Strategy guide

How to Play Medium Pocket Pairs in Criss Cross Poker

Pocket 6s through 10s start both hands with a pair that pushes rather than wins. We ran all 63,562,800 ways the board can fall to price exactly what that is worth.

  • 6-6
  • 7-7
  • 8-8
  • 9-9
  • 10-10
Across hand
Down hand
Your cards
  • 3x MAX Across
  • 3x MAX Down
  • 3x MAX Middle

A pair of 6s through 10s reaches the push line in both directions before a single community card turns over.

1

The Push Is Already Secured

Both hole cards play in both directions, so a pocket pair of 6s through 10s reaches the push line in the Across hand and the Down hand at once.

Every hand in Criss Cross Poker uses your two hole cards twice: once with the three horizontal community cards to make the Across hand, and once with the three vertical cards to make the Down hand. Hold two cards of the same rank and both of those hands begin with a pair already made. What happens to that pair is decided entirely by where the pay table splits.

Where a bare pair lands on the Across, Down and Middle pay table
Pair you hold Result
Jacks or better Pays 1 to 1
6s through 10s Pushes
5s or lower Loses

Five starting pairs sit in that middle band. Each one of them clears the losing line in both directions the moment it is dealt, and none of them clears the paying line.

  • At least a pair in the push range
  • At least a pair in the push range
  • At least a pair in the push range
  • At least a pair in the push range
  • At least a pair in the push range

No community card can remove the starting pair.

Your floor is a push, not a win

That single word is the whole difference between this page and jacks or better. With J-J, Q-Q, K-K or A-A the worst board you can be dealt still pays, so the ante cannot lose. With 6-6 through 10-10 the worst board returns your stake and nothing more. You are not protected from losing money on the hand, you are protected from losing it on the boards that miss you, which is a smaller promise and a genuinely useful one.

The board itself behaves identically for the two groups. Nothing about how often a pair improves changes with its rank. Only the line on the pay table moves.

The same boards, three different prices

Run the same exhaustive calculation for three starting pairs and the effect of that split is easy to read off. These are expected values per unit staked, over every board.

Expected value per unit staked, computed over every possible board
Starting pair Ante EV per unit Line bet EV per unit
J-J or better +1.000000 +1.580408
6-6 to 10-10 +0.281633 +0.862041
2-2 to 5-5 −0.436735 +0.143673

Read the middle row against the one above it. The medium pair keeps a clearly positive expectation on every wager, but it is a fraction of what the high pair earns, and the reason is one word on the pay table. That is the strongest argument on this site for reading the pay tables before you sit down.

2

Make the Maximum Wagers

There is nothing to weigh up. Post both antes, then take the maximum on all three lines.

  1. 1 Post both antes $10 $10
  2. 2 Maximum Across bet 3x
  3. 3 Maximum Down bet 3x
  4. 4 Maximum Middle bet 3x
Bet sizes offered 1x 2x 3x Betting options can vary between tables. Confirm the rules before you sit down.

The Middle bet is graded on the stronger of your Across and Down hands.

Why maximum is the right answer here

Because every one of these wagers has a positive expectation, and the ones you get to choose are the strong ones. Over every possible board, holding a pair of 6s through 10s, the ante is worth +0.281633 per unit. Each line bet is worth +0.862041 per unit. The Middle bet, which is graded on the better of your two hands and so collects whichever direction improves, is worth +1.485835 per unit. When a wager returns more than it costs on average, every extra unit you are allowed to put behind it adds to your expectation, so you take all of them.

Notice which number does the work. The ante is the weakest of the three at +0.281633, well short of the +1.000000 a high pair earns, precisely because the bare pair pushes instead of winning. The ante is also the one wager you cannot size: it is posted before you see a card. Everything you actually decide, the Across bet, the Down bet and the Middle bet, is worth far more per unit than the ante is, and the Middle bet is worth more than either line. So the whole of your decision runs one way, and it runs to the maximum. Its grading rule is set out in full on the Middle bet page, and the hand sits in the 3x band on the strategy chart alongside every other pocket pair.

Maximum means the table maximum on these three bets, which is normally 3x the ante. It does not mean raising your ante. Your ante size is a bankroll decision you make before the deal, and this hand cannot change it, because you have no way of knowing the hand is coming when you post.

3

A $10 Ante Example

Eleven units on the table, priced against every board the deck can produce.

Put the plan from section 2 into money. At a $10 ante you post two antes and take 3x on each of the three lines, which is the standard maximum exposure of eleven units.

