Ask a counter what a double deck game feels like next to a six deck shoe and the answer is a metaphor: the count is livelier. Metaphors hide the real questions, which are how much livelier, measured where, and at what cost in precision. The difference is arithmetic that can be simulated, and it touches three skills: converting the count, estimating the tray, and the basic strategy chart itself.
To put numbers under the metaphor, we ran a simulation: 200,000 shoes each for two decks and six decks, counted with Hi-Lo, where 2 through 6 count plus 1, 7 through 9 count 0, and 10 through ace count -1. The running count was tracked card by card, and the true count is the running count divided by the exact decks remaining, unrounded, measured at the moment the given share of the cards has been dealt. That is the same method as our walk through how the count moves through a shoe, extended to two decks, with the division covered in our piece on running count versus true count.
The numbers, one table
Here is what the two simulations looked like at four depths. The spread columns are standard deviations, computed from the exact formula rather than the simulation: the running count's spread after k cards of an N card shoe is the square root of k x 40/52 x (N minus k) / (N minus 1), and the true count spread divides that by the decks remaining. The share columns are the percentages of simulated shoes in which the true count sat at or beyond each level at that exact moment.
| Share dealt | Two decks (104 cards) | Six decks (312 cards) | ||||||
|---|---|---|---|---|---|---|---|---|
| TC spread | +2 or higher | +3 or higher | -2 or lower | TC spread | +2 or higher | +3 or higher | -2 or lower | |
| 25 percent | 2.59 | 26.1 | 12.5 | 26.0 | 1.49 | 10.2 | 2.2 | 10.3 |
| 50 percent | 4.49 | 37.0 | 29.0 | 36.9 | 2.59 | 23.9 | 13.7 | 24.0 |
| 65 percent | 6.18 | 36.3 | 28.0 | 36.3 | 3.53 | 27.1 | 19.0 | 27.2 |
| 75 percent | 7.78 | 45.0 | 35.1 | 44.8 | 4.48 | 35.4 | 25.0 | 35.5 |
At a quarter dealt, a six deck shoe sits at a true count of +2 or higher in 10.2 percent of simulations; a two deck game does in 26.1 percent, between two and three times as often. At +3 or higher the gap widens further: 2.2 percent against 12.5 percent, close to six times as often. The minus columns mirror the plus columns almost exactly, because Hi-Lo tags as many cards plus 1 as -1, so a shoe that drifts high is as likely as one that drifts low.
The same depth, a wider spread
The relationship between the two games comes straight from the formula. At the same share dealt, the two deck true count spread is the six deck spread times the square root of 311/103, which is about 1.74, and the table carries it up to rounding: 1.49 times 1.74 is 2.59, and 2.59 times 1.74 is 4.51 against the table's 4.49. The 65 percent row sits a little off the ratio, because it uses 68 of 104 cards and 203 of 312, not quite the same share. The divisor, the decks remaining, is small enough that the same running count is a much louder signal.
The speed shows up in the steps. One point of running count moves the true count by one divided by the decks remaining. In two decks that is 0.67 at a quarter dealt, 1.00 at half and 2.00 at three quarters dealt; in six decks the same three figures are 0.22, 0.33 and 0.67. Late in a double deck game, single cards move the true count by whole points. This is the concrete version of the dilution principle our piece on counting more decks describes: each card carries more information when fewer cards remain around it, and the two deck game has fewer cards remaining at every share dealt.
One footnote on the numbers. Our published six deck table reported 13.6 percent for a true count of +3 or higher at the halfway point; this run gives 13.7. That gap is simulation noise, the kind of wobble our piece on how counting numbers are tested exists to explain.
Why the share columns wobble
The spread columns climb smoothly with depth, but the share columns do not. The two deck +2 column, 26.1, 37.0, 36.3, 45.0, dips at 65 percent dealt before rising again; the six deck column shows no such dip. The running count moves in whole steps of one, while the divisor, the decks remaining, shrinks with every card. A true count of +2 needs a running count of twice the decks remaining, and as the divisor changes, whole running counts slide in and out of reach of each threshold. In a small shoe those steps are large relative to the divisor, so the share at a given depth depends on exactly where the whole numbers land. The six deck divisor is big enough to average the effect away.
