The question hides a category error
A casino does not lose money on a game the way a player loses money on a hand. The operator's position is built from the rules themselves, and every wager carries a small mathematical disadvantage for the person making it. When someone asks which game a casino loses the most money on, they are usually asking something more precise: which game gives a player the smallest expected loss per unit wagered, or which game produces the most short-term variance that can make the casino's cage pay out large sums on a given night.
Both readings are answerable, but they lead in different directions. The first is a question about house edge. The second is a question about volatility and the size of the bets a table will accept. Neither is a question about a game where the operator is systematically losing.
How the edge is built into the bet
House edge is the mathematical advantage the operator holds on a given game, expressed as a percentage of each wager. It is not a fee deducted from your balance before you play. It is the difference between the true odds of an event and the payout the game offers when that event occurs. The gap is small on any single wager, but it repeats every time a bet is settled.
Return to Player, or RTP, is the same relationship viewed from the other side. RTP is a theoretical percentage of total wagered money a game returns to players over a very large number of spins or hands. For the same bet under the same rule set, house edge and RTP sum to 100%. If a game returns a high percentage of coin-in over its lifetime, the house keeps a small percentage. If the house edge is larger, the long-run return is smaller. The two numbers describe one mechanism.
What matters for the question is that the edge is not a property of the casino's mood or the day of the week. It is fixed in the payout table and the rules. A roulette wheel pays winning single-number bets at a rate that is slightly short of the true odds against hitting that number. A blackjack table pays a natural at a premium but still resolves all other hands under rules that leave the dealer with a structural advantage. The game does not need to cheat. The arithmetic is enough.
Comparing the games people usually mean
Different games carry different edges, and within a single game different bets can carry dramatically different edges. The table below compares the structural position of the games most often named in this conversation. No percentages are given here because the exact figure depends on the rule set in force at a particular table or the configuration of a particular machine. The relationships are what matter.
| Game | Where the edge sits | What the player controls | What the player does not control |
|---|---|---|---|
| Blackjack | Dealer acts last; player can bust before the dealer plays | Hit, stand, double, split decisions against a published basic-strategy chart for the specific rule set | Deck count, dealer stand rules, doubling and splitting restrictions |
| Baccarat | Payout on the banker bet is short of true odds after the commission; the tie bet carries the highest edge on the table | Which of the three main bets to make | Card drawing rules are fixed and automatic |
| Roulette | The zero pocket or pockets; payouts are calculated as if the zeros did not exist | Which bets to place; European single-zero versus American double-zero layout where both are offered | Wheel physics, pocket order, payout multiples |
| Craps | Different bet types on the same table carry materially different edges; some bets resolve on a single roll, others over many rolls | Which bets to place; some bets can be removed or reduced after the point is established | Dice probabilities are fixed; 36 equally likely combinations |
| Slots | RTP is set in the game's configuration; volatility determines the shape of the return | Game choice, stake size, number of paylines where selectable | RNG outcomes, hit frequency, jackpot contribution rates |
The game with the smallest built-in edge is usually the one where a player decision can change the expected loss on a hand. That points to blackjack played with a published basic-strategy chart matched to the table's specific rules. But the edge is still there. A perfect decision on every hand reduces the rate of loss; it does not reverse the sign of the expectation.
Why a short session looks nothing like the long run
Expected value is a long-run average. It says nothing reliable about the next hour. A game with a small house edge can still produce a large player win in a single session, and a game with a larger edge can grind a player down slowly. The reason is variance: how payouts are distributed over time. High volatility means rarer but larger payouts; low volatility means more frequent but smaller ones. Volatility does not change a game's RTP, but it changes what a session feels like and how often a player walks away ahead.
Suppose a player sits at a blackjack table and wagers 100 units across an evening. The expected loss on that action is a small fraction of the total wagered, but the actual result will fall somewhere on a wide range around that expectation. The player might finish 30 units up or 40 units down. Neither outcome tells you the house edge. The edge is only visible after enough decisions that the short-term swings wash out, and most players never reach that point in a single visit.
