House Edge on Pass/Don't Pass

Hi Folks,

I have been trying to do a little research on this and I'm sure it is something that you guys have considered already so wanted some input.

I'm looking at playing the pass and don't pass, no odds - and trying to figure out the edge. I had read (here) that it was assumed that the house edge would be 2.77% (1/36) however I feel like this might not be accurate.

While you would bar the 12, if that occurred you would only lose 50% of your bet.

Question 1: Should severity of loss be considered when calculating the risks of this strategy, because losing half isn't the same as losing all...
Question 2: Stating a 2.77% advantage doesn't factor in that the 12 has no impact once a point is established, as such wouldn't a more accurate house edge factor in how many rolls would be come out vs point established and use that as a multiple?

It just seems that if we had another hypothetical table next to a regular table where the bet was lost 100%, any time that a 12 was rolled regardless of point establishment that it would be a different game and have different house edge....

Thanks for entertaining my question!

Cheers,

Matt

https://wizardofodds.com/games/craps/
Welcome to the forum Matt,
I suggest checking out the Wiz's website 1st, your numbers seem way off.
Afterwards, you can get alot of help here.

Thanks for the welcome!

Maybe I didn't explain it very clearly... this post discusses the issue and raises the 2.77% number that I thought was off:


The only difference is the full odds aspect but I consider that to be a wash.

Quote: gamblingmatt

Thanks for the welcome!

Maybe I didn't explain it very clearly... this post discusses the issue and raises the 2.77% number that I thought was off:

https://wizardofvegas.com/forum/gambling/tables/1432-craps-betting-pass-and-dont-pass-and-full-odds-on-pass/

The only difference is the full odds aspect but I consider that to be a wash.

I missunderstood your question,
I'm thinking casino advantage would be average of 1.41 and 1.36 for a 1.385 casino advantage

Quote: terapined

I missunderstood your question,
I'm thinking casino advantage would be average of 1.41 and 1.36 for a 1.385 casino advantage

The edge on the don't pass is 1.36% per bet resolved (which includes a tie as a "resolution").

https://wizardofodds.com/games/craps

It's also 0.40% per roll on average. I prefer to look at edge per roll just to keep everything "normalized" into the same domain. Also, I think the edge per roll is slightly under 0.40% and it's just rounded up but I may not be remembering that correctly.

The only non-zero house edge bet that is lower per roll is the buy bet on the four or ten for $25 or $50 with vig paid on the win. That can go as low as 0.3333% in an unusual situation where the vig payment is actually rounded down instead of up. Something that is relatively unusual.

Thanks for this, and sorry for maybe being a bit dumb about it, but it still seems like we could be looking at different numbers.

I was trying to find a simulator that would allow me to play both sides and see what came up, do you know of any? I tried to get wincraps but I couldn't figure out intuitively if it had a monte carlo type sim on it.

Quote: gamblingmatt

Thanks for this, and sorry for maybe being a bit dumb about it, but it still seems like we could be looking at different numbers.

I was trying to find a simulator that would allow me to play both sides and see what came up, do you know of any? I tried to get wincraps but I couldn't figure out intuitively if it had a monte carlo type sim on it.

If you bet both sides, on average 29% of all rolls are comeout rolls.

2.78% of all comeout roll are twelve.

You lose 1/2 of your bets when the twelve rolls on a comeout roll.

0.29 * 0.0278 * 0.5 = 0.004031 = 0.4% edge per roll.

There are lots of ways to see the edge per roll. But I don't know if that helps to know that every 1/(.29*.0278) rolls (124 rolls) you lose half of the money you have (the pass line) when you bet both sides.

?

Compare to roulette -- 2/38 rolls are 0 or 00. You lose both (100%) of red and black when you bet both red and black when 0 or 00 comes up. Every other time you don't lose anything.

2/38 = 0.05263 = 5.263% edge per spin.

Total percentage cost difference between line bets in craps and double zero roulette (per random event which is defined as the roll or the spin) is 7.6589 times as high for double zero roulette than craps.

That's brilliant, thanks - was thinking that we needed the come out roll % too for this, but didn't realize that got it to the 0.4% much appreciated!

Giving this a little more thought, thanks to your pointers. Would it be the same result if we changed the math to more specific:

Chance of a 12 rolling after a 7 (with the roll of a 2,3,11 to infinity in between the two events being counted too?

EG

7,12 (event) which would be
7,2,2,3,11,2,11,2, (non point) 12)

etc...

EDIT - just worked out the 7,12 combo would be 1/216, which is 0.4% neat :)

EDIT again.. but that doesn't factor in impact. while 1/216 for a 7/12 combo is the correct probability of a loss, the impact of a loss should be factored in again, because it would be unfair to compare equally the house edge of losing 100% of your bet with losing 50%. So should it be drawn down to 0.2% for this?

For the statisticians out there, additionally, is there a difference between comparing a 2 roll 7,12 with probabilities of multiple rolls of craps/yo that create a 7 [craps,yo] 12 scenario?