The amount you bring to a casino is a loss limit, not the automatic price of the session. A better estimate of the price is expected loss per hour: the probability-weighted cost of the wagers you place during that hour.
The estimate will not predict what happens tonight. A player with a $20 hourly expectation can win $500 or lose $600 in one hour. The calculation answers a different question: How much value does this pattern of play transfer to the house on average under the stated assumptions?
The core formula
For one repeated wager type:
Expected loss per hour = Average wager × Decisions per hour × House edge
Define:
B= average amount wagered per decision;N= number of decisions per hour;H= house edge as a decimal.
Then:
Hourly expected loss = B × N × H
Suppose a player averages $20 per hand, plays 60 hands per hour, and faces an effective 1% house edge:
$20 × 60 × 0.01 = $12 expected loss per hour
That is a long-run average for that model. It is not a guaranteed $12 deduction.
Total action is the number players usually miss
The formula can also be written:
Expected loss = Total action × House edge
where:
Total action = Average wager × Number of decisions
In the example above:
$20 × 60 = $1,200 hourly action
$1,200 × 0.01 = $12 expected loss
A player may buy in for only $200 while cycling far more than $200 through the game. Wins are re-wagered, chips move back and forth, and the same money can support many decisions. The buy-in therefore understates the amount exposed to the house edge.
More than one wager requires more than one calculation
Many games combine wagers with different edges. Do not average them casually. Calculate each component separately:
Total hourly expected loss = Σ(Bᵢ × Nᵢ × Hᵢ)
The symbol Σ means add the expected loss for every wager category i.
Suppose a blackjack player makes:
- a $20 main wager on 60 hands at an estimated 0.8% disadvantage;
- a $5 side bet on the same 60 hands at a 7% house edge.
Main-game expectation:
$20 × 60 × 0.008 = $9.60
Side-bet expectation:
$5 × 60 × 0.07 = $21.00
Combined hourly expectation:
$9.60 + $21.00 = $30.60
The side bet is only one quarter of the base wager, yet it contributes more than twice the expected loss in this example. “I only bet $20” is not an accurate description of the action.
Worked estimates for different games
The figures below are illustrations, not universal rates. Actual pace, rules, decisions, paytables, occupancy, dealer speed, and jurisdiction can change the result.
Blackjack
Assume:
- $25 average total main-game wager;
- 55 hands per hour;
- 0.7% effective house edge.
$25 × 55 × 0.007 = $9.63 expected loss per hour
A player making strategy errors, using weaker rules, or adding side bets can face a much larger cost.
Roulette
Assume:
- $30 total spread across the layout each spin;
- 35 spins per hour;
- 5.26% edge on a double-zero wheel for the wagers used.
$30 × 35 × 0.0526 ≈ $55.23 expected loss per hour
The number of chips is irrelevant. Total money placed across all positions is the wager.
Slots
For slots, use wager per spin and spins per hour:
Hourly expected loss = Wager per spin × Spins per hour × (1 − RTP)
Assume:
- $1.50 per spin;
- 500 spins per hour;
- 94% theoretical RTP.
House margin:
1 − 0.94 = 0.06
Hourly expectation:
$1.50 × 500 × 0.06 = $45
The actual result can be dominated by rare bonuses and jackpots. The calculation estimates cost, not volatility.
Craps with multiple bets
Craps is difficult to summarize with one edge because wagers resolve at different speeds and can remain working for several rolls. A pass-line bet, odds, place bet, and proposition bet each require their own decision frequency and edge.
Using “total chips on the layout × one average house edge” can be badly misleading. Why expected value needs context explains why the unit—per roll, per decision, or per hour—must be stated.
Pace is often as important as the edge
Doubling the number of decisions doubles expected loss when wager and edge remain constant.
A $10 wager at a 2% edge played 40 times per hour:
$10 × 40 × 0.02 = $8 per hour
The same wager played 400 times per hour:
$10 × 400 × 0.02 = $80 per hour
This is why a low-edge game can be expensive when played rapidly and why speed can matter more than edge.
Pace estimates should reflect actual play:
- full tables are usually slower than heads-up tables;
- breaks and dealer changes reduce decisions;
- autoplay or rapid electronic interfaces can increase decisions;
- bonus rounds may add time without adding the same number of paid base spins;
- thinking time in strategic games can reduce pace;
- side bets may be resolved on every round even when the base game remains slow.
