Multiple Outcomes and Conditional Probability

When Outcomes Are Not Independent

Many sports betting situations involve outcomes that are not statistically independent. Failing to account for this is one of the most costly analytical errors a bettor can make, and it underlies why accumulator mathematics is frequently misapplied.

Two events are independent if the occurrence of one has no effect on the probability of the other. Flipping two separate coins is independent: the first flip has no bearing on the second. In sports betting, many pairs of events are also independent — the result of a La Liga match has no statistical relationship with the result of a Bundesliga match played simultaneously.

The Multiplication Rule for Independent Events

For independent events, the probability that both occur equals the product of their individual probabilities:

P(A and B) = P(A) × P(B)

Arsenal to win at 60% AND Tottenham to win at 55% (if independent): P = 0.60 × 0.55 = 33%. At fair odds, a double on these two selections should be priced at 1 ÷ 0.33 = 3.03. A bookmaker offering this double at 2.80 is applying a margin on top of the combined fair price.

Conditional Probability: When Events Are Not Independent

Conditional probability measures the probability of event B given that event A has already occurred: P(B|A). When A and B are not independent, this differs from P(B). Sports examples where conditional probability applies:

• Same-game multis (bet builders): "Team A wins AND Team A scores first" — these are correlated because teams that score first are more likely to win. The correct probability of both occurring is higher than P(A wins) × P(A scores first), and bookmakers typically price same-game multis by exploiting bettors' tendency to multiply the odds as if the events were independent.
• Half-time and full-time result: If a team is winning at half-time, their probability of winning the full match is substantially higher than the unconditional pre-match win probability. P(Win FT | Winning HT) ≠ P(Win FT).
• Asian Handicap and total goals: A match with a large handicap line (e.g. −2.5) typically implies a high-scoring match. The correlation between the favoured team covering and the total going over is positive.

The Bookmaker's Edge on Correlated Markets

Bookmakers routinely profit from bettors who multiply implied probabilities as if all events were independent. In the PunterStat dataset, analysing same-game multi combinations shows that the bookmaker's true margin on a three-leg same-game multi is typically 12–18%, versus 2.5–5% on a three-leg single-match accumulator from independent matches. The correlation adjustment is real and the bookmakers' pricing captures it — most bettors' multiplied-odds calculation does not.

Key Takeaway

The multiplication rule (P(A and B) = P(A) × P(B)) only applies when A and B are genuinely independent. In sports betting, correlated events — same-game multis, half-time/full-time combinations, Asian handicap with totals on the same match — require conditional probability reasoning. Ignoring correlation leads to systematic underestimation of bookmaker margin on these products: a bet builder that looks like it has a 3% margin per leg actually has a 12–18% combined margin once correlations are properly accounted for. This is among the most important structural lessons in all of betting mathematics.