This is why a casino can pay out a large jackpot or a run of winning hands and still operate profitably. The payout is a single data point in a distribution that is already priced into the game. The house does not lose money on the game because one player won; it pays out a sum that the aggregate of all other wagers has already funded in expectation.
Decisions per hour change total exposure
House edge is expressed as a percentage of each wager, which means the total expected loss depends on how much is wagered in total. Total wager is a product of stake size and the number of decisions per hour. A game that resolves quickly exposes more money to the edge in the same period than a slower game at the same stake.
As an example, suppose a player wagers 5 units per decision. At a table game that resolves 60 decisions in an hour, the player puts 300 units into action. At a slower table resolving 30 decisions in an hour, the same stake puts 150 units into action. If both games carry the same house edge, the faster game produces twice the expected loss in the same sitting, even though the percentage on each bet is identical. Speed is not a separate disadvantage; it is a multiplier on the disadvantage that already exists.
Slots make this explicit. A slot player can spin hundreds of times in an hour at a stake that is small per spin but large in aggregate. The RTP is applied to the total coin-in, not to the starting balance. A game returning a high theoretical percentage still returns less than 100%, and the shortfall applies to every unit wagered, including units recycled from earlier wins.
Progression systems redistribute outcomes, not edge
The query about which game a casino loses the most money on sometimes comes from a different place: a belief that some game or betting pattern can force the house into a losing position. The most common version is a progression system, where the player increases the stake after a loss so that a single win recovers all previous losses plus a small profit. The arithmetic looks convincing on paper, which is why the idea persists.
What a progression does is reshape the distribution of outcomes. It creates many small wins and a small number of large losses. The expected loss across all sessions is unchanged because the house edge applies to every unit wagered, regardless of the sequence of stake sizes. A player using a progression is not removing the edge; they are concentrating it into the rare session where the losing streak runs long enough to exhaust the bankroll or hit the table maximum.
Suppose a player doubles a stake after each loss, starting at 1 unit, and the table allows a maximum stake of 64 units. The player can absorb six consecutive losses before the next required stake exceeds the limit. The probability of that losing streak is not zero, and when it arrives, the loss is large enough to erase all the small wins that came before. The system does not change the expectation; it changes the timing and the shape of the loss. No betting pattern changes the built-in edge of a negative-expectation game.
Where the actual figures live
Every number that would answer this question precisely for a specific venue or a specific game is published somewhere the reader can check. For a table game, the payout multiples are printed on the layout, and the rule variations that move the edge are posted at the table or available from the dealer. For a slot, the game information panel shows the configured RTP and the volatility class, where the jurisdiction requires disclosure. For an online game, the same information sits in the game's help or information screen, usually under a heading like paytable or game rules.
The reader does not need to memorise a list of percentages. The useful skill is reading the payout table and recognising the gap between true odds and paid odds. Where the gap is small, the game is less expensive to play over the long run. Where the gap is large, the game is more expensive. The operator's own published rules carry the figure that applies to the game in front of the reader, and that figure is the only one that matters for their own play.
FAQ
Do casinos ever lose money on a single game?
On a single night, yes. A table can pay out more than it takes in, and a progressive jackpot can exceed the coin-in from the machines feeding it. These are short-run variance events. Over a large number of wagers, the game's built-in edge means the aggregate result moves toward the house's expected share. A single losing night is a data point, not a trend.
Which casino game gives a player the best chance to leave ahead?
The game with the smallest house edge gives the best long-run expected result per unit wagered, but a short session is dominated by variance, not edge. A player can leave ahead on any game, including games with a large edge, because the edge is a long-run average. The question of which game gives the best chance to leave ahead on a given night is answered by volatility and session length, not by the edge alone.
Can a player reduce the house edge by changing how they bet?
On some games, yes. In blackjack, using a published basic-strategy chart matched to the table's specific rules reduces the expected loss per hand compared with arbitrary decisions. In craps, choosing bets with a smaller edge over bets with a larger edge changes the expected loss per unit wagered. In roulette, choosing a single-zero wheel over a double-zero wheel where both are offered changes the edge on every bet placed. None of these choices removes the edge; they change its size.
If you or someone you know is struggling with gambling, help is available. Contact local support services for confidential advice and resources.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.