Craps - pass line + don't come

Here is an interesting strategy I have been playing with recently. It is slightly 'hedgy', but no so bad, and can be kind of fun. What you do is bet the pass line like always. When a point is established, bet the don't come. If another point is rolled, take odds on both the don't number and the pass line wager. What this does is move the bad number from 7 to whatever you rolled on the don't come. For example:
If the point is 6, and then you roll an 8 on the don't come, you win if you roll a 6 and loose if you roll an 8. A 7 is a wash. You route for the point like everyone else, but you don't want to see an 8. If one of those numbers is rolled, you just place another wager to get a new number.

Is there a name for this? I simulated this in wincraps and found that for one hour (120 rolls) the house edge is about the same as if you did a pass plus one come bet, but with double odds the variance was less for the pass /don't come case (about 20%), that is why I say it is slightly hedgy. (The actual sim results for this for a $5 pass line with double odds for one hour: the expected loss is $5 (+- $1) with a standard deviation of $100. I see $120 for simulations of the pass + 1 come with double odds)

Quote: DrEntropy

Here is an interesting strategy I have been playing with recently. It is slightly 'hedgy', but no so bad, and can be kind of fun. What you do is bet the pass line like always. When a point is established, bet the don't come. If another point is rolled, take odds on both the don't number and the pass line wager. What this does is move the bad number from 7 to whatever you rolled on the don't come. For example:
If the point is 6, and then you roll an 8 on the don't come, you win if you roll a 6 and loose if you roll an 8. A 7 is a wash. You route for the point like everyone else, but you don't want to see an 8. If one of those numbers is rolled, you just place another wager to get a new number.

Is there a name for this? I simulated this in wincraps and found that for one hour (120 rolls) the house edge is about the same as if you did a pass plus one come bet, but with double odds the variance was less for the pass /don't come case (about 20%), that is why I say it is slightly hedgy. (The actual sim results for this for a $5 pass line with double odds for one hour: the expected loss is $5 (+- $1) with a standard deviation of $100. I see $120 for simulations of the pass + 1 come with double odds)

The simple fact is that the more money you bet on negative expectation wagers, the greater your loss will be. Betting the Pass Line alone is better than betting twice that amount, split between the Pass Line and the Don't Come. The more you bet, the more you will lose. Period.

What you are doing is constructing an outcome where your results will be less "swingy" than otherwise. The penalty, of course, is that you will rarely obtain large wins. This is bad strategy for a negative EV game, since flattening out your results has the long-term effect of pretty much ensuring a moderate loss. If your objective was to not lose (rather than, to win), however, then the best strategy would be to stay away from the game altogether. The losing player (one who plays craps or any other -EV game) should EMBRACE variance, rather than avoiding it, because variance is the only thing that can save him.

Quote: mkl654321

The simple fact is that the more money you bet on negative expectation wagers, the greater your loss will be. Betting the Pass Line alone is better than betting twice that amount, split between the Pass Line and the Don't Come. The more you bet, the more you will lose. Period.

Agree 100%

Quote: mkl654321

The losing player (one who plays craps or any other -EV game) should EMBRACE variance, rather than avoiding it, because variance is the only thing that can save him.

I do embrace variance, don't get me wrong. I consider the fact that this betting scheme is less swingy to be a negative. When I said it was 'hedgy', read that as if I had said it had a bad odor :).
However, it is only slightly less swinging then my usual strategy (1 pass + 1 come, double odds) and the same expected loss. But it can be fun, a little something different to do! If I really wanted to maximize my variance I would just bet the pass line and bet max odds, but that is not as much fun (for me).

I'm not as big a fan of the variance. I like playing craps, rolling dice, getting drinks. It's a fun game. This strategy has the possibility of reducing your chance of blowing your bankroll as quickly.

I had tried out this strategy a couple months ago on the Wizard's craps game. I didn't feel like there was too much difference in the swings up and down. After all, money on the table will belong to the house eventually. But obviously playing a couple hours online won't be the same as a large number of simulations.

If it does have less variance, I may try it out next time I play for real.

I like this strategy, especially if playing for comps at a property that factors in odds bets. At such a casino, it would be a good way to get a good rating, at fairly low risk.

Quote: Wizard

I like this strategy, especially if playing for comps at a property that factors in odds bets. At such a casino, it would be a good way to get a good rating, at fairly low risk.

I like the strategy, too. If you make two offsetting bets every shooter at $5, your expected loss will be the same as you would making a line+come bet, with less variance. I don't play any crap tables over $5, so I could place most of the action on the odds. I've been playing with a pass with odds + 1 come with odds, which I thought would be low variance, but the swings are incredible; I can't stand it sometimes. So, I like your method for low stress and fun.

