Craps Strategy

Here's a dumb question:
If the house advantage on a pass line bet is over 1 percent, can't you just
place your money on the don't pass line and have an advantage?

No, because the 12 pushes.

Quote: amidnitespecial

Here's a dumb question:
If the house advantage on a pass line bet is over 1 percent, can't you just
place your money on the don't pass line and have an advantage?

The house advantage on the don't pass/don't come is very close to that on the pass/come, due to the fact that DP/DC does not win (or lose) on a 12, while the pass/come does lose.

The "perfect 1980" for the "dark side" is almost, but not quite, the mirror image of that for the pass/come.

Here it is for rightside betting:

result	ways	comment
comeout win	440
comeout loss	220	660 comeout decisions
win on 6	125
loss on 6	150
win on 8	125
loss on 8	150
win on 5	88
loss on 5	132
win on 9	88
loss on 9	132
win on 4	55
loss on 4	110	1320 point decisions
win on 10	55	536 point wins
loss on 10	110	784 seven-outs
	----
	1980

If you add up all the winning ways, you get 976, against 1004 losing ways. This yields a win probability of .492929, loss probability of .507071, and the difference between those is the (in)famous .01414, or 1.414%.

For betting the so-called dark side, here is the distribution of expected results:

result	ways	comment
comeout loss	440
comeout win	165
comeout push	55	660 comeout decisions
loss on 6	125
win on 6	150
loss on 8	125
win on 8	150
loss on 5	88
win on 5	132
loss on 9	88
win on 9	132
loss on 4	55
win on 4	110	1320 point decisions
loss on 10	55	536 point loss
win on 10	110	784 seven-outs
	----
	1980

If you add up the wins, you get 949, against 976 losses and 55 pushes. There are two ways to figure the HA, depending on whether you consider the push bets in the divisor. Personally, I do not, so I get 949 / 1925 = .492987 for win probability and .507013, for an HA of .014026, or 1.4026%. If you consider the push bets as "risked", you get 1.364%. You will see both figures quoted.

The essential information is that you can expect to lose 28 units out of 1980 bets on the pass line and 27 units out of 1980 bets on the don't. However, that is a long-term average expectation and subject to considerable variance, of course.
Cheers,
Alan Shank

Shank's response is right.

But to simplify matters:

The pass has a HA of 1.414%.
Therefore, the Don't pass should have a Player advantage of 1.414%.
So the casino says "aha, on a 12, you don't win (you lose on 12 on the pass), you push." And that happens on 1/36 of comeout rolls.

1/36 = 2.778%

1.414 PA - 2.778% = 1.364% House advantage on don't pass.

Quote: amidnitespecial

Here's a dumb question:
If the house advantage on a pass line bet is over 1 percent, can't you just
place your money on the don't pass line and have an advantage?

Its NOT a dumb question at all. Its very sharp of you to think of this. Its just that others have already thought of it and the casinos were real sharp too. So they made the "Don'tPass" bet be not quite the mirror image of the PassLine bet. That little "not quite" is the 12 is a Push wrinkle and that little wrinkle makes the house have an advantage on the DontPass bets.
I tend to like the Don't-side partially for that teensy little difference in the edge and partly because once I am over the hurdle of that very first roll... I'm rooting for a Seven which is the most likeliest number to be rolled.