A house edge is a rate. Your hourly expected loss is a dollar estimate built from that rate and the amount of action you create. Treating the two as if they were the same number is one of the easiest ways to misunderstand casino cost.
A game with a 1% house edge does not mean “I will lose 1% of my buy-in every hour.” The percentage applies to the money wagered through the game. Chips can be won back, re-bet, split, doubled, moved to side bets, and recycled many times. A $500 bankroll can therefore generate several thousand dollars of total action before the session ends.
The core relationship is:
Expected loss = total action × house edge
An hourly version is:
Expected hourly loss ≈ average amount wagered per decision × decisions per hour × effective house edge
That is a planning model, not a forecast of what your next hour must produce.
Buy-in and action are different numbers
Suppose you buy in for $500 and make 100 wagers of $25.
Your starting bankroll is $500. Your total action is:
$25 × 100 = $2,500
If the wagers carry a 1% house edge, the mathematical expected loss associated with that volume is:
$2,500 × 0.01 = $25
You could still finish the session up $300, down the full $500, or anywhere between. The $25 is not a promise about the session result. It is the long-run average cost of that amount of action under the assumed edge.
This is why Total Action matters more than buy-in when you are pricing repeated casino play.
A low edge can become expensive at high speed
Compare two illustrative situations.
Game A
- average amount wagered per decision: $100
- decisions per hour: 60
- house edge: 1%
Expected hourly loss:
$100 × 60 × 0.01 = $60
Game B
- average amount wagered per decision: $10
- decisions per hour: 600
- house edge: 4%
Expected hourly loss:
$10 × 600 × 0.04 = $240
The second player bets much less on each decision but buys far more decisions and pays a larger percentage price. The expected hourly cost is four times as large in this example.
The speeds are illustrative rather than universal. Real pace depends on the game, number of players, dealer speed, machine settings, player decisions, interruptions, and whether several wagers are resolved at once. The point is the multiplication.
The related Decisions Per Hour page explains why pace belongs beside any house-edge chart.
House edge is the price per dollar; pace determines how many dollars you buy
Think of house edge as a percentage fee embedded in the game over repeated play. The fee is not charged once against your wallet. It is embedded each time money enters a negative-expectation wager.
If the average bet doubles and everything else stays constant, expected hourly loss doubles. If the number of decisions doubles, expected hourly loss doubles. If playing time doubles, expected session loss doubles. If the effective edge doubles, expected loss doubles.
That gives four direct levers:
Expected session loss ≈ average wager × decisions per hour × hours played × effective house edge
A player often focuses almost entirely on the final term—the edge—because it looks sophisticated. In real money, the first three terms can be just as important.
Mixed bets require a weighted calculation
The simple formula works best when the same wager is repeated at one house edge. Real sessions often mix several bets.
A blackjack player might combine a relatively low-edge main wager with a much more expensive side bet. A baccarat player may switch among Banker, Player, and Tie. A craps player may combine line bets with odds and proposition bets. A roulette player may place several wager types at once.
In those cases, estimate each category separately:
Expected loss = Σ(action on each bet type × house edge of that bet type)
The symbol Σ means “add the separate pieces together.”
Suppose a session produces:
- $4,000 of action on wagers carrying a 1% edge;
- $1,000 of action on a side bet carrying a 10% edge.
Expected loss from the main action:
$4,000 × 0.01 = $40
Expected loss from the side bet:
$1,000 × 0.10 = $100
Combined expected loss:
$140
Only one-fifth of the total action came from the side bet, but it created most of the expected loss in this example. Saying “I play a low-edge game” can therefore be misleading if the actual wager mix is expensive.
Extra money on splits, doubles, raises, and odds complicates average-bet shortcuts
Some games let the amount at risk change after the initial wager.
In blackjack, splits and doubles put additional money into action. In carnival games, an Ante may be followed by a larger Play or Raise decision. In craps, free-odds wagers may be added behind a line bet. In poker-style games, side bets can exist beside the main decision structure.
A crude “initial bet × hands” calculation can understate the amount actually wagered if those extra amounts are ignored.
The best model follows dollars rather than labels: how much money actually entered each wager category, and what edge applies to that category?
One hour of actual play will not look like the expected value
Expected Loss describes a long-run average. Actual casino results are noisy.