House edge is not always the same as hold
These terms are often mixed together.
- House edge is the mathematical expected percentage of the initial or resolved wager under specified rules.
- Hold is an accounting result, often casino win divided by a measure such as drop, handle, or turnover.
- RTP is return divided by total wagered under the defined calculation.
- Actual session loss is what happened to one player.
A casino can hold 20% of buy-ins during a period without the game having a 20% house edge. Players recycle chips, leave with winnings, buy in again, and play different lengths of time.
The UK Gambling Commission’s guidance on calculating actual RTP from wins and turnover demonstrates why turnover—not starting cash alone—is central to performance measurement.
Theoretical loss is not a forecast of tonight
The expected-loss model says nothing about the width of the outcome distribution unless variance is also estimated.
Two games can both cost $20 per hour in expectation while one usually finishes near the starting balance and the other produces frequent large swings. Bankroll needs and emotional pressure can be very different.
This distinction matters because players often reject the model after a win: “I was expected to lose $20, but I won $300.” The win is entirely compatible with a negative expectation. Expected value is an average across possible outcomes, not a schedule imposed on every hour.
Likewise, losing $400 in an hour does not prove the game has a $400 hourly expected loss. It may be a high-variance result.
Your effective edge may be unknown
The published or theoretical edge assumes specific rules and behavior. Your real estimate may differ because of:
- strategy errors;
- uncertain paytable or RTP configuration;
- variable side-bet use;
- changing wager sizes;
- promotions or rebates;
- free play with wagering restrictions;
- incomplete pace estimates;
- game states that affect value;
- tips, fees, travel, and other external costs.
A range is often more honest than one number.
If a slot session is estimated at 400–600 spins per hour, $1–$2 per spin, and a 4%–7% house margin, the hourly expectation spans:
Low estimate:
400 × $1 × 0.04 = $16
High estimate:
600 × $2 × 0.07 = $84
That range is wide because the inputs are uncertain. Pretending the answer is exactly $42 would create false precision.
Comps should not be subtracted at face value
A casino may estimate your theoretical loss to determine offers. The reward is usually designed to return only part of the expected value of the play.
If expected loss is $100 and the player receives a room advertised at $150, it does not follow that the session had positive value. The room may have low marginal cost to the casino, may be unused, may require another trip, and may not be worth $150 to the player.
Count only the personal cash value of benefits actually used. What comps are really worth provides a fuller method.
Free play changes cash flow, not necessarily the game edge
Promotional credits can reduce the cash cost of a session, but the correct treatment depends on their rules.
If $50 of free play is converted into $42 cash on average, the benefit is approximately $42—not $50 plus whatever was wagered. If further wagering is required, include the expected cost of that action.
Keep promotional value separate from ordinary wagers so the ledger remains clear.
A practical personal estimate
Use five steps:
- Record the average total wager per decision, including side bets.
- Estimate actual decisions per hour from the game and table conditions.
- assign a defensible edge to each wager category.
- Calculate each component and add them.
- Multiply by planned hours, then compare the result with the maximum loss budget.
For a session of T hours:
Session expected loss = Hourly expected loss × T
If the hourly estimate is $35 and the planned session is three hours:
$35 × 3 = $105 expected session loss
The maximum possible or plausible loss may be much larger. A budget should be based on affordability, not only on the average.
The session loss calculator can organize the basic inputs, but it cannot repair inaccurate assumptions.
Track actual results separately
Maintain two columns:
- Expected cost, calculated from action, pace, and edge;
- Actual result, calculated from cash in and cash out.
Over a small number of sessions they may differ sharply. Over a long, consistent sample, the actual average may move toward the model if the assumptions are correct. Changes in game mix, stakes, promotions, or behavior can prevent clean comparison.
Do not alter the expected-value formula after seeing the result just to make it fit.
The hard truth
Your casino cost is not defined by the size of the first buy-in or by how small one chip looks. It is created by the total amount wagered, how quickly it is wagered, and the edge attached to each decision.
Expected hourly loss is an estimate with uncertainty. Used honestly, it converts “I only played for a while” into a measurable price. That price may still be acceptable as entertainment, but it should be visible before the session—not reconstructed after the money is gone.