Payback percentage is the long-run share of wagered money that a video poker game is expected to return under the strategy and paytable assumptions used in the calculation. It is commonly another way of expressing RTP. The number is useful for comparing games, but it is one of the easiest casino statistics to misread as a promise about a short session.
Translate the percentage into a price first
A 99% payback does not mean a machine takes $100 and gives exactly $99 back to the same player. It means the game’s long-run mathematical return is 99% of total wagers under the stated assumptions.
The complementary number is the house edge:
House edge = 100% - payback percentage
So a 99.5% theoretical payback corresponds to a 0.5% theoretical house edge. That conversion is useful because it turns a promotional-sounding return figure into the cost side of the same equation.
For a deeper comparison of the terms, see RTP versus house edge.
The denominator is total wagers, not the cash in your pocket
Payback applies to coin-in, the total amount wagered, not simply the cash originally inserted.
A player can arrive with $200 and generate $2,000 or more in coin-in by repeatedly wagering credits returned from earlier hands. If the average bet is $5 and 400 hands are played, coin-in is $2,000 regardless of the original bankroll.
Coin-in = bet per hand × hands played
Expected loss = coin-in × house edge
This is why a small mathematical edge can matter over a fast session. The coin-in and expected loss per hour pages connect the percentage to actual play volume.
Video poker adds a strategy condition that slots may not
In many video poker analyses, the published payback assumes a defined strategy—sometimes optimal play, sometimes a specified practical strategy. That means the number is not only about the machine’s paytable. It is also about the player’s hold decisions.
A machine can display a strong schedule while a player gives up return through repeated strategy errors. The game name alone therefore does not guarantee the player will achieve the quoted percentage.
That is one reason a video poker analyzer is useful. It can compare the expected value of holds under the exact paytable rather than relying on intuition.
Paytable changes can move payback dramatically
Two machines labeled Jacks or Better can look almost identical while paying different amounts for full houses, flushes, or other hands. Those differences alter the weighted average return because each payout is multiplied by the probability of reaching that result.
The general structure is:
Payback = Sum of (final-hand probability × payout for that hand)
The final-hand probabilities themselves depend on strategy. That is why a paytable comparison should include both the reward schedule and the strategy assumption.
The Wizard of Odds video poker summary shows how returns differ across variants and schedules, while the 9/6 Jacks or Better strategy reference demonstrates the strategy side of a specific high-return schedule.
A session can finish anywhere around the average
Suppose a game has a theoretical payback near 99%. A player might still lose 20% of the bankroll in one short session, double the bankroll, or finish close to even. Those results do not prove the published return wrong.
Payback is an average over a very large number of equivalent wagers. Short play is dominated by variance—the uneven arrival of pairs, straights, quads, royal flushes, and losing hands.
The difference becomes especially important when rare premium hands carry a large share of the game’s return. A player who does not hit those hands during a session may experience results far below the long-run average. Read RTP versus variance for that distinction.
Higher payback does not automatically mean the better personal choice
If two games are otherwise similar, a higher theoretical payback is generally a better long-run price. But “otherwise similar” can hide several practical differences.
- One game may require a strategy you cannot execute accurately.
- One may be much more volatile.
- One may be offered only at a denomination that forces a larger wager than you want.
- One may require an extra feature bet to achieve the quoted return.
- One may be played much faster because the decisions feel simpler.
A slightly lower-payback game at a comfortable wager can cost less per hour than a higher-payback game played at several times the action.
This is why percentage ranking should be followed by an hourly-cost check rather than treated as the final answer.
Compare exact configurations, not game names
A useful comparison table includes:
| Item | Game A | Game B |
|---|---|---|
| Variant | exact title | exact title |
| Paytable | exact payouts | exact payouts |
| Theoretical payback | stated/calculated % | stated/calculated % |
| Strategy assumption | optimal/simple/other | optimal/simple/other |
| Bet per hand | money amount | money amount |
| Estimated hands/hour | realistic pace | realistic pace |
| Expected hourly loss | calculated | calculated |
| Variance | lower/higher context | lower/higher context |
That comparison prevents a common error: choosing a machine because “99.7 is bigger than 99.4” while ignoring that the 99.7 game is being played at twice the wager and with a strategy the player does not know.
What “99% payback” does and does not say
It does say that under the modeled rules and strategy, average return is 99 cents per dollar wagered over the long run.
It does not say that every $100 session loses $1.
It does not say that the casino changes results to force the percentage back into line after a player wins.
It does not say that every machine with the same title has the same schedule.
It does not say that a player using weaker strategy will receive the same theoretical return.
Those boundaries are more valuable than memorizing the percentage itself.
The casino uses theoretical and actual results for different purposes
Casino operations can compare theoretical hold with actual hold, coin-in, denomination, occupancy, player mix, and other performance measures. A machine that has run unusually high or low over a short period does not automatically indicate a configuration problem; variance can produce substantial deviations.
At the same time, device configuration and integrity must be controlled. References such as GLI standards and Nevada technical standards address regulated gaming-device requirements. Those technical controls are separate from a player’s misunderstanding of what a long-run return percentage means.
For management, “payback” is therefore not a promise to an individual player. It is one input into game configuration, financial expectation, and product comparison.
A worked cost comparison
Assume two hypothetical games are both played at 500 hands per hour.
Game A: 99.5% payback, $5 per hand.
Hourly coin-in = 500 × $5 = $2,500
House edge = 0.5%
Expected hourly loss = $2,500 × 0.005 = $12.50
Game B: 99.0% payback, $1.25 per hand.
Hourly coin-in = 500 × $1.25 = $625
House edge = 1.0%
Expected hourly loss = $625 × 0.01 = $6.25
Game A has the better percentage but the higher expected hourly cost because the wager is four times larger. This does not make Game B universally “better”; it shows why the percentage must be combined with wager size and pace.
Actual results in either hour can be far above or below those expectations.
Use payback as a filter, not a forecast
A disciplined player can use payback percentage in three steps.
First, eliminate obviously poor schedules when stronger comparable ones are available. Second, verify that the return assumes a strategy you can realistically follow. Third, translate the house edge into expected cost at your actual wager and pace.
That approach keeps the number useful without asking it to predict the next hour.
If you want to continue, read video poker RTP for the calculation side, RTP versus variance for session behavior, and high-RTP games can still lose fast for the bankroll lesson. The house edge calculator can help convert return percentages into the casino-price view.