An expected hourly loss of $40 does not mean the game removes $40 smoothly every hour. One hour might end +$500. Another might end -$700. A volatile game can swing far above and below its expected value for a long time. For the underlying probability concept, the OpenStax treatment of expected value explains the probability-weighted average that repeated outcomes tend toward.
That is why two games can have the same expected hourly loss but feel completely different. One may produce many small results clustered around the average. Another may produce long losing stretches interrupted by large wins.
The expected value tells you where repeated outcomes center. Variance tells you how widely actual outcomes can spread around that center. See Variance for that second concept.
RTP is the mirror image of house edge, not a short-session refund rate
On many electronic games, players see return to player (RTP) rather than house edge.
A 96% RTP corresponds conceptually to a 4% house advantage before considering rule-specific details:
House edge ≈ 100% - RTP
That does not mean a machine will return $96 from every $100 during your session. RTP is a long-run average across a large amount of play. Short sessions can end far above or far below it. The UK Gambling Commission player guide to RTP makes the same distinction between long-run return and individual sessions.
The same hourly-cost logic still applies:
Expected loss ≈ coin-in × house edge
If fast play creates $5,000 of coin-in, a 4% long-run edge implies $200 of expected loss associated with that amount of action. The bankroll used to create the $5,000 could be much smaller because credits are repeatedly returned and wagered again.
Casino “theo” is related but not identical to your exact mathematical cost
Casinos often estimate a rated player’s Theoretical Loss, commonly called theo, from average bet, time, game speed, and an assumed house edge or hold model.
That number is useful for player development, hosts, and comp decisions. It is still an operational estimate.
The casino may use standardized decisions per hour. A floorperson may round the recorded average bet. A player may switch wager types. A slot system may calculate value from tracked coin-in and product configuration. The casino’s model may therefore differ from a precise reconstruction of every dollar the player actually wagered.
Keep the questions separate:
- Mathematical expected loss: What is the expected cost of the wagers actually made?
- Casino theoretical value: What expected value does the casino’s rating system assign to the recorded play?
They overlap, but they are not guaranteed to be identical.
The same house edge can produce very different hourly costs
Imagine two players at the same game with the same 2% edge.
Player 1 bets $10 and makes 50 decisions in an hour:
$10 × 50 × 0.02 = $10 expected hourly loss
Player 2 bets $100 and makes 100 decisions in an hour:
$100 × 100 × 0.02 = $200 expected hourly loss
The house edge is identical. The expected dollar cost differs by a factor of twenty because the amount of action differs.
This is the cleanest proof that “the house edge is 2%” is not an answer to “what might this game cost me per hour?”
Comps do not change the basic exposure equation
Players sometimes subtract free drinks, points, food, free play, or room value from gambling cost and conclude that the edge has effectively disappeared.
Comps can offset part of expected cost, but they should be valued realistically. A $20 meal you would never have bought for cash is not necessarily worth $20 to you. Free play is not always equivalent to cash. A room has value only if it replaces a cost you intended to pay.
More importantly, playing extra time or increasing bets merely to earn a reward adds new action. The extra expected loss can easily exceed the value of the comp.
See Why Comps Hide Real Losses and What Comps Are Really Worth for that distinction.
The most useful comparison is an exposure budget
If the goal is cost control, ask four questions before the session:
- What is my average total wager per decision? Include recurring side bets.
- How many decisions per hour am I likely to buy? A fast machine can create far more action than a crowded table.
- What edge applies to the wagers I actually choose? Do not use the best bet on the game to describe every bet you make.
- How long do I expect to stay in action? Time multiplies everything above.
Then estimate:
Expected session loss ≈ average wager × decisions per hour × hours × effective edge
If the wager is cut from $50 to $25 and everything else remains constant, expected cost is cut roughly in half. If a two-hour session becomes one hour, it is cut roughly in half again. Slower play can reduce exposure further.
Those choices do not improve the underlying odds. They reduce the quantity of negative-expectation action purchased.
House edge is necessary information, but it is not the invoice
A house-edge table answers: How expensive is each dollar wagered on average?
An hourly-loss estimate answers: How many dollars are likely to be wagered through that price during this amount of time?
You need both.
The percentage matters, but average bet, pace, wager mix, extra wagers, and duration determine what that percentage becomes in money. House edge is the rate. Total action is the base it applies to. Expected hourly loss is the result of putting those pieces together.
For a direct estimate, use the Expected Loss Calculator and compare it with Expected Loss in Real Sessions.