I looked around but couldn't find anything online about it. I tried to figure out the edge myself and am coming up with -0.707075916.. I hope my calculations are way off...
The card has a terrible explanation of how the bet works, and I'm not sure I can do much better, but it's kind of a combination of the All Small/All Tall/All or Nothing at All bets and the Fire bet. The 6 box/point numbers are marked on the layout, along with the dice combinations that make up that number (as shown on the card.) The box person is responsible for marking each combination.
So 4 is divided into 1-3 and 2-2, 5 is 1-4 and 2-3, 6 is 1-5, 2-4, 3-3.. and the same for 8, 9, and 10. The dice have to hit every combination of a number, and then that number is marked as "complete." Any 7 (seven-out or come-out natural) resolves the bet. You must "complete" at least 2 numbers to get a payout. The pay table is:
Completed numbers | Payout |
---|---|
<2 | -1 |
2 | 5 to 1 |
3 | 10 to 1 |
4 | 20 to 1 |
5 | 50 to 1 |
6 | 200 to 1 |
I played for about 2.5 hours and saw 2 numbers get hit twice (nobody bet it these times) and never anything more. This bet was receiving almost no action. One dealer said earlier in the day all 6 had been hit on a $5 bet for a $1000 payout.
Here are images of the card. Sorry for the quality; I put it in my pocket and forgot about it so it's a bit beat up. The min/max was 1-25.
Front: http://tinypic.com/r/10xrbjq/5
Back: http://tinypic.com/r/10hq9eg/5
Completed Points | Pays | Probability | Return |
---|---|---|---|
6 | 200 | 0.000412 | 0.082424 |
5 | 50 | 0.002219 | 0.110966 |
4 | 20 | 0.007519 | 0.150382 |
3 | 10 | 0.021192 | 0.211924 |
2 | 5 | 0.056294 | 0.281471 |
1 | -1 | 0.160625 | -0.160625 |
0 | -1 | 0.751738 | -0.751738 |
Total | 1.000000 | -0.075196 |
So, house edge of 7.52%.
and no w2-g if you hit the top bet
I was scratching my head trying to come up with formulas. I'm reassured to see that Mike had to do a simulation. And nice to see that the edge isn't unreasonable.Quote: WizardHere are my results based on a quick simulation.
The Fire bet only pays out 1.1% of the time, but the 999:1 payout for all 6 is a lot more alluring than 200:1.
I was also thinking about ways you could hedge this bet, and could only come up with one thing that puts it into JP territory. Hope for easy 4 or easy 10 to be the last dice combination to complete all 6 numbers, lay the number to hedge the hot roller bet, and then hedge the lay with a hardway bet. LOL! If the easy way comes, you lose the lay and the hard way, but win all 6 on the hot roller. 7 comes you win the lay but lose the hardway, and still get paid for 5 completed numbers on hot roller. Hard way comes you lose the lay but win the hardway, and then replace the lay and start over.
Sorry you had to see it.Quote: wudgedI visited Dover Downs casino yesterday and saw a new bet called Hot Roller Craps.
I looked around but couldn't find anything online about it.
Maybe In Bet Gaming is not so proud of it. or maybe they are and they think this will dominate Craps tables all over the world?
I played this in WinCraps for a NBA basketball game duration.
Crap.
The bet has nothing to do with the game of Craps or even popular bets that can be made in Craps..
Is not exciting or even fun at all, for me,
I mean who cares a 4 rolls 1-3 or 2-2 unless the hardway is bet.
and someone somewhere got a kick out of having 14 combinations of the dice roll before a 7.
well, good for them
My $400 Buy4 could not care less ;)
This is an awful bet IMO
and it forces box to track it. that is really great if that is true.
would probably do much better as a bet in Sic Bo
now that game needs a multi-roll bet to gain some popularity.
Tall-Small-All would be better, jeez.
Quote: WizardHere is my write up at Wizard of Odds. Based on a larger simulation, I've adjusted the house edge to 7.50% (down from 7.52%).
And here is my analysis using the same method I used for doing the fire bet (2^14 states instead of 2^6):
Points | Probability | Pays | Return |
---|---|---|---|
0 | 0.7516771933 | -1 | -0.7516771933 |
1 | 0.1606831552 | -1 | -0.1606831552 |
2 | 0.05628734369 | 5 | 0.2814367185 |
3 | 0.02119189666 | 10 | 0.2119189666 |
4 | 0.007528312681 | 20 | 0.1505662536 |
5 | 0.002219893807 | 50 | 0.1109946903 |
6 | 0.0004122047067 | 200 | 0.08244094133 |
1 | -0.07500277814 |
Quote: mipletQuote: WizardHere is my write up at Wizard of Odds. Based on a larger simulation, I've adjusted the house edge to 7.50% (down from 7.52%).
