A natural-card game with three quad tiers
Double Bonus uses 52 ordinary cards and no wild cards. After the initial five-card deal, a player keeps any cards and draws replacements once from the remaining 47. A pair of jacks, queens, kings or aces is the minimum paid result. The final hand earns one category award, not the sum of overlapping poker combinations. Our automatic simulator samples the final categories associated with optimal decisions.
Nine, seven and five all matter
The three numbers mean full house 9, flush 7 and straight 5 per unit wager. This version also pays 50 for a straight flush. The nearby 55-straight-flush schedule is a different preset and has a different return. Do not identify this model merely as Double Bonus 9/7: a straight paying four changes both expected value and the best decisions on some starting hands.
The complete five-coin paytable
| Final category | Credits won (5 coins) |
|---|---|
| Royal Flush | 4,000 |
| Straight Flush | 250 |
| Four Aces | 800 |
| Four 2s–4s | 400 |
| Four 5s–Kings | 250 |
| Full House | 45 |
| Flush | 35 |
| Straight | 25 |
| Three of a Kind | 15 |
| Two Pair | 5 |
| Jacks or Better | 5 |
| All other hands | 0 |
These awards include the original wager. Five credits returned means break-even, not five credits of profit. At $0.25 per credit, a hand costs $1.25 and a 4,000-credit award returns $1,000. Changing denomination scales the dollars without changing probabilities. This model always wagers five coins, preserving the maximum-coin royal; it does not apply the same return to smaller coin-count schedules.
Big quads do not make two pair profitable
Four aces return 160 times the stake; four twos, threes or fours return 80; all other quads return 50. The fifth card is irrelevant to those awards. Two pair returns only the original stake, unlike the two-for-one return in Bonus Poker 8/5. That tradeoff is why two games with similar names can feel different over the same session length and denomination.
Use the right strategy reference
Flush and straight draws are worth more than in many familiar Jacks or Better schedules, while the richer ace award increases the importance of some ace holds. These observations explain why a borrowed strategy can cost return; they are not a complete ranking of draws. The linked 9/7/5 exceptionless strategy is reported at 99.1012%, slightly below the table’s 99.1065% optimal model. The simulator’s exact frequencies should not be read as a promise that a short list of tips achieves optimal play.
A kicker decision that really is simple
For A♠ A♥ A♦ A♣ 7♣, keeping the four aces guarantees 800 credits on five coins. Drawing a new fifth card cannot produce a higher category, and no kicker can increase the quad payout. This differs sharply from Triple Double Bonus. A final hand of 5♣ 6♦ 7♥ 8♠ 9♣ pays 25 credits here, not 20; that extra straight value is precisely why the third number belongs in the preset name.
High return still allows losing sessions
The selected integer-weight model has an optimal theoretical RTP of 99.1065%. Its calculated variance is 28.547130 in squared total-stake units; standard deviation is 5.342951 stakes per hand. Those figures assume independent hands and optimal decisions on this exact paytable, with no progressive jackpot, cashback, promotions or strategy mistakes.
For a fixed number N of hands, expected loss is N times the total hand wager times one minus the return fraction. The standard deviation of total net return is the hand wager times the square root of N times variance. Rare jackpots make short-session outcomes asymmetric, so this formula is not a promise that a particular session ends near its expectation. An affordable bankroll can still run out before the requested hand count.
Compare several sessions with the same settings, then change one setting at a time. A faster speed changes only the pace. Increasing the stake magnifies dollar swings without making a jackpot due. Set an affordable entertainment limit; simulated results are neither income projections nor a reason to chase losses.
Sources & calculation notes
The selected source is the standard 9-7-5 table with straight flush 50. We use its published integer combination weights, not rounded six-decimal probabilities. The 55-SF and 9/7/4 tables are excluded.
The counts preserve the published analyst’s precision. They verify the implemented distribution, not an independent re-enumeration of every possible optimal hold. Displayed percentages are rounded; sampling uses the original integer weights.
Reviewed October 4, 2026. The worked examples explain particular rules and decisions; they are not a complete optimal strategy chart.