Glossary

Every term, in plain English

B C D E F G I K L M O P R S T V W

Back: Betting that something WILL happen. You win the stake times (odds − 1) if it does.
Backtest: Testing a strategy on past data to see how it would have done.
Bell curve / normal distribution: The hill-shaped spread of results you get when many small random effects add up. In the Model Library →
Calibration: Whether your predicted chances come true at the stated rate. Your 30% calls should win about 30% of the time. In the Model Library →
Closing line value (CLV): How much better your odds were than the final price before kick-off. The honest measure of an edge. In the Model Library →
Commission: Betfair's charge on your net winnings in a market. Ours is 2%. In the Model Library →
Confidence interval: The range your true result probably sits in, given your sample. In the Model Library →
Drawdown: The fall from your bank's highest point to its lowest point after it. In the Model Library →
Edge: How much your true chance of winning beats the chance implied by the price.
Expected value (EV): Your average profit or loss per bet if you placed it thousands of times. In the Model Library →
Fat tails: Extreme results happening far more often than the bell curve predicts. In the Model Library →
Green up: Backing and laying the same selection at different prices so you make the same profit whatever happens. In the Model Library →
Implied probability: The chance a price is suggesting: 1 ÷ decimal odds. In the Model Library →
Kelly criterion: A staking formula that sizes each bet by your edge and the odds. In the Model Library →
Lay: Betting that something will NOT happen. You act as the bookmaker: you win the backer's stake, or pay out stake × (odds − 1).
Liquidity: How much money is available to match at a price.
Mean: The average.
Overfitting: A model that has learned the noise in past data rather than a real pattern. In the Model Library →
Overround: How much a book's implied probabilities add up to more than 100%: the margin built into the prices. In the Model Library →
p-value: How often a no-edge bettor would do at least as well as you by luck alone. In the Model Library →
Risk of ruin: The chance of losing your whole bank (or a set part of it). In the Model Library →
ROI: Return on investment: profit ÷ total staked.
Sample size: How many bets or results your conclusion is based on. In the Model Library →
Significance: A result unlikely enough under "no edge" that luck is an unconvincing explanation. In the Model Library →
Slippage: Getting a worse price than the one you saw when you placed the bet.
Standard deviation (σ): How much individual results swing around the average.
Standard error (SE): How much your average result swings: σ ÷ √(number of bets).
t-distribution: A bell curve with fatter tails, used when your sample is small.
t-statistic: How many standard errors your result sits above zero.
Tick: One step on Betfair's price ladder. The step size grows with the price: 0.01 below 2.0, up to 10 above 100. In the Model Library →
Variance: Standard deviation squared, a measure of how noisy results are.
Walk-forward testing: Building on older data and testing on newer, unseen data, repeated through time. In the Model Library →

18+ only. Educational content, not financial or betting advice. Past results do not guarantee future returns. If gambling stops being fun, get free, confidential help at BeGambleAware.org.