Repeated wagers increase the amount exposed
A $100 deposit can support more than $100 in bets when payouts are used again. If total wagers reach $2,000 at 96% RTP, fixed-turnover expected loss is $80. Applying 4% only to the original deposit would miss most of the wagering.
The ability to complete a planned amount of play depends on the payouts and remaining funds. A calculation that assumes a fixed number of bets does not guarantee that the budget will last long enough to place them.
See the worked example showing why turnover can exceed a deposit.
Compare returns at the same volume
| RTP | House edge | Expected loss |
|---|---|---|
| 99% | 1% | $10 |
| 98% | 2% | $20 |
| 96% | 4% | $40 |
| 94% | 6% | $60 |
These are expected costs for a specified amount of play. They do not mean a player at 99% will necessarily lose less in one session than a player at 94%. Random outcomes can reverse that short-term ordering.
The long run is not a recovery mechanism
In an independent game, earlier losses do not improve later probabilities. A growing sample can make average results more representative under the mathematical model, but it does not require your personal cash loss to shrink. More bets can add further expected losses.
Dividing a budget by average expected loss per bet is not a reliable estimate of bankruptcy time. Different payout distributions and stopping rules can produce very different outcomes despite the same RTP.
Watch for the thinking trap where high RTP becomes a reason to keep spending.
Control the quantities you can choose
- Set a money limit before play; do not use expected return as a spending budget.
- Check the full cost of each selected mode.
- Consider the effect of bet size, number of bets and playing pace together.
- Stop at your limit rather than adding wagers to chase theoretical return.
Read common thinking traps if the idea of a game being due to repay you is influencing a decision.
Why equal averages can produce different bankroll paths
A session ends when your chosen stopping rule is reached or the funds needed for the next bet are unavailable. Those path-dependent events cannot be inferred from RTP alone. A distribution concentrated in small outcomes behaves differently from one whose mean includes rare prizes.
A game with the same mean and smaller payout variation may have a different chance of reaching a target number of rounds. A proper bankroll model needs the full distribution, full stake, available balance and stopping rule.
Recycled winnings are new exposure to the edge
Withdrawing a payout removes it from future game exposure. Betting it again creates another wager. Calling it “house money” does not change the expected cost of the next bet. Money already won and money deposited can fund the same wager under the same rules.
For a fixed total of $5,000 wagered at 98% RTP, expected loss is $100 whether the wagers were funded by one initial deposit or by several deposits and recycled payouts. Actual funding paths affect whether that total can be reached, but they do not turn the wager model into a 2% charge on deposits.
This is why a measure based only on deposits cannot be described as game RTP. Keep transaction and game ledgers separate, and use the weighted calculation when combining different stakes or returns.
Can stopping after a win remove expected losses?
A stop rule can limit exposure or define when you end a session. It does not improve the probability of the next independent round. A plan to stop at a small profit can produce many small successful sessions and occasional larger losses; the proportion of successful sessions alone is not enough to judge expected value.
For example, nine sessions winning $10 each and one losing $150 have an 90% session success rate but an overall $60 loss. This is a bookkeeping example rather than a simulation of a particular strategy. To evaluate a real rule, every possible loss path and the stake required along the way must be included.
Limits are useful as spending controls. They should not be sold as a way to turn a negative-expectation random game into income. Read the progression example before interpreting a high success rate as an advantage.
Reference: MGA: player protection and dispute resolution. The calculations in this section are illustrative unless a named product is explicitly identified.
Common questions
Will playing long enough make me reach the advertised RTP?
There is no guaranteed personal endpoint. Theoretical RTP describes expectation, and continuing to wager on a positive-edge game adds expected cost.
See the explanation: The long run is not a recovery mechanism →
Explore the explanation: Why chasing the advertised return can increase spending →
Can I calculate exactly how long my bankroll will last?
Not from RTP alone. You would need the payout distribution, stakes, available funds and stopping rules, and the result would still describe probabilities rather than certainty.
See the explanation: Control the quantities you can choose →
Is playing with winnings different from playing with deposited money?
For the same wager and rules, the funding source does not change expected return. Re-betting a payout creates new exposure to the house edge.
See the explanation: Recycled winnings are new exposure to the edge →
Does winning most sessions prove a betting strategy works?
No. Infrequent large losses can outweigh frequent small wins. Expected value must include the complete distribution of outcomes.
See the explanation: Can stopping after a win remove expected losses? →
Sources & further reading
Primary sources checked on 17 November 2026. The examples above use the stated rules and assumptions. Product pages describe published versions; they do not verify every operator’s configuration. Regulatory guidance applies within its stated scope.