And here is my analysis using the same method I used for doing the fire bet (2^14 states instead of 2^6):
Points Probability Pays Return 0 0.7516771933 -1 -0.7516771933 1 0.1606831552 -1 -0.1606831552 2 0.05628734369 5 0.2814367185 3 0.02119189666 10 0.2119189666 4 0.007528312681 20 0.1505662536 5 0.002219893807 50 0.1109946903 6 0.0004122047067 200 0.08244094133 1 -0.07500277814
Thanks, really appreciate it. I assume 6 completed numbers is the easiest to calculate. Can you show the actual formula for it?
Quote: wudged
Thanks, really appreciate it. I assume 6 completed numbers is the easiest to calculate. Can you show the actual formula for it?
I did this in Google docs.
A: index. Keeping track of what state the game is in. Index 0 is the start of a new bet. Index 1 is the state after rolling only a hard 10. Index 2, just an easy 10. Index 3, completing the 10 point. 16383 is completing all points. 16384 is default.
B: denominator. Number of rolls that affect the bet. at the beginning all rolls except craps and 11 count (30). If you've already made the 36, 46, and 55 (index 11) then there are only 25 rolls that count.
C: get to. The probability that you will reach this index. Just summing rows T to AG.
D: final end. The probability you will 7 out on this index. 6/B*C
E: Points. The number of completed points.
F to S: The index number in binary. 1 = roll made.
T to AG: Just some fancy vlookup stuff. It calculates the probability the current index number came from rolling each of the previous rolls. An example using index 5:
Index 5 is where we have already rolled 45 and 55. row AE checks to see that 45 was one of the required rolls (a 1 in row Q). we then look up the probability that a previous state was without the 45. This is index 1. The probability of being in state 1 * the probability of rolling a 45 next : (get to from index 1)*2/(dem from index 1).
row AG checks to see that 55 was one of the required rolls (a 1 in row S). we then look up the probability that a previous state was without the 55. This is index 2. The probability of being in state 2 * the probability of rolling a 45 next : (get to from index 2)/(dem from index 2).
Feel free to ask any questions.
The approach I was taking was just comparing each dice combination vs 7 instead of vs 7 and other countable outcomes, since the rolls are independent. This led to the assumption the chance of completing the 10 was simply 1/18 ([(1/36) / (6/36)] * [(2/36) / (6/36)],) which is actually true, but this is the chance of completing 10 in general (other numbers could or could not be completed) and not the chance of completing ONLY the 10. These assumptions compounded on themselves and led to my ridiculous answer.
Quote: miplet
And here is my analysis using the same method I used for doing the fire bet (2^14 states instead of 2^6):
Points Probability Pays Return 0 0.7516771933 -1 -0.7516771933 1 0.1606831552 -1 -0.1606831552 2 0.05628734369 5 0.2814367185 3 0.02119189666 10 0.2119189666 4 0.007528312681 20 0.1505662536 5 0.002219893807 50 0.1109946903 6 0.0004122047067 200 0.08244094133 1 -0.07500277814
Nice work! I thought for doing a recursive analysis, but with so many different states I decided to be lazy and do a simulation.
Mind if I quote your figures on my Odds site? If so, I'd be interested in seeing the exact combinations.
Quote: Wizard
Nice work! I thought for doing a recursive analysis, but with so many different states I decided to be lazy and do a simulation.
Mind if I quote your figures on my Odds site? If so, I'd be interested in seeing the exact combinations.
Yes you may. I linked and tried to explain my spreadsheet a few posts above this here.
Quote: wudgedI visited Dover Downs casino yesterday and saw a new bet called Hot Roller Craps.
Also available at Delaware Park in Wilmington, DE.
The bet was not played by more than 1 or 2 players on a full table. I did play it; it won more than it lost but I never got any higher than 3 numbers covered. The bet stayed up after a seven out when it was a winning bet.
Thank you
Quote: WizardofnothingAre you serious in wanting to play it or just curious?
How is this Hot Roll side bet compare to the Fire, Small-Tall-All side bets?
Quote: WizardofnothingAre you serious in wanting to play it or just curious?
I already play it on the East Coast. Inquiring which Casinos have the bet in Las Vegas.
Quote: wudgedI visited Dover Downs casino yesterday and saw a new bet called Hot Roller Craps.
link to original post
I just saw this bet at the Golden Nugget in Laughlin in June 30, 2023.