VPK 111: Video Poker House Edge — Paytable Plus Strategy

Video poker house edge is unusual because it depends on two things at once: the machine’s exact paytable and the player’s strategy. A full-pay 9/6 Jacks or Better game is commonly calculated at about 99.5439% return with optimal play, which corresponds to about 0.4561% house edge. The same cabinet running a weaker paytable can have a much larger edge, and poor hold decisions can push the player’s achieved return lower still.

So the useful question is not simply “What is the house edge of video poker?” It is: Which game, which paytable, how many coins, and what strategy?

The basic conversion is simple

For a specified game and strategy:

House Edge = 1 - RTP

Using percentages:

House Edge % = 100% - RTP %

For 9/6 Jacks or Better at 99.5439% theoretical return:

House Edge = 100% - 99.5439%
           = 0.4561%

That means an optimally played $1,000 of total action has an expected loss of about:

$1,000 × 0.004561 = $4.56

It does not mean every $1,000 session loses $4.56. A royal flush, four of a kind, or long dry sequence can move a short session hundreds or thousands of dollars away from expectation. House edge prices the long run; video poker variance describes the size of the ride around that average.

Paytable changes can matter more than the game name

“Jacks or Better” is not one mathematical product. The full-house and flush rows are enough to distinguish common versions.

The two schedules look similar, but the edge differs by roughly 2.25 percentage points.

At five $1 credits per hand for 500 hands, coin-in is $2,500:

9/6 expected loss = $2,500 × 0.004561 ≈ $11.40
8/5 expected loss = $2,500 × 0.027016 ≈ $67.54

That is about a $56 difference in long-run expected cost for the same number of $5 hands.

The video poker paytables guide explains why apparently small changes to recurring hands such as flushes and full houses have a large cumulative effect. The dedicated 9/6 Jacks or Better page covers that benchmark game separately.

House edge is not the chance that a hand loses

Video poker can produce many hands that return less than the amount wagered, including complete misses and small pays that do not compensate for previous losses. None of that makes “percentage of losing hands” interchangeable with house edge.

House edge weights every final hand by both its probability and its payout. A rare royal flush contributes a large amount to long-run return even though it occurs infrequently. Frequent low-paying pairs contribute in a different way. This is why removing one unit from a full-house or flush payout can matter materially: those hands occur far more often than a royal, so a modest reduction is applied repeatedly across the long run.

The same logic explains why a game with a high hit rate can still have a worse house edge than a game that pays less often. Frequency tells you how often something returns credits; expected value tells you how much the complete payout distribution returns per unit wagered.

Strategy is part of the mathematical return

This is the point that separates video poker from a conventional reel slot.

After the initial five cards are dealt, the player chooses which cards to hold. Different holds lead to different sets of possible final hands, so each decision has its own expected value. Optimal strategy chooses the available hold with the highest expected return for that exact hand and paytable.

In conceptual form:

EV of a Hold = Σ (Probability of Each Draw Result × Payout of That Result)

For every dealt hand, the correct strategy compares the EV of all legal holds and takes the maximum. The game’s published optimal RTP is the weighted average of those best decisions across all possible initial hands.

A strategy mistake does not change the machine’s printed paytable. It changes the branch of outcomes the player chooses to enter.

Suppose a game has an optimal return of 99.5439%, but a player’s mistakes cost an average of 0.8 percentage point over a large sample. The player’s strategy-dependent return would be approximately:

99.5439% - 0.8% = 98.7439%

and the effective house edge for that pattern of play would be about:

100% - 98.7439% = 1.2561%

On $2,500 coin-in, the expected loss would rise from about $11.40 under optimal play to about $31.40. The 0.8-point error cost here is an illustration, not a universal estimate; actual strategy loss depends on the specific mistakes and how often they occur.

For learning the decision process, use Video Poker Strategy Basics and How to Read a Strategy Chart.

“Max coins” is a paytable question, not a ritual

Many classic video poker schedules award a disproportionately large royal flush for the maximum coin wager. A common structure pays 250 credits per coin for a royal at one through four coins, but 4,000 credits for a five-coin royal. In that design, playing fewer than five coins changes the effective paytable and lowers theoretical return.

