Three Card Poker Odds

Three Card Poker looks mathematically small because each hand contains only three cards. The underlying analysis is not small. The main Ante/Play game depends on two three-card hands, the player acting first, dealer qualification, a fold/raise decision, and an Ante Bonus paytable. Pair Plus uses the player’s hand only and has a different return calculation.

The safest way to read Three Card Poker odds is therefore to keep hand probability, main-game expected value, and side-bet paytable value in separate columns.

Begin with the 22,100 possible player hands

From a standard 52-card deck, the number of unordered three-card combinations is:

C(52,3) = 52! / (3! × 49!) = 22,100

Under the standard three-card ranking used by the commonly analyzed game, the hand frequencies are:

These figures match the current published Three Card Poker analysis at Wizard of Odds.

Three-card ranking reverses one familiar poker relationship

In five-card poker, a flush outranks a straight. In standard Three Card Poker, a straight outranks a flush.

The reason is visible in the table: a three-card straight occurs in 720 of 22,100 hands, while a three-card flush occurs in 1,096. The straight is rarer, so it ranks higher.

That single reversal matters twice. It affects main-hand comparison and it affects how a Pair Plus paytable should be read. A player who imports five-card rankings can misread both the result and the advertised payout.

The main game is not just the probability of beating one random hand

If the game were simply “deal two random three-card hands and compare them,” the analysis would be much easier. Standard Ante/Play adds a first-mover disadvantage: the player must decide whether to fold or place the Play wager before seeing the dealer’s hand.

Under the common rules:

• the player starts with an Ante;
• after seeing the player hand, the player folds or makes a Play wager equal to the Ante;
• the dealer reveals the dealer hand;
• the dealer needs queen-high or better to qualify;
• if the dealer does not qualify, the Ante wins and the Play pushes;
• if the dealer qualifies, Ante and Play are compared against the dealer;
• an Ante Bonus may pay on strong player hands under the posted paytable.

So “What are my odds of winning?” has no single useful answer until you specify whether you mean hand category, dealer comparison, net outcome of the Ante/Play contract, or side-bet result.

Queen-six-four is a decision boundary, not a lucky hand

For the commonly analyzed standard rules, the well-known strategy is to raise with Q-6-4 or better and fold weaker hands.

The boundary exists because continuing with Q-6-4 has a slightly better expected result than surrendering the Ante, while continuing with Q-6-3 is slightly worse than folding. It is an expected-value threshold, not a claim that Q-6-4 is likely to beat the dealer.

This is why “always raise on any queen” is close but not optimal. Current published analysis puts the house edge of that looser mimic-the-dealer approach around 3.45% under the analyzed common rules, compared with about 3.37% for the standard full-pay Ante Bonus table and optimal Q-6-4 strategy.

Use Three Card Poker strategy when you want the decision rule rather than the probability table.

The famous 3.37% needs its denominator beside it

For the commonly analyzed Ante Bonus schedule of:

• straight: 1 to 1;
• three of a kind: 4 to 1;
• straight flush: 5 to 1;

the published main-game analysis gives a house edge of about 3.37% measured against the initial Ante.

That percentage is not the same as average loss divided by all money eventually wagered. Because the player sometimes folds and sometimes adds the Play wager, average total action exceeds one Ante. The same analysis reports an element of risk around 2.01% for that paytable.

Both numbers can be correct because they use different denominators.

House edge on Ante
= Expected loss / Initial Ante

Element of risk
= Expected loss / Average total amount wagered

This denominator problem is common across carnival games and is the reason house edge should never be quoted without saying what the percentage is measured against.

Dealer qualification changes settlement, not the strength of your cards

Dealer qualification is often misunderstood as an extra probability bonus attached to a weak player hand. It is not.

Your cards are what they are. Qualification changes how the Ante and Play settle when the dealer does not reach queen-high or better. Under the common rules, the Ante wins and the Play pushes in that case.

That means a player can receive a favorable main-game settlement without ever comparing hand strength against the dealer. It also explains why simply computing “player hand beats random dealer hand” does not reproduce the actual game edge.

Pair Plus is a different probability-to-paytable problem

Pair Plus ignores dealer qualification and the fold/raise decision. It asks whether the player’s three cards fall into a paying hand category.

Under one widely analyzed schedule:

The published house edge for that exact schedule is about 7.28%.

The important words are that exact schedule. Other Pair Plus paytables exist, and the return changes when any payout changes. The underlying 22,100 hand frequencies stay the same while the price attached to each category changes.

That is why pair-based side bets and paytables explained should be read together.

A hit rate is not a return percentage

On the example Pair Plus schedule above, any pair-or-better hand occurs in:

48 + 52 + 720 + 1,096 + 3,744 = 5,660 hands

5,660 / 22,100 ≈ 25.61%

So the wager produces a qualifying pair-or-better result roughly one quarter of the time. That does not make its return 25.61%, because most of those hits are low-paying pairs and the remaining roughly 74.39% lose the stake.

Probability tells you how often an event occurs. Expected value tells you what the probability is worth after the paytable is applied.

Combining Ante and Pair Plus creates a blended cost, not a better game

Players often bet both because both circles are physically close on the felt. Mathematically, that creates two simultaneous contracts.

If you wager one unit on the Ante and one unit on Pair Plus, your expected loss per starting round is the sum of the expected losses of those two wagers, assuming the corresponding rules and strategies. The blended percentage depends on how you choose the denominator.

A useful session model is:

Expected session loss
≈ (Ante action × Ante-game edge on its chosen base)
 + (Pair Plus action × Pair Plus edge)
 + (Other side-bet action × its edge)

This is why a $10 minimum Three Card Poker table can behave like a $20, $30, or larger repeated decision once optional bets and Play wagers are included.

Paytable changes can matter more than small strategy errors

A player can memorize Q-6-4 perfectly and still choose an expensive table if the bonus or side-bet paytable is poor.

Likewise, comparing a 3.37% main-game figure with a 7.28% Pair Plus figure without reading the actual felt can be misleading if the casino offers a different schedule. Current regulated rules and approved variants can differ. Nevada’s current approved-games list includes multiple Three Card Poker-related versions and progressive products, which is enough reason to verify the version in front of you.

Read Three Card Poker odds in this order

For a clean analysis, ask the questions in sequence:

1. What exact rules and version are being played?
2. What is the three-card hand ranking?
3. What Ante Bonus paytable is posted?
4. What strategy threshold applies to the main game?
5. What denominator is used by the quoted house-edge percentage?
6. Is Pair Plus being played, and what is its paytable?
7. Are there progressive or six-card side bets that need separate math?

Only then does a number such as 3.37%, 2.01%, or 7.28% become meaningful. Three Card Poker is simple to deal, but its odds are a collection of separate probability and pricing problems sharing the same three cards.