What a 96% RTP does and does not mean in one session

A 96% RTP means that, across a very large number of rounds, a game returns on average KES 96 of every KES 100 staked and keeps KES 4. It is not a prediction for one session. In a short session the result is usually far from 96%, and over many rounds the cost still falls on every shilling wagered.

By Nairobi Ultimate Editorial (prepared with AI tools under our editorial policy) · Last updated 2026-10-14

The short answer

RTP, return to player, is a property of the game's rules. It is worked out by multiplying every possible payout by its probability and adding the results, as described in the guide to how RTP and house edge work. A 96% figure means the remaining 4% is the house edge.

The figure says what happens on average when a huge number of rounds is pooled. It does not say what happens to one person on one evening.

What it does mean

For every KES 100 wagered, the average return is KES 96 and the average cost is KES 4. The cost scales with the amount wagered, not with the amount deposited.

expected loss = stake per round × number of rounds × (1 − RTP)

This is the same idea as expected value: a long-run average per unit staked, here equal to minus 4 cents per shilling.

What it does not mean

• It is not a share of your deposit. Nothing guarantees that KES 96 of every KES 100 you bring comes back.
• It is not a schedule. The game does not track your session and adjust payouts to reach 96%. Each round of a game of chance is independent of the last.
• It is not a measure of how often you win. A win rate is a separate number.
• It is not a measure of risk. Spread of results is a different property, covered under volatility.

One session, worked in full

To see how far one session can sit from 96%, take a deliberately simple game. Each round costs KES 100. With probability 0.4 the round returns KES 240 (a 2.4 times payout), otherwise it returns nothing. The RTP is 0.4 × 240 / 100 = 0.96, so exactly 96%.

Now play five rounds, staking KES 500 in total. If k is the number of winning rounds, the return is 240 × k and the net result is 240 × k minus 500. The probability of k wins is the number of ways to choose which rounds win, times 0.4 to the power k, times 0.6 to the power 5 minus k.

Three things stand out. First, no row returns exactly KES 480, the 96% figure: the average appears only when the rows are weighted by their probabilities. Second, the player finishes ahead when there are three or more wins, with probability 23.04% + 7.68% + 1.024% = 31.744%. Third, the player loses the whole stake in 7.776% of sessions.

The weighted average of the net column is a check on the arithmetic: -38.88 - 67.392 - 6.912 + 50.688 + 35.328 + 7.168 = -20. That is KES 20, which is 4% of KES 500, exactly as the formula gives.

Why a good session proves nothing

The 31.744% of winning sessions are real, and they are the reason a short run can feel like evidence that a game pays well. It is not. Each such session sits in the same distribution as the losing ones, and the average across all of them is still a loss of 4% of the amount wagered.

The same holds in reverse. Five losses in a row, which happens in 7.776% of sessions in this example, does not make the next round more likely to win. The game is not due, as explained in the guide to the gambler's fallacy and session limits.

What changes as sessions get longer

The expected loss grows in proportion to the number of rounds: 5 rounds at KES 100 cost KES 20 on average, and 500 rounds cost KES 2,000. The scatter does not vanish quickly, but relative to the total wagered it narrows, so the average result of many sessions settles near minus 4%.

That is why a platform with thousands of players sees something very close to the published percentage, while one player sees a lumpy, unpredictable path. The player is on the short-run side of the same arithmetic.

Reading an RTP figure for your own session

1. Find the rules or information screen of the game and check whether the figure is labelled theoretical.
2. Write down the stake per round and the number of rounds you expect to play in the time you have set.
3. Multiply the two, then multiply by one minus the RTP. The result is the average price of that session.
4. Decide on a limit you can afford to lose entirely, because the table above shows that a full loss is an ordinary outcome.

Speed deserves attention here. The number of rounds per hour, not only the percentage, sets the bill. A game that settles a round every few seconds reaches a large total wagered quickly.

The percentage also does not describe how often a round pays, which is a different measure called hit frequency.

What Kenyan rules say

The Conduct of Gambling Operations Regulations, 2026 provide in regulation 20(2) that a licensee "shall display clearly rules of games, odds, house edge and average return to the player". Regulation 20(3) applies the provision to online gambling activities, with the necessary modification where applicable.

Regulation 45(2) requires that the random number generator used for games of chance be certified to statistically test the randomness and the theoretical return to player. Regulation 45 also says the theoretical return to player percentage for each game is to be publicly disclosed in the licensee's online platform or website. Testing is covered in the guide to RNG testing and certification in Kenya.

Gambling is addictive. A published RTP below 100% describes a cost, not an income, and it is sensible to play responsibly and only with money set aside for entertainment.