That does not mean every video poker machine in every denomination has the same max-coin rule. Some modern games express wagers differently or use alternative jackpot structures. Read the actual paytable before assuming that “five coins is always optimal.”

There is also an important bankroll distinction: if the correct five-credit wager is too large for your budget, switching to a lower denomination at the full required coin count can be mathematically preferable to playing an impaired royal schedule at a higher denomination—provided the lower denomination offers the same paytable.

Regulators treat video poker as a skill-affected game

The role of strategy is not merely casino folklore. The Nevada Gaming Control Board’s New Gaming Device Submission Package requires documentation of theoretical payout and specifically asks manufacturers of skill-affected games such as video poker to describe the optimal strategy and how it produces the best long-term average return.

That is the correct way to read a return figure: it is tied to a mathematical configuration and a defined method of play.

If a site, sign, or game information screen quotes a high video poker RTP without stating the paytable and strategy assumption, the number is incomplete.

Strategy charts must match the paytable

A correct hold on one version is not guaranteed to be correct on another. Changing the value of a full house, flush, four of a kind, wild-card hand, or progressive award can alter the relative EV of competing holds. The differences are often concentrated in borderline hands, which is exactly where intuition is least reliable.

That means “I know Jacks or Better strategy” is incomplete unless the player knows which schedule the chart was built for. Using a near-match chart may still be better than guessing, but its published return no longer describes perfect play of the machine in front of you. The common strategy mistakes guide shows how small decision leaks accumulate over many hands.

Coin-in can overwhelm a very small percentage

A low edge is valuable, but speed and stake still matter.

A player making 600 hands per hour at $5 per hand generates:

Hourly Coin-in = 600 × $5 = $3,000

At a 0.4561% edge:

Expected Hourly Loss ≈ $3,000 × 0.004561 = $13.68

At a 2.7016% edge:

Expected Hourly Loss ≈ $3,000 × 0.027016 = $81.05

Those are theoretical averages. The actual session can be far more volatile, particularly because a meaningful share of long-run return can come from rare high-paying hands.

The calculation also shows why “under 1% house edge” is not the same thing as “cheap forever.” Playing faster, raising denomination, or extending the session increases coin-in. The Expected Loss Calculator lets you model the dollar effect directly.

A progressive jackpot can change the edge, but not automatically

Some video poker games feed a progressive jackpot into the royal or another rare hand. As the jackpot grows, the expected value of the game can increase because one payout has become larger while its probability remains governed by the game and strategy.

In principle, a sufficiently large progressive can push the theoretical return toward or above 100%. But finding the break-even point requires more than looking at the meter. You need:

• the exact base paytable;
• the jackpot amount and qualifying wager;
• the probability of the progressive hand under the correct strategy;
• any strategy changes caused by the larger jackpot;
• eligibility, meter-reset, and payout rules.

A progressive therefore does not turn every nearby machine into a positive-expectation game, and a positive theoretical edge would not eliminate short-term risk.

House edge, hold, and actual results are different numbers

Three percentages are often mixed together:

Theoretical house edge is 100% minus the theoretical return under a defined strategy.

Player-achieved edge reflects the strategy actually used. Mistakes can make it worse than the optimal figure.

Actual machine hold is an accounting result observed over a period. It can be above or below theoretical because real hands include variance, different players, different strategies, jackpots, and changing amounts of action.

A machine winning 8% of coin-in over a short period does not prove its mathematical house edge is 8%. Likewise, a machine losing money to players during a jackpot period has not become a negative-edge game by definition.

How to compare two video poker games properly

If your goal is to minimize mathematical cost, compare in this order:

1. Identify the exact game family. Jacks or Better, Bonus Poker, Deuces Wild and others use different pay structures and strategies.
2. Read the entire paytable. Do not judge only the royal flush row.
3. Check the qualifying wager. Confirm how coin count or bet size affects major payouts.
4. Find the return for that exact schedule. A generic game-name RTP is not enough.
5. Use the matching strategy. A strategy chart for one paytable can be wrong for another.
6. Then consider variance and bankroll. The lowest edge may still come with swings you do not want to fund.

The video poker RTP guide focuses on return as a percentage. This page’s narrower point is that house edge in video poker is conditional on both the paytable and the decisions made against it.

If you remember only one number, remember the formula. If you remember one operational rule, remember this: never compare video poker machines by title alone.