Casino math becomes useful when it answers a decision, not when it produces a page of symbols. Before playing, a person mainly needs to know five things: what can happen, how often each result can happen, what each result pays, how much money will be wagered in total, and how widely the session can swing.
The questions below connect those pieces without pretending that a long-run average predicts tonight’s result.
What is probability?
Probability is the chance of an outcome. It can be written as a fraction, decimal, percentage, or odds.
On a fair six-sided die, the probability of rolling a 6 is:
1 ÷ 6 = 0.1667 = 16.67%
Casino games often involve several stages, conditional events, or unequal payouts, so the probability of winning is not enough by itself. A wager can win frequently and still be poor value if its wins are too small relative to its losses.
What is the difference between probability and payout?
Probability describes how often an event occurs. Payout describes what the casino returns when it occurs.
Suppose an event has a true probability of 1 in 6. Fair net odds would be 5 to 1: one winning result must compensate for five losing results. If a casino pays only 4 to 1, the missing unit creates the casino advantage.
This is the foundation of house edge: the difference between the probability-weighted fair return and the return offered by the rules.
What is expected value?
Expected value, or EV, is the probability-weighted average result per wager:
EV = Σ(Probability of outcome × Net result of outcome)
If a $10 bet wins $10 with probability 48% and loses $10 with probability 52%, then:
EV = (0.48 × $10) + (0.52 × −$10) = −$0.40
The bet’s expected value is negative 40 cents per $10 wager. That does not mean every bet loses 40 cents. Each individual result is still plus or minus $10. The 40-cent figure is the average that emerges from repeated wagers.
For a full explanation, see What Is Expected Value?.
What is house edge?
House edge is the casino’s expected profit expressed as a percentage of the original amount wagered. A 2% house edge means an average theoretical loss of $2 for every $100 of resolved action over extensive repeated play.
House edge = −EV ÷ Initial wager
A lower house edge is generally less expensive per dollar wagered, but it does not guarantee a lower session cost. A fast low-edge game played for hours can create more expected loss than a slower high-edge wager made only a few times.
House edge also depends on the exact rules and player decisions. “Blackjack house edge” is not one universal number because deck count, dealer rules, surrender, doubling rules, payouts, and strategy all matter.
Is RTP just the opposite of house edge?
For a simple house-banked game measured on the same wagering basis:
RTP = 100% − House edge
A 96% RTP corresponds to a 4% house edge. The relationship is straightforward, but the interpretation causes trouble.
RTP is a long-run design average across large amounts of play. It is not the percentage of a player’s bankroll that must return in one session, and it does not describe how the return is distributed. Two games can both have 96% RTP while one pays frequent small amounts and the other concentrates much of its return in rare features.
Read What Is RTP? for the difference between published return and personal session results.
What is total action?
Total action is the sum of all wagers made, including money recycled after wins.
Total action = Average wager × Number of wagering decisions
A player can enter with $300 yet create several thousand dollars of action. If the average bet is $10 and 210 decisions are made, total action is:
$10 × 210 = $2,100
The original $300 is the bankroll. The $2,100 is the amount exposed to the game’s edge. Casino mathematics prices the latter.
This is why a small visible bet can become costly through speed and time. See What Is Total Action? for how tables, slots, and hosts count it differently.
How do I estimate expected session loss?
The basic estimate is:
Expected loss = Average wager × Decisions × House edge
Consider a player averaging $10 per decision, making 70 decisions per hour for three hours at a 1.5% house edge.
- Decisions:
70 × 3 = 210 - Total action:
$10 × 210 = $2,100 - Expected loss:
$2,100 × 0.015 = $31.50
The $31.50 is a long-run average attached to that pattern of play. The actual session can finish ahead, lose the full bankroll, or land anywhere between. Use the Expected Loss Calculator to change stake, pace, duration, and edge without treating the output as a forecast.
Why can actual loss be so different from expected loss?
Because expected value identifies the center of a distribution, while variance describes its spread.
A $31.50 expected loss does not arrive as a smooth deduction. In blackjack it arrives through full-bet wins, losses, pushes, doubles, splits, and blackjacks. In slots it may arrive through many losing spins, small awards, bonuses, and occasional large prizes.
Short samples are noisy. A good wager can lose quickly; an expensive wager can win immediately. Neither event changes the wager’s expected value.
Does a larger sample remove variance?
No. The absolute amount by which results can vary generally grows as more bets are made. What improves is the stability of the average result per wager.
For independent, identically distributed wagers, expected result grows in proportion to the number of wagers, while standard deviation grows approximately with the square root of that number. This is why the average result tends to settle even though total dollar swings can remain large.
A thousand wagers are more informative than ten, but “more informative” does not mean “guaranteed to match expectation.” The NIST explanation of measures of scale describes variance and standard deviation as measures of data spread.
Does a bigger bankroll improve the odds?
No. Bankroll does not change probability, payout, RTP, or house edge. It changes how much ordinary variance the player can absorb before running out of money.
A $300 bankroll with $10 bets contains 30 betting units. The same bankroll with $50 bets contains six. Those are radically different risk profiles even if the game rules are identical.
This is why bankroll size and bet size must be considered together. Bankroll management controls exposure; it does not create positive expectation.
What is theoretical loss, and why do casinos use it?
Theoretical loss is an estimate of a player’s expected value to the casino:
Theoretical loss = Average bet × Decisions × House edge
The formula resembles expected loss because it describes the same mathematical relationship from the opposite side of the table. Casinos may use a version of theoretical loss when rating players, planning reinvestment, or reviewing game performance.
Actual loss is not the same thing. A player can win and still generate substantial theoretical value, or lose heavily while generating modest theoretical value. Comps are usually intended as a fraction of expected value, not reimbursement for what happened that night.
Can a betting system change expected value?
Changing bet size according to previous results does not change the probability or payout of the next independent wager. A progression can change the distribution of session results—many small wins and occasional large losses, for example—but it cannot remove the underlying edge.
A strategy changes expected value only when it changes something mathematically relevant, such as:
- choosing a different wager;
- using a decision strategy that affects probabilities or payouts;
- receiving a promotion with real monetary value;
- exploiting information or conditions unavailable in ordinary play;
- avoiding an unfavorable rule or paytable.
Bet sequencing by itself is not one of those changes. See Why Betting Systems Fail.
Are slot outcomes independent?
Approved random games are designed and tested so that outcomes follow the permitted random process. In an independent-spin model, previous losses do not increase the next spin’s probability of winning.
The UK Gambling Commission’s standard for generation of random outcomes requires game results and random-number generation to be acceptably random. The exact technical and regulatory framework depends on jurisdiction and product, but a player should not assume that a losing history creates a debt the next result must repay.
Progressive jackpot amounts, promotions, or must-award conditions can change the value of a wager under specified rules. That is different from a normal machine becoming due because it has recently lost.
Which number should I check first?
Use this order:
- Rules and payout: What exactly wins, loses, or pushes?
- House edge or expected value: What is the average price of the wager?
- Stake: How much is at risk per decision?
- Pace and duration: How much total action will be created?
- Variance: How rough can the path be?
- Bankroll limit: Can the planned loss be absorbed without chasing or adding money?
No single percentage answers all six questions. Casino math is a chain: probability and payout create expected value; stake and pace create exposure; variance shapes the session; bankroll determines how much of that session a player can withstand.
Useful next steps
Start with House Edge Explained, Expected Value Explained, and RTP Explained. Then connect the averages to Variance, Total Action, and Risk of Ruin. For casino operations, How Casinos Calculate Comps shows how the same numbers are applied to player ratings rather than personal predictions.