I re-read it, and realized it may differ from what I had tried.

So you don't take odds on the pass line right off the bat? You wait for the roll after you bet Don't Come? I'm not great at doing this stuff in my head. Does this create a house edge on the odds, by waiting one roll? Or does it not actually make a difference?

Quote: dudestupid

So you don't take odds on the pass line right off the bat? You wait for the roll after you bet Don't Come? I'm not great at doing this stuff in my head. Does this create a house edge on the odds, by waiting one roll? Or does it not actually make a difference?

I prefer to wait on the odds bet until after the the Don't come, so that I am not so exposed to the seven on that roll. On the down side, if the shooter rolls his point right away, you would regret not having the odds up. :) It doesn't really make much of a difference though.... it makes no difference to the expected value, and has negligible effect on your standard deviation.

Quote: DrEntropy

Here is an interesting strategy I have been playing with recently. It is slightly 'hedgy', but no so bad, and can be kind of fun. What you do is bet the pass line like always. When a point is established, bet the don't come. If another point is rolled, take odds on both the don't number and the pass line wager. What this does is move the bad number from 7 to whatever you rolled on the don't come. For example:
If the point is 6, and then you roll an 8 on the don't come, you win if you roll a 6 and loose if you roll an 8. A 7 is a wash. You route for the point like everyone else, but you don't want to see an 8. If one of those numbers is rolled, you just place another wager to get a new number.

Can you give a better example?
Since a player "takes odds" on the pass line and "lays the odds" on the don't side, what amounts do you use?

Your example:
"For example:
If the point is 6, and then you roll an 8 on the don't come, you win if you roll a 6 and loose if you roll an 8. A 7 is a wash. "

Only a "wash" if you have not taken any odds on the pass line and "laid" the odds on the don't come bet.
A point-Seven causes both bets to lose.
You say you "take odds" on both bets. Do you take the odds right after the dont come goes to a number?
How can it then be a wash?

I was trying to code your system of play also in WinCraps but you left out many of the needed values.

A better example would be:
On the come out roll bet $10.
Comeout roll is a #6.
While . . .
A point is established on any number
then . . .
Bet $ 10 on Don't Come
Next roll is a #8
While . . .
Don't Come 8 is greater than $ 0
Total Don't Come bets are equal to ($)1
then . . .
Bet $ 10 on Pass Line Odds
Bet $ 12 on Don't Come 8 Odds

Quote: nope27

Can you give a better example?
Since a player "takes odds" on the pass line and "lays the odds" on the don't side, what amounts do you use?

Your example:
"For example:
If the point is 6, and then you roll an 8 on the don't come, you win if you roll a 6 and loose if you roll an 8. A 7 is a wash. "

Only a "wash" if you have not taken any odds on the pass line and "laid" the odds on the don't come bet.
A point-Seven causes both bets to lose.

I can send you my script later (via PM) when I get home.
But yes, point-seven causes both bets to lose, so I wait and take odds after the don't pass gets a number as well. Let me give a detailed example:
Example 1:
Bet $5 on the pass line.
Roll: 6
bet $5 on the don't come.
Roll: 7
Result: Lose $10

Example 2:
Bet $5 on the pass line.
Roll: 6
bet $5 on the don't come.
Roll: 9
Now i have two numbers, one I want and one I don't :) Take $10 odds on the 6. Lay the $15 odds on the 9.
Now if a 7 rolls, you loose 15 on the pass line + odds, and you win 15 on the don't come +odds.. start over. If a 6 rolls, you win $17, and place a new pass line bet. If a 9 rolls, you lose $20 and place a new don't come bet.
Cheers!

Quote: DrEntropy

I can send you my script later (via PM) when I get home.
But yes, point-seven causes both bets to lose, so I wait and take odds after the don't pass gets a number as well.

Why not flip them? Make the don't pass bet first, then the come bet. Then point-seven causes both bets to win. That's good on two fronts: one, you win two bets while everyone else is grumbling, and two, you're in exactly the same position otherwise after the first two numbers. Then you can just do what you'd normally do.

Quote: mkl654321

The simple fact is that the more money you bet on negative expectation wagers, the greater your loss will be. Betting the Pass Line alone is better than betting twice that amount, split between the Pass Line and the Don't Come. The more you bet, the more you will lose. Period.

This, of course, is not true. Expected value (loss) is not the same as results. You cannot lose 14 cents on a bet (the ev for a $10 pass bet). The ev is a theoretical figure that, in the case off the line bets, derives from the slight difference in the probability of winning or losing the bet, since the payoff is even money. This betting strategy, like many, is "chasing a pattern", and if the pattern occurs, the player will win.

