A progressive machine is an ordinary paytable with one row replaced by a counter. On a Jacks or Better progressive that row is almost always the royal flush. The meter starts at a reset value, usually the standard 4,000 coins for a five-coin bet, and a small fixed share of every bet placed on the machine, or on a whole bank of linked machines, is added to it until somebody hits the royal. Then it drops back to the reset and starts climbing again.
Every other row stays exactly where it was. That makes the return of a progressive easy to reason about, because only one payout is moving, and it means the question "when is this machine worth playing?" has a precise answer.
Where the meter starts
Progressives are usually built on a trimmed base table. The common one is 8/5 Jacks or Better: 8 per coin for a full house, 5 for a flush, everything else as on the full-pay machine. With the royal at its 4,000-coin reset it returns 97.30% with correct play, and the royal arrives about once in 40,200 hands. That is 2.70 points short of break-even, and the meter has to make up all of it on its own.
Whatever the meter says, read the table on screen first; Video Poker: Jacks or Better keeps its paytable in view for every variant.
The simple estimate, and why it is wrong
The obvious calculation divides the shortfall by the royal's frequency. Each extra coin per coin on the royal row is worth 1 ÷ 40,170 of a coin per hand, so closing 2.70 points needs 0.0270 × 40,170 ≈ 1,085 more per coin. That puts break-even at 1,885 per coin, or about 9,430 coins on a five-coin meter.
That figure is correct only for a player who keeps playing exactly as they would at 4,000. Nobody should. A bigger royal makes royal draws worth more, so hands that were close calls tip towards holding cards to the royal. Re-solving the whole game for each meter value, and playing the best strategy at each one, gives a lower crossing point:
| Royal meter (coins) | Per coin | Return | Royal once in |
|---|---|---|---|
| 4,000 (reset) | 800 | 97.30% | 40,200 |
| 5,000 | 1,000 | 97.81% | 36,400 |
| 6,000 | 1,200 | 98.39% | 33,700 |
| 7,000 | 1,400 | 98.98% | 33,000 |
| 8,000 | 1,600 | 99.59% | 32,600 |
| About 8,700 | About 1,730 | 100.00% | 32,600 |
| 10,000 | 2,000 | 100.82% | 32,200 |
Break-even with adjusted play is roughly 8,670 coins, or 1,733 per coin. On a quarter machine, where the reset is $1,000, that is a meter of about $2,170 rather than the $2,360 the simple estimate gives. Treat it as approximate, about 8,700, since published figures differ a little in rounding and any change to another row of the table moves it.
The adjustment is worth about 760 coins on the meter. Play an 8,670-coin royal with the 4,000-coin strategy and you get 99.62%, not 100%. The gap is not a rounding error; it is the value of changing your holds.
What changes in the strategy
At break-even the best hold differs from the reset strategy on about 3.6% of deals. Almost all of those changes go the same way: towards the royal, and away from small made hands and non-royal draws.
The clearest example is three to a royal against a high pair. Dealt J♠Q♠K♠J♥4♦ at reset, holding the pair of jacks is worth 1.526 per coin and holding the three spades 1.449, so you keep the pair. The three-card draw completes the royal on exactly 1 of the 1,081 possible two-card draws, so every extra coin per coin on the royal adds 1 ÷ 1,081 to its value. The two holds cross at 884 per coin, a meter of about 4,420 coins. At break-even the three spades are worth 2.31 and the pair is not close.
That is why the royal frequency in the table drops from one in 40,200 to about one in 32,600. You are not getting luckier. You are giving up some full houses, flushes and paying pairs to buy more royal draws, and at a big enough meter that trade pays.
If you want to see these holds for yourself, Video Poker: Jacks or Better has strategy hints that show which cards to hold, and you can compare them with your own choice.
What "positive" does and does not mean
A machine above break-even has a long-run return over 100%. That is a statement about averages over tens of thousands of hands, and the royal carries more of the return the higher the meter goes. At 8,670 coins, the royal row alone is worth about 5.3 points of the return. Three consequences follow:
- The wait is long. At one royal in 32,600 hands and 400 hands an hour, the average wait is about 81 hours of play. The chance of a royal in any one hour is about 1.2%, and over a full 32,600 hands there is still a 37% chance of seeing none, the same figure as in the royal flush post.
- On a linked bank, other players are drawing at the same meter. If one of them hits it, the meter resets and your machine is back to 97.30% straight away, whatever you had hoped to collect.
- The variance rises with the meter, because more of the return sits in one rare hand. A positive machine still loses for most people over most sessions, and the stack it takes to play one through is larger than at reset (see bankroll and session length).
The meter's own arithmetic also explains why progressives exist. Over a full cycle, from reset to hit, the money paid into the meter goes back to players as a group, so the average return across the cycle is roughly the base return plus the meter's contribution rate. Most of that cycle is spent below break-even, which is where most of the play happens.
There is no jackpot meter in Video Poker: Jacks or Better. None of its four variants (Jacks or Better, Deuces Wild, Bonus Poker and Double Bonus) has a meter row on its paytable, so none of the meter values above apply to it. What does transfer is the habit: read the table, work out which row carries the return, and hold accordingly. It plays offline, and the chips have no cash value.