Two Dice Combo: How Many Combinations?
Thirty-six. That is the hard number every pit boss, table game designer, and advantage player must tattoo into their skull before they touch a single dice layout. Two standard six-sided dice produce exactly 36 combinations, and every payout structure, house edge, and side-bet trap in the casino ecosystem is built on that foundation. Get this wrong and your math model collapses before the first roll.
Break it down mechanically. Die A shows 1 through 6. Die B shows 1 through 6. The sample space is 6 × 6 = 36 ordered pairs. Not 21. Not 12. Thirty-six. The 21 figure is the count of unordered outcomes, and half the amateur analysts on the planet confuse the two and then wonder why their dice probability tables bleed money. Ordered pairs matter because the dice are physically distinct objects — a 3 on the red die and a 5 on the blue die is a different microstate than a 5 on red and a 3 on blue, even though the sum is identical.
The Sum Distribution Nobody Respects
Here is where the real technical meat sits. The 36 combinations do not distribute evenly across sums. Seven is the king with 6 ways to land (1-6, 2-5, 3-4, 4-3, 5-2, 6-1). Six and eight each carry 5 ways. Five and nine carry 4. Four and ten carry 3. Three and eleven carry 2. Two and twelve are the dead-ends with a single combination each. That distribution curve is the entire engine behind craps pass-line odds, and any operator who flattens it into uniform payouts is handing sharp players a free lunch 🎲.
Run the percentages. Seven hits at 6/36, roughly 16.67 percent. Snake eyes and boxcars each sit at 1/36, about 2.78 percent. The gap between those extremes is where the house margin lives and dies. A field bet paying even money on a 2 or 12 is a sucker's proposition precisely because the true two dice combinations math says those outcomes are three times rarer than a seven.
Where Combination Counting Breaks Down in Live Play
Physical dice are not perfect random number generators. Weight distribution, edge wear, and throwing technique introduce bias that shifts the theoretical 36-combination grid. I have audited tables where a worn die favored one face by enough to move the seven frequency off its 16.67 percent anchor by a measurable fraction — and that fraction, compounded over thousands of rolls, is pure profit or pure bleed depending on which side of the bet you sit. Precision dice with sharp edges and balanced cores exist for exactly this reason. The 36-combination model is the baseline, and every deviation from it is a signal.
Combinatorial Counting Beyond Two Dice
Scale the logic and the formula holds: n dice with s sides give s to the power of n total combinations. Three six-sided dice explode to 216 outcomes. Four dice hit 1,296. The sum distribution tightens toward the center as you add dice, which is why multi-dice side bets with fat multipliers are almost always priced against a player who never ran the combinatorial count. Know the 36. Respect the 36. Every payout table in the building is a negotiation against that number, and the house never negotiates blind 😤.