Video Poker Strategy
Once the stake is set, the hold is the only decision left to you, and for any hand and pay table one set of cards returns more on average than the others. This guide shows how it is found, what mistakes and borrowed strategies cost, and what perfect play changes over a session. Values are credits returned per credit bet; costs are in points, one point being one percentage point of return.
Take 5♠ 5♥ 8♠ J♠ K♠ on 9/6 Jacks or Better. Keep the four spades and 47 cards remain. Nine are spades and complete a flush paying 6, which is 54 credits, and six more (three jacks, three kings) make a paying pair worth 1 each. That totals 60 credits over 47 possible draws, or 1.28 per credit bet. The pair of fives, averaged over its 16,215 possible three-card draws, is worth 0.82. The flush draw is correct, and choosing the pair gives up 0.45 per credit, or 2.26 credits on a five-credit bet. The correct hold has the highest average, not the most frequent win or the biggest possible one.
How Perfect Strategy Is Found
There are 2,598,960 different five-card deals from a 52-card deck, or 2,869,685 when a joker is added, and each can be held in 32 ways. For every deal and each of those holds, the pay table is averaged over every possible draw: 47 draws when one card is replaced, 1,081 for two, 16,215 for three, 178,365 for four and 1,533,939 when all five go. The best of those averages is the correct hold, and the average of the best values over all deals is the game's return under perfect play: 99.54% for 9/6 Jacks or Better. Nothing is sampled, so the answer is exact, and each pay table gets its own.
The same calculation prices every mistake. What a misplay costs over time is its size multiplied by how often the hand comes up. Four to a royal is the correct play on only 0.04% of deals, while keeping two unsuited high cards is correct on 15.1% and a single high card on 14.6%. Small misjudgments on those common hands add up faster than rare spectacular ones.
Worked Decisions: The Best Hold Against the Tempting One
Values are the average credits returned per credit bet across every possible draw, at the maximum bet. Four to a royal beats a made flush because one of the 47 unseen cards completes a royal worth 800 per credit.
The Cost of Borrowing a Strategy
Each row plays every deal with the hold that is perfect for the first game and scores it with the second game's pay table, at the maximum bet.
Why Some Strategies Travel and Others Break
Between Jacks or Better tables, 9/6 strategy travels almost free, as the first row of the table shows. A cheaper full house or flush moves the values without changing their order, except on a few rare hands. Q♠ J♠ 9♠ from Q♠ J♠ 9♠ 8♦ 3♣ is worth 0.72 on 9/6 and 0.68 on 8/5, and beats keeping Q♠ J♠ (0.59 and 0.58) on both.
The rule: A strategy breaks when the table changes a key line: which hands pay at all (a pair of tens in Tens or Better), what two pair pays, a large four-of-a-kind bonus, or a wild card.
Take A♠ A♦ K♠ K♦ 5♣. On Bonus Poker, where two pair pays 2 per credit, keeping both pairs wins clearly. On 10/7 Double Bonus, where two pair pays only 1, both pairs and the aces alone are almost exactly tied. On Double Double Bonus, where four aces with the right kicker pay far more, the aces alone win. Each of those pages gives the values. What happens when a wild card is added is measured on the Deuces Wild page.