The trend across depths is real and so are the dips, so read each column as a whole rather than one cell in isolation.
21 TrainerFree on iPhone and Android. Charts that redraw for the selected rules, with instant feedback and an explanation on every hand.
Get the app
What this changes for the counter
Three practical consequences follow. The first is deck estimation. The true count divides by the decks remaining, and the same error hurts more as the denominator shrinks: a half deck error at one deck remaining is a 50 percent error in the divisor, far larger in proportion than the same half deck at four decks remaining. Our guide to estimating decks remaining treats the tray read as a skill with its own drills, and in a double deck game it decides whether the livelier count is even being measured correctly. The related habit, what to do with the fraction that division leaves behind, is the subject of our piece on rounding conventions.
The second is penetration. How deep the dealer deals is not a rule printed on the felt but a piece of dealing procedure, expressed as the share of the shoe dealt or as decks cut off, per our penetration explainer. In a double deck game the same percentages describe far fewer cards, and the deepest rows of the table above, where the two deck true count spreads widest, exist only when the deal gets there.
The third consequence is that counting a two deck game is still counting. The minus columns are as wide as the plus columns, and nothing in the arithmetic removes risk from the game. Counting with your head is legal in the United States, and casinos are private businesses that may refuse service to anyone they suspect of it. Our pieces on when counting is not worth it and on its legality are honest companions to this table.
The chart changes too
The two deck game also plays from a slightly different basic strategy chart, as our overview of what changes with the rules explains. In the app's charts, with the dealer standing on soft 17, the two deck game differs from the four to eight deck game in exactly these cells: hard 9 against a 2 doubles, hard 11 against an ace doubles, hard 16 against a 9 hits instead of surrendering, 6,6 against a 2 is a plain split rather than a split that depends on doubling after split, and 6,6 against a 7 and 7,7 against an 8 are splits if double after split is allowed, otherwise hits.
With the dealer hitting soft 17 the list shifts: hard 9 against a 2, hard 16 against a 9, soft 14 against a 4 doubling, 6,6 against a 2 and against a 7, 7,7 against an 8, and 8,8 against an ace, which surrenders only when double after split is not allowed and splits otherwise. Everything else on the chart is the same, the pattern our piece on single deck versus many traces: fewer decks move a handful of borderline cells, not the chart's logic.
A counter learning deviations has two moving parts here: a different base chart, and a true count that moves faster than it does in shoe practice. 21 Trainer's Double Deck preset sets one such game, two decks with the dealer hitting soft 17 and doubling after split allowed, and the Pro upgrade includes a true count drill alongside its four counting systems, so the conversion can be practiced with no money in play.
Frequently asked questions
Is card counting easier with two decks than six?
It is not easier to keep, but the true count moves more. At the same share of the shoe dealt, the true count in a two deck game spreads about 1.74 times as wide as in a six deck game, so high and low counts arrive sooner and more often, and the count moves in both directions equally. The work is not trivial: the running count still has to be kept perfectly, and deck estimation has to be finer.
How much faster does the true count move in a double deck game?
One point of running count moves the true count by one divided by the decks remaining. In two decks that is 0.67 at a quarter dealt, 1.00 at half and 2.00 at three quarters. In six decks the same steps are 0.22, 0.33 and 0.67. Late in a double deck game, single cards shift the true count by whole points at a time, which is why the count there is described as faster and further.
Does basic strategy change in a double deck game?
Yes, in a small set of cells. When the dealer stands on soft 17, the two deck chart differs from the four to eight deck chart in the hard 9 against a 2, hard 11 against an ace, hard 16 against a 9, and the 6,6 and 7,7 rows. When the dealer hits soft 17, the list changes slightly again, including soft 14 against a 4 and 8,8 against an ace. The rest of the chart is identical.
Is counting cards in a double deck game legal?
Counting with your head is legal in the United States. Casinos are private businesses that may refuse service, and a suspected counter can be backed off or banned regardless of the deck count. No technique removes risk from the game, and nothing about a two deck game changes any of that.
Do you need better deck estimation for a double deck game?
Yes, because the same estimation error is a larger share of a smaller denominator. A half deck error with one deck remaining is a 50 percent error in the divisor, while the same half deck error with four decks remaining is far smaller in proportion.