What RTP actually means over 1,000 bets
Expected return, actual return and the variance between them — with a simple worked example.
01RTP is an expected return
Return to player describes the expected gross payout as a share of the money staked under a game’s rules. A 99% RTP means an expected $0.99 returned per $1 wagered, including any returned stake.
It does not mean a 99% chance of winning, that 99% of players profit, or that your balance will settle at 99% of its starting value.
02The arithmetic for 1,000 bets
Here is a hypothetical example with a constant $1 stake and a 99% RTP. These are illustrative assumptions, not results from a casino or a promise about a session.
| Total amount wagered | 1,000 × $1 = $1,000 |
|---|---|
| Expected gross payout | $1,000 × 99% = $990 |
| Expected net result | $990 − $1,000 = −$10 |
The $1,000 is turnover: the sum of all stakes. Money won and wagered again counts again. It need not equal the original deposit.
The calculation assumes all 1,000 bets are placed at that stake under the same rules. An actual player can run out of funds earlier. Expected values do not provide the bankroll needed to complete a session.
031,000 bets is not a guarantee
Games with the same RTP can distribute payouts very differently. Frequent small returns and rare large returns can have the same mathematical average, but very different session results.
There is no universal bet count at which a player must receive the advertised RTP. The payout distribution determines how much results vary; rare outcomes can remain absent over many bets. Past losses do not make an independent future win due.
For an example using a documented game and reproducible replay, read our Duel Plinko variance article. The underlying audit supplies the game-specific RTP and paytables.
04Measure what actually happened
Observed RTP = total gross payouts
÷ total amount wagered × 100%Use all bets in the chosen period and count payouts consistently, including returned stakes. If $1,000 was wagered and $850 paid back, observed RTP is 85%; the net result is −$150. This is another arithmetic example, not measured data.
A low observed return alone does not show that an outcome was manipulated. Provably fair verification checks whether outcomes reproduce from the recorded inputs; RTP analysis checks the distribution and payout math. Both matter, and they answer different questions.