Short Lists, Perfect Play and What Either Buys
A strategy list compresses perfect play into ranked patterns: you play the highest line your hand matches. Our 25-line list for 9/6 Jacks or Better plays at 99.50%, against 99.54% for perfect play. Its largest remaining gap is a single pattern. On 0.478% of deals the list holds a three-card straight flush draw with no high card where perfect play holds a single high card, and that exception alone costs 0.0203 points, nearly half the list's total shortfall. The other lists, each on its variant page:
- Jacks or Better 7/5: 96.10% against 96.15% (0.045 points)
- Tens or Better 6/5: 99.08% against 99.14% (0.058)
- Bonus Poker 8/5: 99.12% against 99.17% (0.046)
- Double Bonus 10/7: 99.97% against 100.17% (0.198)
- Double Double Bonus 9/6: 98.80% against 98.98% (0.182)
- Deuces Wild Full Pay: 100.68% against 100.76% (0.083); NSU: 99.57% against 99.73% (0.159)
- Bonus Deuces Wild 10/4/4/3: 99.66% against 99.86% (0.200)
- Joker Poker 20/7/5: 100.42% against 100.65% (0.228)
The right list for the game beats a good list for another one: the Double Bonus 10/7 list gives up 0.198 points on its own table, while Jacks or Better strategy played there gives up 0.540. The choice of table matters more than either. Over 2,000 hands, a few hours at several hundred hands an hour, perfect 9/6 play finishes ahead in 31% of sessions, with a median of −46 bets (one bet being the stake on one hand) and an expected loss of 9.1 bets. The same skill on a 7/5 table, the version in Habanero's Jacks or Better, finishes ahead in only 12.1%, with a median of −114 and an expected loss of 77 bets. Eight sessions in ten end between −155 and +92 bets on 9/6, and between −217 and +16 on 7/5. A royal turns up in 4.83% of those sessions on 9/6 and 4.86% on 7/5, so the gap comes from the everyday hands.
Takeaway: The 25-line list costs 0.043 points; choosing 7/5 over 9/6 costs 3.39, far more than any list error, and no skill wins it back.
To practice, pick a free game from our library that runs the table you mean to play, learn that one list until it is automatic, and deal until you have met each worked decision above. Start on single-hand play, since a multi-hand game copies one mistake to every line, and treat any hold a game suggests on screen as a hint rather than a rule.
Video Poker Myths
A royal is due after a long drought
Thousands of hands have passed since the last royal, so one must be close.
Reality Check
Each deal starts from a newly shuffled deck, and no approved game may change its odds to make up for past payouts (dealing is covered in the how-to-play guide). The chance of a royal on the next hand is the same after a long drought as after a royal: about 1 in 40,391 hands with perfect 9/6 play.
Switching machines changes your luck
After a cold run, moving to another machine gives a fresh start.
Reality Check
No machine remembers your run, so a new seat changes nothing about the cards. The only good reason to move is a better pay table.
Bet size changes the odds
Betting more when a hand is "due", or less when cold, shifts the chances.
Reality Check
The stake does not touch the deal, and on a table where every line scales with the bet it does not change the correct hold either. What it can change is the royal flush prize on some games, which is a pay-table question. The pay tables guide gives the numbers.
Keep a kicker just in case
Holding a high card beside your pair adds a chance of something better.
Reality Check
It removes a draw card instead. From A♥ A♦ K♣ 7♠ 3♦ on 9/6 Jacks or Better, the aces alone are worth 1.54 per credit and the aces plus the king 1.42. Where a kicker does pay extra, the answer depends on the hand, and the Double Double Bonus page works through those decisions.
Video Poker Strategy: FAQs
No. Each line draws from its own deck, so every line has the same average value for a given hold, and the hold that is best for one hand is best for all of them. Multi-hand play changes how widely a session swings, not what to keep.
No. The double-up is offered only after a hand is settled and is an even-money bet, so it adds nothing to the value of any hold. Play the hand as if the feature did not exist.
It can, for a few hands built around the royal flush. On games whose royal drops from 800 to 250 per credit when you bet less than the maximum, draws to a royal are worth less, so the balance tips toward other holds on some close calls. The lists on this site assume the maximum bet, which is also where the return is highest.
In the long run and on average, yes, but a session is far too short to count on it. Full Pay Deuces Wild returns 100.76% under perfect play, and after 2,000 hands it expects to be 15.2 bets ahead, one bet being the stake on one hand. Yet only 40.5% of those sessions finish ahead, and the median session is 41 bets down, because much of the return waits in rare hands.