Odds

What Is RTP? House Edge Explained Without the Marketing

WagerTruth Team August 15, 2026 10 min read 0 views

RTP stands for return to player. It is the percentage of all wagered money a game pays back to players over the long run. If a game has a 97% RTP, then for every $100 wagered across all players over time, the game returns about $97 in winnings and keeps $3. That kept portion is the house edge, and the two numbers are just two views of the same thing: house edge = 100% - RTP. A 97% RTP game has a 3% house edge. That is the entire relationship, and everything else on this page follows from it.

Casinos and case sites advertise RTP because the bigger number sounds good. Players should learn to read the smaller one instead, because the house edge is what actually describes what a game costs you.

What RTP actually means (and what it does not)

The most misunderstood word in the definition is "over the long run." RTP is a statistical average measured across millions of rounds. It says nothing reliable about your session tonight.

Say you bet $10 on a 97% RTP game. The expected return of that single bet is $9.70, but you will never actually receive $9.70. You will win $20, or lose the $10, or hit some other outcome the game allows. The 97% only emerges when you average an enormous number of bets. In the short run, variance runs the show: you can double your money in ten minutes on a terrible game, or go broke on a generous one. Both are normal.

So treat RTP as a measure of how fast a game drains you on average, not as a prediction. A 97% RTP does not mean you personally get 97% back. It means the game is built so that, over enough play, players as a group get 97% back and the house keeps 3%.

House edge is the business model

Every gambling site you have ever seen, from a Vegas casino to a CS2 case site, runs on the same engine: the payouts are set slightly below the true odds, and the gap is the profit. If a fair coin flip pays 2x, the house version pays 1.98x. If a fair roulette number pays 36 to 1, the house version still pays 35 to 1 while the wheel has 37 pockets. The games feel fair because wins happen constantly. The edge is invisible in any single round, which is exactly why it works.

The practical way to think about house edge is as a toll on every bet. Bet $10 on a 3% edge game and the expected cost is 30 cents, win or lose. The toll is charged on every bet, not on your deposit, which is why it compounds so viciously. Wager the same $100 ten times over an evening (easy to do) and the expected cost of a 3% game is about $30, even if your balance barely moved the whole night.

Calculator: what does a game actually cost you?

Expected loss = bet size x number of bets x house edge. Play with the numbers, then notice how the number of bets does the damage.

Total wagered

Expected return

Expected loss

Typical RTP ranges by game

Not all games are priced the same. Here are the ranges you will realistically encounter, from the friendliest to the most expensive. These are industry-typical figures; the exact number for a specific game should come from that site's own documentation.

  • Blackjack with correct strategy: ~99.5% RTP. About the best deal in any casino, which is why so few skin sites offer it. Play it badly and the edge climbs fast.
  • Baccarat (banker bet): ~98.9% RTP. Boring game, great price.
  • Dice on crypto casinos: ~99% RTP. The classic 1% edge setup most crypto sites use.
  • Crash and Mines: usually 97% to 99% RTP. Depends on the site; check the fairness or game info page.
  • European roulette: 97.3% RTP. One zero. American roulette with the extra double zero drops to 94.74%, which is a massive difference for one extra pocket. Never play American if European exists.
  • Online slots: typically 92% to 98%. The same slot title can ship in multiple RTP configurations, and the casino picks which one to run. Two sites can offer the identical game at different prices.
  • Coinflip and jackpot on skin sites: usually 90% to 95% effective RTP. These take a rake on the pot, often 5% to 10%, which is the edge in disguise.
  • Case openings: often 40% to 80% RTP. Yes, really. More on this below, because it is the worst mainstream deal in the entire space.

A rough rule of thumb: anything at 97% or above is priced like a serious casino game. Anything below 95% is expensive entertainment. And anything where the site will not tell you the number at all should be assumed to be terrible until proven otherwise.