Full maximum exposure at a $10 ante
Wager Amount
Across Ante$10
Down Ante$10
Across Bet at 3x$30
Down Bet at 3x$30
Middle Bet at 3x$30
Total wagered$110
  • +$101.93 Expected profit on this hand
  • $211.93 Average total returned on $110
  • 192.66% Return on the amount staked

Computed by evaluating all 19,600 three-card combinations per line and all 63,562,800 board arrangements. Nothing is sampled.

The ordinary outcome costs you nothing. A line finishes as just the pair on 71.8367% of boards. When that is what a line does, the ante and the bet on it push and the money comes straight back. If both lines finish that way, the Middle bet is graded on the better of two pushing hands, so it pushes as well, and all five wagers return your stake. That is the shape of this holding: the common result is flat, and the profit comes from the boards that improve one line or the other on the pay table.

This is one starting hand, not the game. None of this makes Criss Cross Poker beatable. The overall house edge is 4.33% on the ante, or 1.48% as element of risk. Both antes are posted before you see a card, so there is no way to wait for a good starting hand: you pay the full price of every other hand in order to reach this one. Maximum exposure in a single hand is 11 units, which is $275 at a $25 ante or $110 at a $10 ante. The Five Card Bonus is a separate wager at a 3.53% house edge and your hole cards never affect it. The full picture is on strategy and probability.

4

How the Pair Can Become a Paying Hand

The pair on its own only pushes, so every dollar this holding makes comes from the board. The five community cards give it two separate chances to grow, the horizontal three making the Across hand and the vertical three making the Down hand, with the centre card counted in both.

Take the showcase hand, . Across all 19,600 three-card boards one line can be dealt, it finishes in one of five ways. Four of them are improvements that pay. The fifth is the pair you started with, and it pushes.

Two pair

The board brings a second pair alongside your tens.

16.1633% of boards

Three of a kind

One community card matches your pair.

10.7755% of boards

Full house

A matching card plus another pair on the board.

0.9796% of boards

Four of a kind

Two community cards match your pair.

0.2449% of boards

Baseline

No improvement, just the pair

The board misses you completely and the pair stands on its own. A pair of 6s through 10s does not win here. It pushes, and the money on that line comes back.

71.8367% of boards

Those five outcomes cover every board there is, and they sum to exactly 100%. Four of them pay, and together they are 28.1633% of boards. They are also identical for 6-6, 7-7, 8-8, 9-9 and 10-10, and identical again for the high pairs: the rank of the pair changes nothing about how often the board improves it, only what the improvement is worth.

The ceiling: no straight, no flush

Some improvement lists for pocket pairs include a straight or a flush. That is wrong, and it is worth saying plainly. Your two cards share a rank, so a five-card hand built from them holds at most four distinct ranks, and a straight needs five. Your two cards are of different suits, so the hand holds at most four cards of any one suit, and a flush needs five. Neither is reachable, in either direction, on any board.

The pairs ladder is the whole of your upside, which is exactly why two pair and three of a kind carry so much of the value in this holding: between them they are 26.9388% of boards, and they are the first two rungs above a push. The full ladder is set out in the hand rankings guide.

A matching centre card improves both hands at once, because the centre belongs to the horizontal row and the vertical column.

That is why the centre card is worth watching more closely than the other four. It is the only card on the table that can lift the Across hand, the Down hand and the Middle bet in the same moment, because the Middle is graded on the better of the two lines and both lines run through the centre. A blank in the top or bottom position leaves one line pushing; a ten in the centre turns both lines into paying hands at once. The Middle bet guide covers how that shared card shapes the wager.

5

Across, Down and the Shared Centre Card

You are playing two five-card hands out of one pair, and they are not independent of each other. One card sits in both.

The Across hand is your pocket pair plus the three cards lying horizontally. The Down hand is the same pocket pair plus the three cards lying vertically. The card in the middle of the cross belongs to both of them, which is what makes the layout a cross rather than two separate rows. The full deal is walked through on how to play.

Across Hand
+
Pair plus the three horizontal cards
Down Hand
+
Pair plus the three vertical cards

The centre community card is counted in the Across hand and in the Down hand.

Shared card, correlated outcomes

Because one community card feeds both hands, the two results move together. A ten in the centre makes three of a kind in both directions at the same time. A blank in the centre leaves both directions leaning on the two cards they do not share. The lines are not two separate trials of the same experiment, and that is the reason you cannot price a round by taking one line’s five-card probabilities and doubling them. Doubling assumes independence, and there is none here.

It is also why the Middle bet needed the long calculation. The Middle is graded on the better of the two hands, so its value depends on the joint behaviour of Across and Down, not on either one alone. Pricing it meant enumerating all 63,562,800 ways the five community cards can fall into the cross: choose the centre first, then two horizontal and two vertical from what is left. There is no shortcut that gets the same answer, and the 19,600 single-line combinations, on their own, are not enough to produce it.

