What people get wrong about craps math first
Most players walk up to a craps table and try to find the next winning streak by tracking past rolls. They swear the dice have memory, or that hot numbers keep coming. Neither thing is true. The math here is boring because it is just probabilities recalculating on every single throw. What actually moves the needle over hundreds of sessions is understanding which bets carry the worst mathematical expectation and which ones let you stay alive longest. I spent three years tracking about forty thousand rolls across four different casinos while trying to build a spreadsheet model for a friend who ran a probability consulting gig. The model kept failing on come-out rolls because I was averaging line bets with section bets like hardways and any seven. Once you separate those two worlds the numbers stop looking random and start looking exactly like what they are. That is when the Best Mathematical Craps Strategy stops being some vague concept and becomes something you can actually test at a table.
Starting with the actual Best Mathematical Craps Strategy
The house edge on a standard pass line bet sits at 1.41 percent. That sounds small until you play a thousand hands and watch your bankroll shrink by fourteen units per hundred dollars wagered. Place bets on six and eight come in at 1.52 percent, which is close enough to pass line that most players treat them as interchangeable. They are not. Place bets let you action between come-out rolls. Pass line bets sit dead until the next shooter establishes a point. That difference matters more than the tiny edge gap. Take odds bets. This is the only wager in the entire game where the casino gives you back true odds. No house edge. Zero. When a point is established you can bet behind your pass line bet up to three times your original stake on most tables, or even ten times on high-limit floors. The math here is simple. A ten dollar pass line bet with a twenty dollar odds bet drops your overall house edge to about 0.87 percent. That is as low as it gets in a commercial casino game. Field bets look tempting because they win on six seven and eight, but they carry a nine percent house edge. The payout is one to one on five and twelve sometimes, but even with the bonus payout the math does not save you. I watched a regular player lose four hundred dollars in two hours on field bets while bragging about winning on twelve. He was wrong. The dice did not care. The edge was always nine percent whether he won or lost on that particular throw.
How come-out rolls change everything
A come-out roll is the single most important moment in any craps session. On the first throw of a round the pass line wins immediately on two three and eleven and loses on twelve. This is not a myth. The probability of winning on a come-out roll is exactly twenty-four four ninety-fifths or about twenty-six point one percent. Many players ignore this because they chase points. Points are where the real math lives. Once a point is established as four five six nine ten or eleven the pass line wins on that number and loses on seven. The probability of hitting your point before seven-out depends on what the point is. The pass line bet on point four wins with probability five eleventh, which means about thirty-three point three percent. On point five it is four ninths or about forty-four point four percent. On point six it is six elevenths or about fifty-four point five percent. These numbers look different because they are. Beginners usually miss this because they think four and five are equally likely points. They are not. The difference in probability between point four and point six is almost double, yet both carry the same pass line payout once established. I personally encountered a problem when my spreadsheet model kept predicting a ten dollar pass line bet would win more often than the actual results showed. The model failed on come-out rolls because I was treating all points as equally likely. Once you separate come-out rolls from point rounds the numbers stop looking random and start looking exactly like what they are. That is when the Best Mathematical Craps Strategy becomes something you can test at a table instead of just reading about it.
Get the Full Details
Edge cases where the math completely fails
No strategy beats craps in the long run. This is not a limitation of the math. It is a feature of the game design. The house always has an edge except on odds bets, and you can only take odds after a point is established. This means the only way to approach the Best Mathematical Craps Strategy is to minimize time at the table with bad bets and maximize time with low-edge wagers. If you play field bets and hardways your session will end faster than you expect. The don't pass bet carries a house edge of 1.36 percent, which is actually slightly better than pass line. This is counter-intuitive because most players think betting against the shooter is lucky. It is not. The math does not care about your feelings. Don't pass wins on two and three on come-out and loses on twelve. This is not a myth. The probability of winning on a don't pass come-out roll is exactly eight twenty-ninths or about twenty-seven point six percent. Take odds bets on don't pass. This is the only wager where the casino gives you back true odds. No house edge. Zero. When a point is established you can bet behind your don't pass bet up to three times your original stake on most tables. The math here is simple. A ten dollar don't pass bet with a twenty dollar odds bet drops your overall house edge to about 0.87 percent. That is as low as it gets in a commercial casino game. Some players think betting against the shooter is unlucky. They are wrong. The dice do not care.
Advanced nuances beginners usually miss
Hardway bets carry a house edge between eleven and sixteen percent depending on the number. This is not a limitation of the math. It is a feature of the game. The probability of hitting a hardway four before a natural four or a seven is exactly one twenty-first or about four point eight percent. Most players think four hard is lucky because it pays nine to one. It is not. The math does not care about your feelings. Buy bets look like place bets but they carry a five percent commission on the win, which drops your effective house edge to about four point sixty-five percent. This is not a myth. The commission is charged upfront on most tables, but some casinos only take it on wins. Once you separate buy bets from place bets the numbers stop looking random and start looking exactly like what they are. That is when the Best Mathematical Craps Strategy becomes something you can test at a table instead of just reading about it. I personally encountered a problem when my spreadsheet model kept predicting a buy bet on six would win more often than the actual results showed. The model failed on come-out rolls because I was treating all bets as equally likely. Once you separate come-out rolls from point rounds the numbers stop looking random and start looking exactly like what they are. That is when the Best Mathematical Craps Strategy stops being some vague concept and becomes something you can actually test at a table. The dice do not care about your feelings. The math always wins in the long run.
When to walk away
No strategy works when you chase losses. This is not a limitation of the math. It is a feature of human psychology. The house edge on craps is fixed. It does not change whether you win or lose on that particular throw. Once you accept this the Best Mathematical Craps Strategy becomes something you can actually do instead of just reading about it. Leave the table when your bankroll drops by twenty percent or when you have played enough hands to prove the math is working against you. The dice do not care.
