Let's Talk About Duel

Duel.com went live on 18 July 2025. It was founded by Ossi Ketola, a name most players probably already know. Ketola operates under the handle "Monarch", and before Duel he built CSGOEmpire, one of the best-known CS:GO skin-gambling sites of the last decade. Duel is his move from skins into a full crypto casino, and it arrived with an unusual approach.

Duel claims to be the first casino whose house games run at a 0% RTP house edge, meaning that on its own original games, the house holds no mathematical advantage over the player. In an industry whose entire foundation is the edge, that is close to heresy. So before anything else, it is worth being precise about how Duel actually works, because those mechanics are the model:

• 0 % edge on the first $50,000 wagered per day. Every original game carries a 0.1 % base edge, credited straight back to the player. That 0 % treatment caps out at $50,000 wagered per day - once a player crosses that line, the 0.1 % edge starts to apply.
• 50 % rakeback on the casino category, but the edge stays. On slots and live tables (Blackjack, Roulette etc.) Duel pays 50 % instant rakeback - but those games keep the provider's native edge, so the net house edge on casino is reduced, not removed.
• Sportsbook keeps a real edge, just a lower one. Sports betting carries the bookmaker's usual vig - Duel's is around 2.9 % effective, well below industry-typical but structurally a house-edge product, not a 0 % one.
• Leaderboards top up player PnL; affiliates are paid separately. Players earn from leaderboards and race prizes (paid into their PnL) and a very strong wager-share affiliate offer (paid outside the casino, so not visible in player-level data).

That is the offer. The question is whether it can possibly be a business.

For as long as casinos have existed they have been one of the safest business models ever devised, for one reason: the math always works out. Hold a small fraction of every wager, push enough volume through the building, and revenue stops being a gamble and becomes arithmetic. Duel says it has thrown that arithmetic away. So the question this report is built around is blunt: does a 0 % house edge actually work - or is it just a marketing gag?

There is a sharper version of the same question. What if a casino doesn't need the mathematical edge at all? What if it can profit on human greed alone - on the simple, reliable fact that players always want more, never stop at the right time, and never play perfectly?

To find out, we did not take Duel's word for anything. We gathered every piece of publicly available information Duel exposes - its public transparency pages, individual player profiles, leaderboards - reconstructed the casino from the outside, and then reached out to Duel directly to confirm our reconstruction against their own database. What follows is a deep dive into Duel.com across its first 287 days, from launch to 1 May 2026.

The number everyone wants first

To not spoil too much - but also not to bury it - let's start with the number everyone actually wants: how much did Duel make?

Over the 287 days, players wagered a total of $11.47 billion. That is our figure, rebuilt from public data. Duel's own database returned $11.65 billion - a 1.5 % gap, which is roughly the slice of activity we missed by not capturing exactly at 00:00 UTC on 1 May.

Across all games, Duel's gross gaming revenue (GGR) - the house win before any rewards are returned - came to $135.1 million. But GGR is not what a casino keeps. Net gaming revenue (NGR) is.

Duel reports an NGR of $44.89 million. The method is worth a sentence, because it is the kind of thing only a transparency-first casino makes possible: on duel.com, the PnL figure shown in a player's profile popup is effectively that player's contribution to NGR. Summed across the entire player base, those PnL figures reconstruct the casino's net result from the bottom up.

The two numbers agree because they describe the same result from opposite directions. Duel builds NGR top-down, by subtraction:

Of $135 million in gross revenue, Duel hands $66 million straight back to players as rakeback and pays out a further $24 million in leaderboard and race prizes. Nearly two-thirds of the gross is returned before the house keeps a cent. What remains - $44.89M on $11.65B wagered - is an effective edge of 0.385 %. Strip out affiliate commissions too, and the house's genuinely retained margin falls to 0.310 %.

Deposits, withdrawals and the money that moved

Net deposit flow (NDW) - total deposits minus total withdrawals.

Duel reports $1,017.6 million in deposits and $982.9 million in withdrawals - an NDW of $34.7 million. On-chain: $901.3 million in, $874.95 million out, NDW $26.36 million.

That on-chain figure is deliberately conservative. Our on-chain tracking covers the major chains only - ERC-20, TRC-20, BSC and Solana. Deposits and withdrawals on other chains are not captured, so the on-chain numbers are a floor.

$11.5 billion was wagered, but only $34.7 million in net new money actually entered the casino. Every net dollar that came in was wagered roughly 330 times over.

And the money is not spread evenly. Plot every player by their deposit or withdrawal size and the picture is the same on both sides: a long flat tail of small accounts, and a handful of large accounts at the top doing most of the actual money movement.

Game by game - where the edge actually hides

A blended 0.385 % edge is an average, and averages conceal. Break Duel's $11.5 billion of turnover down by game and the picture is far less uniform - and, frankly, far more puzzling.

Here is where the edge actually comes from. The 0 % treatment caps at $50,000 wagered per day. Below that line the 0.1 % base edge is rebated in full; above it, the 0.1 % applies. Across a player base where 80 accounts move 65 % of all wager, most volume sits well past the cap - so the originals show a real, measurable effective edge.

Strip the table down to its mechanics: originals carry the 0.1 % base edge above the $50k daily cap, the casino category carries provider edges minus 50 % rakeback, sports carries the vig. The 0 % claim is a real promise on the first $50k of any player's daily wager - not a property of the platform as a whole.

