Loyalty Points Accrue 2.4x Faster on Losing Spins Than Winning

Across a sample of 14 mid-tier and large European-facing casino loyalty programmes, points accrued on losing spins at an average rate 2.4 times that of winning spins, after normalising for stake size. The ratio was not a design quirk of one operator: it held in 11 of the 14 programmes, ranging from 1.6x to 4.1x. The figure emerged from a six-month audit of published programme terms, sample account statements, and operator-side point-ledger documentation supplied under disclosure arrangements with three regulators between January and June 2024.

What makes the finding worth attention is not that losing play earns more — most players intuit that — but the size and consistency of the gap, and what it implies about how loyalty schemes are priced. A 2.4x differential is not a rounding artefact. It is a structural feature, and it recasts the loyalty point as something closer to a rebate instrument than a reward currency.

How the 2.4x figure was constructed

The audit compared points earned per €1 of stake across two conditions: spins that returned less than the stake (net loss), and spins that returned more (net win). To isolate the loyalty mechanic from game-level variance, the analysis held game type, bet size, and session length constant, and used operator-side ledger data rather than player-reported figures, which are unreliable for this purpose.

Three methodological notes matter for interpreting the ratio.

First, "losing spin" here means a single round outcome, not a session. Session-level accounting produces a different and generally smaller ratio, because wins and losses interleave within a session and most programmes aggregate at session or daily level.

Second, the 2.4x average conceals substantial spread. The programme with the 4.1x ratio used a tiered structure in which base points accrued only on net losses above a rolling threshold, effectively zeroing out winning play for most players. The programme with the 1.6x ratio awarded a flat 1 point per €10 staked regardless of outcome, but applied a 0.5x multiplier to wins above a certain size — a narrower differential by design.

Third, the ratio is stake-normalised. Raw point totals would show an even larger gap, because losing players tend to stake more per spin and play longer sessions. Normalising for stake removes that behavioural effect and leaves the mechanical one.

Why winning play is discounted

Operators gave broadly consistent explanations when asked. The most common was margin protection: loyalty points carry a real cost (typically 0.1% to 0.4% of handle, depending on programme), and awarding them symmetrically on wins and losses would compound the cost during high-variance periods. One compliance officer at a Maltese-licensed operator described it as "paying for engagement, not for outcomes" — a framing that is defensible in isolation but sits awkwardly against the 2.4x figure.

A second explanation was behavioural. Several operators argued that winning spins already generate their own reinforcement, so loyalty points are better deployed where they can influence retention — namely, on losing streaks. This is a retention argument dressed as a reward argument, and it is the more interesting of the two for regulatory purposes.

The regulatory grey zone

Loyalty schemes occupy an unusual position in most licensing frameworks. They are typically classified as promotional tools rather than gambling products, which means they fall outside the strictures that apply to the underlying games. That classification is increasingly contested.

In 2023, the UK Gambling Commission signalled in a consultation response that loyalty schemes would be examined under the same lens as other inducements, with particular attention to whether they function as a form of loss-chasing incentive. The Commission stopped short of mandating outcome-neutral accrual, but the direction of travel is clear. A 2.4x differential is exactly the kind of figure that invites scrutiny under a "does this encourage continued play after losses" test.

The practical question for regulators is whether a differential is inherently problematic or only problematic above some threshold. No jurisdiction has published a threshold. Sweden's Spelinspektionen has taken a harder line on bonus mechanics generally but has not addressed loyalty accrual specifically. Malta has been silent. Gibraltar has issued informal guidance suggesting operators document the rationale for any outcome-based multiplier, which is a disclosure requirement rather than a prohibition.

For operators, this creates an asymmetry: the mechanic is legal in most markets, but the documentation burden and reputational exposure vary widely. A programme with a 1.6x differential and a clear published rationale is in a materially different position from one with a 4.1x differential and no disclosure.

What the differential actually costs the player

The intuitive reading — that losing players earn more points, so loyalty schemes favour losers — is correct but incomplete. The more precise framing is that the differential transfers value from winning players to losing players, funded by the house edge.

Consider a simplified model. A player staking €100,000 over a year at a 96% RTP game expects to lose €4,000. If the programme awards points worth 0.2% of handle on losing play and 0.08% on winning play (a 2.5x differential), the expected loyalty value depends on the distribution of wins and losses, not just the net result. Because wins are less frequent but larger, the stake-weighted accrual skews toward the losing side. In practice, the player's expected loyalty return lands somewhere between 0.13% and 0.17% of handle — below the headline 0.2% and well below what a flat programme would deliver on the same handle.

This is not necessarily unfair. It is, however, opaque. Most programme terms describe the differential in language that few players parse correctly, and none of the 14 programmes audited disclosed the expected loyalty return as a percentage of handle in a way that would allow direct comparison.

The comparison problem

Cross-operator comparison is close to impossible for a player. Two programmes can both advertise "1 point per €10" and deliver materially different value depending on outcome weighting, point redemption rates, expiry rules, and tier multipliers. The 2.4x average differential is one input among many, but it is the one least visible in marketing material.

Third-party comparison sites have not solved this. Most compare headline earn rates and redemption catalogues, not outcome-weighted accrual. A player choosing between two programmes on the basis of published terms is, in effect, choosing on incomplete information.

Where the mechanic goes next

The 2.4x figure is a snapshot, not a constant. Three pressures are likely to move it.

Regulatory attention is the most immediate. If any major jurisdiction mandates outcome-neutral accrual — or, more likely, requires disclosure of the differential — programmes with high ratios will face a choice between restructuring and accepting a competitive disadvantage against flat-accrual operators.

Competitive pressure runs the other way. Outcome-weighted accrual is more efficient from the operator's perspective: it directs reward spend toward players most likely to churn. A flat programme is simpler to market but less targeted. The 11 of 14 programmes with a differential above 1.5x suggest the market has already voted.

Player awareness is the wildcard. If the 2.4x figure becomes widely known, the marketing advantage of a high headline earn rate erodes, and operators may shift toward transparency as a differentiator. That has not happened yet in any market the audit covered.

The open question is not whether loyalty points should accrue equally on wins and losses — there are coherent arguments on both sides — but whether a 2.4x differential can be justified to a player who understands it. Operators have so far been asked that question only by regulators, and only in private. The next phase of the debate will require them to answer it in public, and the answer will determine whether the loyalty point survives as a reward or is reclassified, formally, as a rebate.