Variance is a statistical measure of how widely outcomes are spread around their average. In a gambling context, it captures the natural swings a player experiences: even when the long-run expectation is fixed, actual results in the short term can differ sharply from that expectation because of chance.
Two bets can share the same expected value yet carry very different variance. A wager that wins small amounts frequently has low variance, whereas one that wins large amounts rarely has high variance. The expected return may be identical, but the journey — the size and frequency of the swings — is quite different.
Consider two hypothetical games, each with an expected value of zero over time. Game A pays even money on a near coin-flip, so results cluster close to the starting bankroll. Game B offers a rare payout of 100 times the stake at long odds, so most bets lose but occasional wins are large. Over 100 rounds, Game A’s bankroll barely moves, while Game B might be far ahead or deeply down depending purely on whether the rare payout landed. Same expectation, very different variance.
A frequent misconception is treating a winning or losing streak as proof of skill or a broken system. Short-term results are dominated by variance, and even a mathematically sound approach will endure extended downswings. Conversely, a flawed approach can look successful for a while simply through good fortune.
Managing variance is central to sustainable play. Larger bankrolls, smaller relative stakes and longer time horizons all reduce the chance that normal swings end a player’s involvement prematurely. Variance differs greatly between products: a high-volatility slot or a long-odds accumulator carries far wider swings than a low-margin, high-frequency market, so players should match their bankroll and expectations to the volatility of what they are playing.