Why case openings are the worst deal in the room

This is the part case sites do not put on the homepage. When you open a $2 case, the expected value of the items inside is usually well under $2. Across the case sites we have looked at, it is common for a case to return 60 to 80 cents on the dollar, and plenty of high-volatility cases sit far lower. A 60% RTP would get a slot machine laughed out of a licensed casino. On a case site, it is just Tuesday.

The saving grace is that you can check it yourself when a site publishes item odds. The math is one line: multiply each possible item's value by its probability, add those up, and divide by the case price. That ratio is the case's true RTP.

Worked example

A case costs $1.00 and contains only three possible items:

  • Common skin worth $0.20 at 80% chance → contributes $0.16
  • Mid skin worth $1.50 at 18% chance → contributes $0.27
  • Knife worth $40 at 2% chance → contributes $0.80

Expected value: 0.16 + 0.27 + 0.80 = $1.23 against a $1.00 price. That case would be a 123% RTP, meaning it is priced in the player's favor, and in practice it would be pulled or repriced within days. Real cases almost always sum to less than the price. Run the same math on a real case's odds table and the gap between the sum and the price is the house edge, stated in dollars.

Sites that publish per-item odds (the better ones do, often committed through their provably fair system) are giving you everything needed to compute this. Sites that hide the odds are telling you the number is ugly. If you open cases at all, spend five minutes with the odds table first. It is the single most informative thing on the site.

Variance: why the average takes so long to show up

Here is the uncomfortable follow-up. Even on a fairly priced game, your results will swing wildly around the expectation, sometimes for thousands of bets. That is variance, and it is the reason gambling feels winnable in the short term while being a guaranteed cost in the long term.

The demo below simulates sessions of even-money bets on a 97% RTP game (48.5% chance to double your bet, 51.5% chance to lose it). Run it a few times and watch how far individual sessions land from the expected result, even though every bet is priced exactly the same:

Simulator: 10 sessions on a 97% RTP game

Each session is 200 bets of $1 at even money with a 48.5% win chance. Expected result per session: -$6. Watch how rarely reality lands on it.

Two lessons hide in that simulator. First, some sessions finish in profit even though the game is priced against you, and those are the sessions people remember and talk about. Second, the average of all sessions drifts toward the expected loss, and the more sessions you run, the closer it sticks. The house does not need you to lose every time. It just needs you to keep playing.

High-volatility games stretch this effect further. A case with a 2% knife chance and a 90% RTP can still eat your whole balance for hours, because so much of the payback is locked inside a rare outcome. Two games with identical RTP can feel completely different: one drips small wins constantly, the other starves you until a big hit lands. Same price, different ride. Know which one you are strapped into.

How to actually use RTP as a player

RTP will not help you win. Nothing will, because the edge is structural. What it does help you do is lose slower and avoid the worst products. Concretely:

  1. Compare games before you play them. A 99% dice game and a 60% case cost the same per bet and return wildly different amounts. If you enjoy both equally, play the better-priced one.
  2. Check the odds table on every case. Five minutes of arithmetic beats an hour of regret. If a site publishes committed odds, use them. If it does not, assume the worst.
  3. Think in expected cost, not in wins. The calculator above turns any game into a dollar-per-hour price tag. That framing kills most of the marketing magic.
  4. Treat big RTP claims with suspicion. "Up to 99% RTP" usually means one obscure bet configuration pays 99% while the ones people actually use pay far less. Read which bet the number refers to.
  5. Remember that RTP and fairness are different questions. A game can be perfectly provably fair and still be a terrible deal. Verify the randomness, then separately check whether the price is worth paying.

The bottom line

RTP and house edge are the same number wearing two hats: one flatters the game, the other tells you what it costs. Learn to convert between them instantly, learn the rough ranges for the games you play, and learn to compute case value from an odds table. None of that makes gambling profitable, because it is not. But it makes you a customer who knows the price of the product, and in this industry that alone puts you ahead of most of the room. If the cost stops feeling like entertainment money, our responsible gaming page has the resources worth reading.

← Back to the blog