How to Play Video Poker: Paytables, Optimal Play and Full-Pay Machines
Video poker is the rare casino game where skill and machine selection both matter. Full-pay 9/6 Jacks or Better returns about 99.54% with mathematically optimal play — a house edge of roughly 0.46%. The same game on an 8/5 paytable returns about 97.3%. Identical screen, identical rules, five times the cost.
Video poker looks like a slot machine and is not one. You make a real decision every hand, the paytable is printed on the screen, and the return is calculable exactly rather than declared by the manufacturer. That combination makes it worth learning.
The rules #
You are dealt five cards. You choose which to keep and which to discard, the discards are replaced once, and the final hand is paid according to the paytable. That is the entire game — one decision per hand, and it is the decision that determines your return.
In Jacks or Better, the most common variant, a pair of jacks or higher pays even money and anything lower pays nothing. Payouts rise through two pair, three of a kind, straight, flush, full house, four of a kind, straight flush and royal flush.
Reading the paytable, which is the whole game #
Machines are named after two numbers: what a full house pays and what a flush pays, per coin bet.
| Paytable | Full house | Flush | Return, optimal play |
|---|---|---|---|
| 9/6 (full pay) | 9 | 6 | ~99.54% |
| 9/5 | 9 | 5 | ~98.4% |
| 8/6 | 8 | 6 | ~98.4% |
| 8/5 | 8 | 5 | ~97.3% |
| 7/5 | 7 | 5 | ~96.1% |
Nothing else changes. Same cards, same rules, same interface — only those two lines. A player who does not read them can sit at a 7/5 machine believing they are getting the game’s advertised return and pay roughly seven times the house edge of the full-pay version.
Check the full house and flush lines before your first hand. It is the highest-value thirty seconds available anywhere in a casino.
Always bet maximum coins #
The royal flush pays disproportionately more on a five-coin bet than on fewer — typically 250 per coin at one to four coins, and 800 per coin at five. That jump is a meaningful part of the game’s total return, and playing fewer than five coins forfeits it, cutting the return by roughly 1.5%.
If five coins at a given denomination is more than you want to stake, move down a denomination rather than reducing coins. Five coins at 10p beats two coins at 25p, both in cost and in return.
Optimal play #
Every deal has one mathematically best hold, and it is often not the intuitive one. A few that reliably catch people out:
- Four to a royal flush beats a made flush. Breaking up a paying hand to chase the top prize is correct here — the royal’s value outweighs the flush you give up.
- A low pair beats four to a flush. The pair is more valuable than the draw.
- Never hold a kicker with a pair. Holding an ace alongside a pair of sixes reduces your draw for no gain.
- Four to an open straight beats a high pair? No. Keep the pair. Made hands beat most draws in this game.
- Discard everything on a hopeless deal. Five fresh cards beats holding a single high card you cannot build on.
Strategy charts exist for each variant and following one is what converts the paytable’s theoretical return into your actual return. Playing 9/6 Jacks or Better by instinct typically returns a percentage point or two below optimal — which is to say, roughly the difference between the best paytable and a mediocre one.
Variance, and what “99.54%” hides #
That figure includes the royal flush, which occurs roughly once in 40,000 hands. A substantial slice of the game’s return is locked inside an event most sessions will never see.
Day to day, a 9/6 machine returns less than 99.54% because you have not hit the royal. That is not the machine underperforming; it is the arithmetic of a rare, large payout. It also means bankroll matters more than the headline return suggests — the return arrives lumpily.
Other variants #
Deuces Wild, Bonus Poker, Double Bonus and Joker Poker each have their own paytables and their own optimal strategies. Full-pay Deuces Wild is theoretically above 100% with perfect play, which is why full-pay versions of it are close to extinct and any machine bearing the name almost certainly runs a reduced paytable.
The general rule holds: the name on the machine tells you the family, the paytable tells you the return, and only the paytable is binding.
What video poker costs #
At 9/6 Jacks or Better with correct play, roughly 46p per £100 staked — the lowest cost of any game covered here. At 8/5 with the same play it is about £2.70 per £100. At 8/5 with careless play, more still.
The difference between the best and worst version of the same game is larger than the difference between most games. That is the reason to learn it, and it is entirely within your control.
A hand, worked through #
You are playing 9/6 Jacks or Better at five coins. The deal is:
K♥ Q♥ J♥ 9♥ 4♠
Four hearts, including three high cards in sequence. Most players hold the four hearts and draw one, chasing the flush. That is not the best play.
Hold K♥ Q♥ J♥ and discard the 9 and the 4. You give up a strong one-card flush draw, and in exchange you keep three parts of a royal flush with two cards to come.
Why it wins: the royal pays 800 per coin at five coins — 4,000 coins. Nothing else on the paytable is close, and holding three to a royal with two draws is worth more in expectation than the flush draw you abandoned, even though the flush draw completes far more often. You are also still live to a straight, a flush, and any paying pair of jacks or better.
This is the pattern that makes video poker a skill game. The intuitive hold protects the likely small win; the correct hold accepts more losing hands in exchange for a much larger rare one. Deviating “just this once” toward the safe hold is precisely how a 99.54% machine returns 98% in practice.
Bankroll and the royal cycle #
A royal flush appears roughly once in 40,000 hands with optimal play. At 600 hands an hour — brisk but normal — that is around 66 hours of play per royal on average, and “on average” hides enormous spread. Plenty of players go several times that without one; some hit two in a week.
The practical consequence is that a session bankroll sized to the game’s headline return will usually be too small. Between royals you are effectively playing a game returning around 97.5%, and the missing 2% arrives all at once or not at all.
Treat the royal as a bonus you may never collect rather than as part of your expected session return, and size stakes so that a long stretch without one is survivable.
Progressive video poker #
Some machines add a progressive jackpot to the royal flush, funded from play. Because the base return of full-pay video poker is already close to 100%, a sufficiently large progressive can in principle push a machine over the line.
Two cautions. Progressive machines are usually offered on reduced paytables, so the base game is worse and the jackpot has to make up the difference before it is even neutral. And the calculation only holds if you play optimal strategy for that specific variant, which differs from the standard chart when a jackpot is in play. In practice, treat the progressive as a reason to check the paytable more carefully, not as a reason to assume value.
What does 9/6 Jacks or Better mean?
The full house pays 9 per coin and the flush pays 6. That paytable returns about 99.54% with optimal play. An 8/5 version of the identical game returns about 97.3%.
Why must I bet five coins?
The royal flush pays 800 per coin at five coins versus 250 per coin below that. Playing fewer forfeits that bonus and cuts your return by roughly 1.5%. Drop to a lower denomination instead of playing fewer coins.
Is video poker better than slots?
On a good paytable with correct play, substantially — roughly 0.46% house edge against 4% or more on a typical slot. On a poor paytable played casually the gap narrows sharply, so the advantage comes from machine selection and strategy, not from the game type.
Do I need a strategy chart?
To get the advertised return, yes. Playing by instinct usually costs one to two percentage points, which is about the same as choosing a bad paytable. The chart is what turns the theoretical figure into your figure.
Can video poker have a positive expectation?
Full-pay Deuces Wild is above 100% with flawless play, which is exactly why such machines are essentially gone. Assume any machine you find is set below 100%, and verify by reading the paytable.
Why am I not getting 99.54% back?
Because a large part of that return sits in the royal flush, which lands roughly once in 40,000 hands. Sessions without one return noticeably less, and that is the maths working as designed rather than a fault.