What Expected Value Means in Gambling Math
Expected value, often written as EV, is the weighted average of every possible outcome of a bet. To calculate it, multiply each outcome by its probability and sum the results. In a fair coin flip with a $1 bet that pays $1 on heads and loses $1 on tails, the expected value is zero. In roulette the expected value of every standard bet is negative because the wheel has one or two extra pockets compared to the payout structure. Expected value is the single most important concept for understanding any betting system, because it tells you the average cost of the wager regardless of the order of wins and losses.
How Roulette Generates a Negative Expected Value
On a European wheel the red bet wins 18 of 37 spins and loses 19 of 37 spins. A $1 bet pays $1 on a win and loses $1 on a loss. Expected value per $1 bet is (18/37) times 1 plus (19/37) times -1, which equals -1/37, or approximately -2.70 cents. That number is the house edge expressed as a negative expected value. Every standard roulette bet on a European wheel shares this same -2.70% expected value because the payout ratios are calibrated to produce it on each bet type.
European Roulette EV in Detail
European roulette uses 37 pockets including the single zero. Across every standard bet from straight up to red and black, the expected value per unit wagered is approximately -0.027. Multiplied by total wagered volume across a session, this produces the expected loss for that session. A player who wagers $1,000 in total across many spins should expect to lose about $27 on average, regardless of bet type, regardless of strategy and regardless of the order of spins.
American Roulette EV in Detail
American roulette has 38 pockets because of the added double zero. On red, expected value per $1 becomes (18/38) times 1 plus (20/38) times -1, which equals -2/38, or approximately -5.26 cents. Across all standard bets the expected value per unit wagered is approximately -0.0526. The same $1,000 wagered on an American wheel produces an expected loss of about $53. The extra zero almost doubles the long-run cost of playing.
Total Wagered and Expected Loss
The most useful expected value formula in practice is expected loss equals total wagered multiplied by the wheel house edge. The strategy simulator uses this exact calculation at the end of every run. Total wagered is the sum of all stakes placed across all spins, not the net amount bet. A Martingale player who doubles after each loss generates a much higher total wagered figure than a flat bettor playing the same number of spins, which is why Martingale produces a higher expected loss over time, not a lower one.
Why Doubling Stakes Does Not Remove EV
Each individual bet has its own expected value. Doubling the stake on the next bet doubles the absolute expected loss of that bet, but the percentage expected value per dollar wagered stays the same. Adding bets together never changes the average. If every step of a Martingale sequence has the same negative percentage expected value, the whole sequence inherits that same percentage. Strategy cannot reorder probabilities into an advantage.
Why Short-Term Wins Can Hide Negative EV
Variance, not expected value, dominates short sessions. Martingale produces many small winning sessions because the early steps of the progression usually succeed. A player who walks away after a profitable short session feels the system works. The negative expected value reasserts itself once the player keeps playing through enough losing streaks to expose the doubling sequence to its weak point. Expected value is a long-run average, not a guarantee of any specific session.
Why Long Sessions Expose the House Edge
The law of large numbers says that as the number of trials grows, the average result converges on the expected value. A long roulette session with many bets has a net result close to the theoretical expected loss. A short session does not. This is why betting systems that look brilliant in two-hour casino visits steadily lose value in extended simulations like those produced by the strategy simulator on this site.
Expected Loss Formula in Practice
The clean formula is expected loss equals total wagered multiplied by the house edge written as a decimal. For European roulette the multiplier is 0.027. For American roulette it is 0.0526. For French roulette with La Partage actively applied on even-money bets, the multiplier on those bets falls to 0.0135. Pick the wheel, total your wagered volume, multiply and you get a realistic estimate of expected long-run loss.
Calculator Limitations and Honest Assumptions
The Martingale calculator on this site reports a theoretical success probability and risk of ruin. Those numbers are honest given the model, but they are not predictions of any specific session. Expected value still rules the long run. A high success probability for a single sequence does not change the expected loss across many sessions, because each successful sequence wins only one base bet of profit while each failed sequence loses the entire progression's exposure.
Responsible Interpretation of Expected Value
Expected value is a math concept, not a betting plan. It tells you what the average cost of play looks like over time. It does not advise anyone to play. The point of explaining expected value clearly on this site is to make sure no reader confuses Martingale's pleasant short-run pattern with a real mathematical edge. Read the responsible gambling page for the limits anyone studying roulette systems should respect.