99% RTP
Same edge on every pick count from 1 through 10.
Pick up to ten numbers on an eighty-cell grid. The game draws twenty. More matches mean bigger pay. Pick count sets the variance: one number is a coin flip with a small multiplier, ten numbers is a jackpot hunt with long dry spells. RTP holds at 99%.
Same edge on every pick count from 1 through 10.
Twenty numbers drawn per round. Pay table scales with how many you picked.
Run the same selection across hundreds of draws without re-clicking the grid.
Keno is the slow-burn Original. Sessions can run hundreds of rounds with no meaningful hit if you play ten picks and aim for the top row of the pay table. One-pick rounds feel like a choppy even-money game. The draw is provably fair, the pay table is on screen, and hot/cold number tracking is useless because each draw reshuffles the pool logic from seeds, not from history.
Hot and cold number boards are noise. Each draw is seed-derived, not history-weighted.
10,000× on ten-of-ten hits is real math with lottery odds (~1 in 8.9 million). Most ten-pick sessions end at zero to four catches. Budget like a lottery, not like blackjack.
Choose how many numbers to play (1-10) and which numbers from the 80-cell grid. Place stake. The system draws 20 numbers. Payout depends on the pick-count and how many of yours hit. With 10 picks, hitting 5 of 20 pays ~3×; hitting 10 of 20 pays a jackpot multiplier of ~10,000×. Pay-tables are published in-game.
Pay tables change by pick count. Open the in-game table before you select numbers so you know what three catches pays versus seven.
Auto repeat runs the same selection across hundreds of draws. Essential for ten-pick jackpot hunts.
Pick count sets variance. RTP stays 99% across all counts.
Select count, select numbers, draw twenty, pay by matches.
| Rule | Duel Keno |
|---|---|
| RTP / house edge | 99% / 1% |
| Number pool | 80 numbers, 20 drawn per round |
| Pick count | 1-10 numbers per round |
| Top jackpot | ~10,000× on 10/10 match |
| Min bet | $0.10 |
| Provably fair | Draw order from HMAC shuffle of 80 |
RTP is constant at 99% across every pick-count configuration, what changes is variance. Picking 1 number is a near-50/50 (1.4× return on 25% hit rate). Picking 10 numbers is a long-shot lottery (10,000× on a ~1 in 8.9 million hit). Auto mode is essential for hitting enough rounds to converge on the 99% return; one 10-pick session can run hundreds of rounds without a payout.
Pick count is a variance dial, not a skill test. Auto mode exists because you need volume to approach 99% RTP. One manual round tells you almost nothing.
The 20 drawn numbers are a deterministic function of the seed pair and nonce, mapped through a Fisher-Yates shuffle of the 80-position pool. Reproduce the shuffle, take the first 20 positions, confirm they match the casino's published draw.
Duel Keno: how pick count changes variance, jackpot odds, auto mode, and reproducing the draw from seeds.
Keno is Duel's lottery-style Original. Players mark numbers, the system draws twenty from eighty, and payouts follow the published matrix for each pick count.
1% house edge regardless of pick count. More picks increase jackpot size and decrease hit frequency.
Ten-pick jackpot hunting needs a bankroll sized for thousands of empty draws. One-pick play suits smaller swings. Set a session loss cap in dollars.
Versus Beef, Keno removes the instant head-to-head reveal for multi-number suspense. Versus real-world lottery, Keno runs every few seconds with verifiable draws.
Counts toward VIP volume.
Mines and Keno both suit long auto sessions with different tension profiles.
Twenty drawn numbers = first 20 positions from 80-cell shuffle. Match all 20.
Use this table to decide between Keno and Castle Roulette. Links go to full guides on this site.
| Topic | Keno | Alternative |
|---|---|---|
| Pace | Lottery draw | Castle Roulette: Table spin |
| RTP | 99% | Castle Roulette: 100% main wheel |
| Variance | Pick-count driven | Castle Roulette: Low on even money |
No. Each number has equal draw probability. Past draws do not matter.
Catching all picks is much rarer than catching one. Hitting any match actually gets easier with more picks.
~10,000× on 10/10. Extremely rare.
No. Each number once per round.
Yes. Repeat same picks across many draws.
Full grid and verification on mobile.
Recompute shuffle, first 20 positions = drawn set.
Yes.
1-4 picks for frequent small hits. Ten picks is jackpot hunting.
Castle Roulette is 0% on the main wheel. Keno is 99% lottery draw.
Useless. Each draw is seed-derived, not history-weighted.
Yes. Run the same selection across hundreds of draws.
Pick numbers. Watch the draw. Verify the twenty drawn positions.
Play Keno at Duel