6

Keep the 5-Card Bonus Separate

The bonus is dealt in the same hand as your main bets, but it is settled on a different set of cards, so it belongs in a different column of your arithmetic.

The main wagers

  • Graded on your two hole cards plus three community cards.
  • Covers the Across, Down and Middle bets, plus both antes.
  • A medium pocket pair matters here: it guarantees a push in both directions and gives the board four ways to turn that into a payout.

The 5-Card Bonus

  • Settled on the five community cards only.
  • Your hole cards never contribute to it, not one of them, not ever.
  • A medium pocket pair does nothing for it. Pocket nines and a 7-2 offsuit are worth exactly the same to this bet: nothing.

Work out the value of the optional bonus separately. It shares a deal with your main bets, and nothing else.

There is a specific mistake to avoid here, and this holding invites it. Because a medium pair pushes so often, a lot of hands finish flat, and it is tempting to let a bonus that hit stand in for the profit the pair did not make, or to blame a bonus that missed for a hand that only pushed. Do not. The bonus carries a house edge of 3.53% and that number does not move when your hole cards are paired, so folding it into the value of a pocket pair overstates one and understates the other. Judge the bonus on its own terms, using the arithmetic in the 5-Card Bonus guide, and judge the pocket pair on the five wagers it can actually reach.

7

What the Exact Numbers Say

Every wager on a medium pocket pair, priced against every board that can possibly fall.

Most valuations of this holding stop at “a pair is better than no pair”. We priced it instead. For one line, the pair was evaluated against all 19,600 three-card combinations available from the 50 cards left in the deck. For the Middle bet, which is graded on the better of the two lines, the pair was evaluated against all 63,562,800 ways the five community cards can fall into the cross: the centre card first, then two horizontal and two vertical from what remains. Nothing was sampled and nothing was estimated. The figures below are the complete enumeration, and they are identical for 6-6, 7-7, 8-8, 9-9 and 10-10, because all five sit in the same push band on the pay table.

Every wager priced per unit staked, holding a pair of 6s through 10s
Wager Expected value per unit Note
Ante Across +0.281633 the bare pair pushes, it does not win
Ante Down +0.281633 the bare pair pushes, it does not win
Across bet +0.862041
Down bet +0.862041
Middle bet +1.485835 graded on the stronger hand
Two antes plus 3x on all three lines, 11 units +10.1930 units at 11 units
+$101.93 Expected profit at a $10 ante
192.66% Return on the amount staked
+10.1930 Units expected on 11 units risked

Correcting the Figures That Circulate

A particular set of numbers for this holding has been repeated widely enough that you will probably meet it. It does not match an exhaustive calculation. Here is what circulates, and here is what the full computation gives.

Figures that circulate

  • $68.44 profit per hand
  • 85.6% return
  • 72.5% of hands push

None of these match an exhaustive calculation.

What the full computation gives

  • +$101.93 on a $110 stake
  • 192.66% return on the amount staked
  • 71.8367% of boards, per line

Enumerated over every board, not sampled.

The push figure is worth separating out, because the wording matters more than the decimal. One line, Across or Down, finishes as just the pair on 71.8367% of boards. That is a per-line number, not a per-round one. A round contains two lines that share a centre card, so “how often does the whole hand push” is a different question with a different answer, and it cannot be read off the 71.8367% figure. Anything quoted as a share of hands rather than a share of boards per line is measuring something else.

Everything above describes what happens once you are holding a pair of 6s through 10s. It does not describe the game. Both antes go in before you see a card, so you cannot wait for this hand, and across every starting hand Criss Cross Poker still carries a 4.33% house edge on the ante, or 1.48% measured as element of risk. Nothing on this page changes that number. Be precise about which claim is which as well: a bare medium pair guarantees a push on a line, not a win, and the totals here are expected values over every possible board. Individual boards still vary. How the probabilities are worked out Where the low-pair split sits

Play within a fixed bankroll. Set the limit before you sit down, while the numbers are still abstract, and size the ante off that limit rather than off the table minimum. A hand worth betting to the maximum is only worth it if the maximum was affordable before the cards came out. Setting a limit that holds

Play a Few Pocket Pairs Free

The free table deals real hands with the same pay table, and the built-in coach shows the recommended action at each of the three decisions.

Criss Cross Poker is a table game from AGS, developed by Ron LaDuca and In Bet Gaming. Overall house-edge figures follow the published analysis at Wizard of Odds. The pocket-pair expected values on this page were computed with this site’s own engine, across all 19,600 three-card line combinations and all 63,562,800 board arrangements, with nothing sampled. This site is free to play, uses play money only, and accepts no wagers.