The Crypto Cashback Index — Q3 2026

Every partner rate on WagerSave, tracked quarterly, with median and range by vertical. Q3 2026 covers 35 partners across crypto casino, CS2 case sites and mystery boxes, collected TBC. Methodology and raw data below.

Quarter-on-quarter change: TBC.

Publish this section in full. It's what makes the Index citable rather than marketing.

Collection. Rates taken from live partner terms as of TBC. Where a rate is tiered, the entry-tier rate is recorded, with the top tier noted separately. Where a rate is negotiated and non-public, the partner is excluded rather than estimated.

Normalisation. Partners quote in two incompatible units — percentage of theoretical loss and percentage of turnover. Every rate is converted to return per 1,000 wagered using:

Per 1,000 = 1,000 × house edge × rate     (theoretical-loss quoting)
Per 1,000 = 1,000 × rate                   (turnover quoting)

House edge assumptions. Conversion requires an assumed edge. The Index uses a fixed reference figure per vertical so quarters remain comparable:

These are reference figures for normalisation, not claims about any specific game. Actual edges vary widely within each vertical.

Weighting. Rates are recorded at 100% game weighting. Where a partner applies category weighting, the weighted rate for slots is used and the weighting table is noted in the per-partner data.

Exclusions. Partners are excluded where the rate is non-public, where the programme launched within the quarter (insufficient history), or where the payment form is site credit rather than withdrawable value. Excluded partners are listed with the reason.

Median, not mean. Rate distributions are skewed by a small number of outliers at the top of each vertical. The median describes the typical partner; the mean doesn't.

Download: crypto-cashback-index-q3-2026.csv

Crypto casino

21 partners. Median 10%, range 10%–30%.

Rates in this vertical cluster tightly. The affiliate market is mature, commission terms are broadly standardised at 30–50% of NGR, and competitive pressure has compressed the spread. A partner materially above the range is usually quoting in a different unit or capping somewhere unstated.

TBC — 2–3 sentences on what moved this quarter and why.

CS2 case sites

10 partners. Median 10%, range 2%–40%.

The highest rates in the Index, for two structural reasons: house edges of 8–20% against a casino's 4%, and affiliate commission above casino norms in a vertical competing hard for a small, well-informed audience.

TBC

Mystery boxes

4 partners. Median 11%, range 10%–15%.

The widest internal spread in the Index. Digital-prize platforms carry near-zero fulfilment cost and can pay several times what physical-goods platforms can, so the vertical doesn't cluster the way casino does. Rates here require checking individually rather than assuming a norm.

TBC

→Why mystery box cashback rates are lower.

Notable individual moves:

• TBC
• TBC
• TBC

→Largest increase this quarter: Datdrop rate rises to 40% for all members.

Stating the limitations is what separates a dataset from an advertisement, and it's what makes the Index worth citing.

Payment reliability. A rate is a stated intention. Whether an operator pays consistently at volume isn't in this data.

Effective rate after weighting. Rates are recorded at slots weighting. A live-casino or table-games player will realise materially less at most partners.

Caps. Daily, weekly and monthly maximums aren't normalised into the per-1,000 figure. A high rate with a low cap can be worth less than a lower uncapped rate at volume.

Negotiated rates. Non-public terms exist and are excluded. The Index describes advertised rates only.

House edge variance. The reference edges used for normalisation are vertical midpoints. A player on 99.5% blackjack or a 23% case will diverge substantially from the per-1,000 figures shown.

Free to reference with attribution. Suggested form:

Raw CSV available above. If you need the underlying per-partner collection notes for research, TBC.