Quote: mkl654321

What you are doing is constructing an outcome where your results will be less "swingy" than otherwise. The penalty, of course, is that you will rarely obtain large wins. This is bad strategy for a negative EV game, since flattening out your results has the long-term effect of pretty much ensuring a moderate loss.

This is true, but your defining it as "bad" ignores the desires/needs of the individual gambler. A player who values "table time" and finds this method more interesting or fun than just the pass line might understand your point but not consider it "bad", for him/her.

Quote: mkl654321

If your objective was to not lose (rather than, to win), however, then the best strategy would be to stay away from the game altogether. The losing player (one who plays craps or any other -EV game) should EMBRACE variance, rather than avoiding it, because variance is the only thing that can save him.

Again, your presume to decide for others what they SHOULD do. Of course, not to play guarantees you will not lose, but what fun is that? The intelligent player, in my view, understands why he/she is playing and what he/she wants from it and devises a strategy that maximizes his/her chance of "success", as defined only for that individual.

For myself, I like to lower the ev/SD ratio by playing pass only and taking odds, but at the same time having a low probability of busting my bankroll within a couple of hours' play.

It's true that any hedge bet is going to increase the expected loss, because you're making another bet with a negative ev, but it can also reduce the magnitude of losses when luck is bad. Some people may value that. I don't. In my view, playing the straight minimum-bet pass line minimizes expected loss and also has very low variance. For a couple of hours' play (60 bets) at $5 a pop, your most likely outcome is breaking even, you have about a 40% chance of coming out ahead and you're pretty unlikely (probability .058) to lose $70 or more . But, yeah, many people consider that boring, and that's part of the decision, too.
Cheers,
Alan Shank
Woodland, CA

Quote: MathExtremist

Why not flip them? Make the don't pass bet first, then the come bet. Then point-seven causes both bets to win. That's good on two fronts: one, you win two bets while everyone else is grumbling, and two, you're in exactly the same position otherwise after the first two numbers. Then you can just do what you'd normally do.

Right.. I started with the don't pass + come bet , but like rooting with the shooter so flipped it. It actually doesn't matter so much in terms of the math.

Quote: DrEntropy

I can send you my script later (via PM) when I get home.

That would be fine if you would like but I prefer to do my own coding.
Seeing how others code just makes me ask at most times "why did you do it that way?"

Quote: DrEntropy

But yes, point-seven causes both bets to lose, so I wait and take odds after the don't pass gets a number as well. Let me give a detailed example:
Example 1:

Example 2:

Now i have two numbers, one I want and one I don't :) Take $10 odds on the 6. Lay the $15 odds on the 9.
Now if a 7 rolls, you loose 15 on the pass line + odds, and you win 15 on the don't come +odds.. start over. If a 6 rolls, you win $17, and place a new pass line bet. If a 9 rolls, you lose $20 and place a new don't come bet.
Cheers!

Thank you for the examples.

Quote: nope27

That would be fine if you would like but I prefer to do my own coding.

Ok, but please do share your sim results!
Cheers!

Quote: DrEntropy

Ok, but please do share your sim results!
Cheers!

I am visiting the US this week and will be traveling the next 6 days before my return to Spain.
You have already ran simulations and have posted some results in your earlier posts.

Why would my results show anything different from yours?

I do run 10,000 to 20,000 sessions at a time to get overall results, but they rarely differ from shorter length simulations.
Let me know if there is something specific that you would like to see and I will share that with you.

deleted, I had to hash out how it was different from the usual doey-don't

Do you take the Pass Odds down for the one roll when you put up the new DC bet, like you do right after the point is established?
The same risk is there, where a 7 would cause you to lose both the PL and DC bets......

Quote: RaleighCraps

Do you take the Pass Odds down when you put up the new DC bet, like you do right after the point is established?
The same risk is there, where a 7 would cause you to lose both the PL and DC bets......

Yes. I also like to do this when I bet standard Pass + 1 come bet.

Quote: DrEntropy

Example 2:
Bet $5 on the pass line.
Roll: 6
bet $5 on the don't come.
Roll: 9
Now i have two numbers, one I want and one I don't :)

Funny...

In truth, you do want both numbers to win like...
Point 6 winner (5/11) chance of happening

followed right after with a 7 winner. (1/6) chance of happening. Not too much to ask for. (5/66)

Now you have won 3 bets and you are off to the races.

If nope27 does not want to see your WinCraps auto-bet file, I would like to see it.
I can code it to run multiple simulations and have it "spit out" to a text file any stats that you would like to see.
Let me know.