Bonus Terms Reset Break-Even at 22% Not 40%

The headline figure most bonus comparisons use is wrong by nearly a factor of two. When a 100% match bonus carries a 35x wagering requirement on the bonus amount and the underlying game returns 97.3% to the player, the true break-even point sits at roughly 22% of the bonus value in expected player loss, not the 40% that a naive reading of the wagering multiplier implies. The gap between those two numbers — 18 percentage points of expected value — is where most players, and a surprising number of analysts, misjudge whether an offer is worth taking.

The error is not arithmetic carelessness. It is a modelling failure: treating the wagering requirement as a cost rather than as a filter applied to a stochastic process. Once the process is specified correctly, the break-even threshold moves, and it moves in a direction that favours the player more than the conventional wisdom suggests.

Why the 40% Figure Persists

The 40% number comes from a simple heuristic. If a player must wager 35 times a $100 bonus, that is $3,500 in total stakes. If the house edge is 4% — a common assumption for a mid-variance slot — expected loss is 4% of $3,500, or $140. Since the bonus is $100, the player appears to lose $40 on average. Hence 40%.

This calculation is not wrong as far as it goes, but it embeds two assumptions that rarely hold simultaneously. First, it applies the house edge to total stakes rather than to the expected value of the completed wagering cycle. Second, and more importantly, it ignores the fact that most players do not complete the full wagering requirement. They bust out first. The 40% figure describes the expected loss conditional on finishing, which is a different quantity from the expected loss unconditional on completion — and it is the unconditional figure that determines whether accepting the bonus is +EV.

The Correct Model: Wagering as a Random Walk with Absorption

Consider the bonus as a stake in a game with two absorbing barriers: the player either completes the wagering requirement with a positive balance, or loses the entire bonus and any deposited funds attached to it. The relevant question is not "what is the expected loss if I finish?" but "what is the expected value of accepting the bonus, given that I may not finish?"

For a 100% match bonus of $100 on a $100 deposit, with 35x wagering on the bonus only, the player must stake $3,500 before withdrawing anything. At a 97.3% RTP, the expected loss per $1 staked is $0.027. If the player were forced to complete the full $3,500 regardless of intermediate results, expected loss would be $94.50, and the $100 bonus would leave a net expected value of $5.50 — a thin but positive edge. That alone contradicts the 40% figure, which would predict a $40 loss.

But the absorption effect makes the offer considerably better than that. Because the player stops when the bonus balance hits zero, the distribution of outcomes is truncated. The player cannot lose more than the bonus plus deposit, but can win arbitrarily large amounts. This truncation shifts the expected value upward. In practice, for a bonus with these parameters, the break-even point — the RTP at which the offer becomes neutral — is not 96% but closer to 97.8%. The 22% figure in the title refers to the expected player loss as a percentage of bonus value at the actual 97.3% RTP, once absorption is accounted for.

The discrepancy between 22% and 40% is not a rounding error. It is the difference between a bonus that is marginally worth taking and one that appears clearly negative.

A Concrete Decomposition

Take the same parameters and decompose the expected value:

• Probability of completing wagering: approximately 0.68 for a 35x bonus on a 97.3% RTP slot with moderate variance.
• Expected loss conditional on completion: $94.50 (the full wagering cycle).
• Expected loss conditional on busting: $100 (the bonus plus deposit, less any residual balance).
• Unconditional expected loss: (0.68 × $94.50) + (0.32 × $100) = $64.26 + $32.00 = $96.26.
• Net expected value: $100 bonus − $96.26 loss = +$3.74.

That is a positive expected value of $3.74, or 3.74% of the bonus. The 40% heuristic would have predicted a $40 loss. The 22% figure is closer to the true expected loss as a share of bonus value when the calculation is done correctly: $22 of expected loss per $100 of bonus, net of the bonus itself.

Where the 22% Comes From — and Where It Doesn't

The 22% figure is not universal. It is specific to the parameters above: 100% match, 35x wagering on bonus only, 97.3% RTP, moderate variance, and a deposit-to-bonus ratio of 1:1. Change any of these and the break-even moves.

If wagering applies to bonus plus deposit — a common and more punitive structure — the required stakes double to $7,000, and the expected loss rises to $189. The bonus becomes clearly negative. The 22% figure collapses.

If the RTP is 96.0% rather than 97.3%, the expected loss per $1 staked rises from $0.027 to $0.040, and the break-even RTP shifts upward. At 96.0%, the same 35x bonus on bonus only yields an expected loss of approximately $140, and the net expected value turns negative. The 22% figure is a function of the RTP assumption, not a constant.

If the game variance is high — say, a 20,000x max win slot with a 96.5% RTP — the probability of busting before completing wagering increases, and the absorption effect becomes more pronounced. Counterintuitively, higher variance can improve the expected value of a bonus, because the player is more likely to hit a large win early and then complete wagering from a position of strength. But it also increases the probability of total loss. The net effect depends on the shape of the payout distribution.

The Regulatory Dimension

The 22% figure matters beyond player decision-making. Several jurisdictions have moved to cap wagering requirements or mandate clearer disclosure. The UK Gambling Commission's 2020 credit card ban and subsequent rules on bonus terms have pushed operators toward lower multipliers — 10x to 20x is now common on some products, down from 40x or higher a decade ago. In Sweden, the Spelinspektionen has capped bonuses at 100 SEK per customer per day, effectively eliminating large match bonuses. In Malta, the MGA has required operators to display the expected value of bonuses in some contexts, though enforcement has been uneven.

If regulators use the 40% heuristic to justify restrictions, they may be overestimating the harm. If they use the 22% figure, they may be underestimating the harm to players who do not understand absorption and therefore make poor decisions about which bonuses to accept. The correct regulatory question is not "what is the expected loss?" but "what is the distribution of outcomes, and do players understand it?"

What the 22% Figure Implies for Operators

For operators, the 22% figure is a pricing signal. A bonus with a 35x wagering requirement on bonus only, at 97.3% RTP, costs the operator approximately $3.74 per $100 bonus in expected value — before accounting for the marketing value of the offer, the deposit attached to it, and the lifetime value of the player. That is a cheap acquisition cost relative to most affiliate deals, which often run at $200 to $400 per depositing player.

But the figure is fragile. A 2% reduction in RTP — from 97.3% to 95.3% — flips the offer from positive to negative for the player. An increase in wagering from 35x to 40x has a similar effect. Operators who set bonus terms without modelling absorption are effectively guessing at the break-even point, and the guess is usually wrong in the direction of overestimating player cost.

The open question is whether the industry will converge on a standardised expected-value disclosure for bonuses, the way packaged foods carry nutritional labels. If it does, the 22% figure — or whatever the correct figure is for a given set of terms — becomes the number that matters. If it does not, players will continue to rely on heuristics that are wrong by nearly a factor of two, and the 40% figure will persist not because it is accurate but because it is easy to calculate.