Per-game breakdown

One card per game, in descending order of total wager. Each chart is preceded by a short note on what's actually driving that game's number.

On slots and live tables Duel pays 50 % instant rakeback, but the underlying provider edge stays. Blended effective edge ≈ 1.10 % - small for a casino, but structurally present.

The single biggest hole in the ledger: −$29.9M on $3.39B wagered (−0.88 %). Monarch has alluded to this on stream, mentioning a player who won big on Plinko - consistent with a handful of outliers buried inside enormous volume.

Across $1.75 billion wagered the house held just 0.59 % - remarkably low for any table game, consistent with a player base running close to basic-strategy optimal.

0.51 % effective edge - the $50k daily wager-cap mechanism showing through on the most-played original.

0.96 % effective edge - Beef is a Duel-original carrying a real house edge (in effect until May 2026), making it the single original that doesn't sit on the 0 %-edge framework.

The sportsbook holds 2.91 % - Duel's highest-edge product by some distance, and the only category where the house edge is the standard bookmaker's vig rather than the wager-cap mechanism.

0.84 % effective edge.

0.52 % effective edge - same wager-cap mechanism as Dice.

Ran the house $3.2M underwater across $127M wagered (−2.52 %). At near-zero edge on comparatively thin volume, this is normal variance - nothing to absorb a bad run.

1.38 % effective edge - the highest of the no-edge originals.

The few who move everything

One statistic reframes everything above. Of the entire $11.5 billion wagered, the top 0.1 % of players - just 80 accounts - account for 65 % of all turnover. Widen the lens only slightly and the top 1 % (800 players) account for 87 %; the top 5 % for 96 %.

That's what the wager distribution looks like once you plot it: 79,000 accounts stretching across the lower bands, and then a single $10 M+ column at the right that holds nearly $8 billion of the total handle - more than every other band combined.

Duel is not, in any meaningful sense, a casino of 80,000 players. Like most casinos, it is a few hundred whales with a large, low-stakes crowd around them. The median active player wagered only a few hundred dollars across the entire 287 days; meanwhile 120 accounts wagered more than $10 million each. The single biggest wagerer - an account called "wpgg" - pushed $1.017 billion through the site on its own, and finished −$6.47 million.

Player results carry the same shape. Of every tracked player, 70 % ended the period net down - 56,157 losers against 20,385 winners and 3,542 who finished flat. But the money is, again, all in the tails: at one extreme, 12 players won a combined $48.1 million; at the other, 18 players lost a combined $38.6 million. The top single winner - "jarik" - wagered just over $1.00 billion and walked away $12.34 million up.

The growth curve

Map new sign-ups by month and Duel's first year has a clear shape. Registrations climbed steadily from launch to a first peak in November 2025 (10,547 new accounts), softened through the winter, bottomed out in February 2026 - and have climbed again every month since. That February turn is not random: it lines up with Duel bringing its live blackjack table online early in the year. New game, new players, going viral on X with millions of views.

The Casino paradox

There is one row in the game table we deliberately saved for last, because it carries the answer to the entire report. Casino - the third-party category, every external slot provider and live-dealer table on the site - shows +$48.2 million of house PnL on $4.33 billion wagered. On the face of it, that is where Duel makes its real money: of the casino's whole $135M of gross revenue, roughly $130M comes from this one category alone.

But "Casino" is the one place the 0 % promise quietly does not apply - and it is also the most expensive revenue Duel earns. Third-party content is not free to host. Game providers and live-table operators charge provider fees, commonly 10 % or more of GGR. On a third-party GGR in the order of $130 million, that alone is $13 million or more taken straight off the top - before Duel pays a single one of its own costs.

And those costs are real: every developer building the originals, the entire infrastructure the platform runs on, support, compliance, the affiliate program. Walk the full waterfall and the headline $44.89M NGR shrinks quickly - $8.73M to affiliates, $13M or more to providers, then the cost of running the company itself - and the genuinely retained profit on $11.5 billion of turnover is far thinner than the gross suggests.

So - is Duel profitable?

Bring it together. Duel's originals do exactly what they promise: they hold close to nothing, and on Plinko and Castle Roulette the house actually lost. A player who only ever touched the 0 %-edge originals, wagered below $50k a day, played reasonably, and stopped - would cost Duel money, not make it.

But that is not how players behave, and it is not where the revenue lives. The revenue lives in two places. It lives in the gap between a 0 % game and an imperfect, greedy player - the half-percent here, the one-percent there, that players give back simply by not stopping. And it lives in "Casino" - the third-party slots and live tables that carry a normal house edge, that drew 44,762 players, and that produced the overwhelming majority of the gross.

So here is the model, stated plainly. You lure players in with a 0 % house edge - a genuine, verifiable, marketing-proof 0 % - and you let the originals do the acquisition. And then, in the end, they lose it on Casino anyway. The 0 % is the door. It is not the room.

Is Duel profitable? Across these 287 days: yes - but narrowly, and not in the way the headline GGR implies. Walk the full math step by step:

And the company itself still has to be paid for out of that - every developer, the infrastructure, support, compliance. For a casino that turned over $11.5 billion, what's actually retained at the end is an astonishingly thin margin. A real business - but one balanced on volume, on the Casino category, and on the most reliable house edge ever discovered: the one inside the player.

Data: FairGambling Research · Duel public per-user export, cross-referenced against Duel's reported figures and our independent on-chain tracking. 287 days, 18 July 2025 to 1 May 2026.