Another hardway video
The house rounds down the VIG to $1 in every casino in Las Vegas that I know of that has vig on the win.
1.67 would be correct otherwise.
This is why I specifically point out the bet amounts because it is a sweet spot.
Some casinos will charge a $2.00 vig rounding down if you do them for $50 each. But I don't know which casinos do and don't do this.
Thank you drive through!
Quote: guido111https://wizardofvegas.com/forum/gambling/betting-systems/11356-what-are-the-odds/3/#post183638
you have been corrected before
Too much math to try to predict chaos.
Too many people with OCD about beating the game.
;)
Quote: 7craps"A bet which reaches a winning or losing decision is said to have had action.
The amount of money that is utilized when the bet wins or loses is the amount of action.
This amount can also be called the handle or turnover."
= (0.5*98 - 0.5*100) / (0.5*50 + 0.5*100) = 1.33% HE
EV / action (or average total resolved bet)
Many mistakenly calculate this way
= (0.5*98 - 0.5*100) / (100) = 1.0% HE
ev/total bets at risk
It is not really a bad mistake, it just mixes apples and oranges and treats them as one and the same.
The disagreement that we have is that I claim the handle is twice what you claim.
If you are saying my handle is half what I say because it's not marked on the felt, my claim is that I get that bet anyway because that is my bet that I want.
The only thing that supports your argument is that the bet I described isn't valid because it's not marked on the felt.
There is no other way to claim that I don't have a $100 handle from your example, and instead have a $50 handle.
I have one bet, not two bets as you imply. The fact that I constructed one bet from two marked bets on the felt is only of anecdotal relevance and has absolutely no impact on the math.
I think the OP bets were $25Quote: 7craps
$50 Buy 4 AND $50 Buy 10.
Total at risk = $100
(too powerful a bet for most craps players)
4 or 10 hits, the bets come down
He had another example in a different thread of $5000 that ME corrected.
correct
(0.5*49 - 0.5*50) / (0.5*25 + 0.5*50) = -0.013333333*100 = 1.33%
not correct
(0.5*49 - 0.5*50) / (50) = -0.01*100 = 1%
There are remarkable similarities between these numbers and the chances of winning compared to Baccarat. Including the minimum bets.
I have pointed that out too.
You are leveraging the round down to get the lowest edge at $25, and $50 if the casino rounds down from $2.50 vig to $2.00.
It is extremely annoying to be corrected when I was not incorrect to begin with.
MustangSally, 7 craps, and whoever else can feel all smug for themselves in their associating me with others who can't perform the math.
And I don't say I know every damn thing about the math.
But come on now. I agree with the above poster that all this obsessing over math is a little bit boring, repetitive, and ridiculous.
But when it comes to thinking along the lines of "oh that Ahigh guy can't even do basic math."
Yes I absolutely can. Alright.
I just think the reality is that I know a lot more details about this game than guys who just crunch numbers as a specialty.
I actually play this game a lot.
Okay?
I really do hope you play a lot in the casino and actually make some money.
There is no chaos when a DI claims to be able to have dice distributions that are skewed because of his claimed talent for rolling pairs and hardways.Quote: WongBoWow, the craps threads on this site all have a few things in common....
Too much math to try to predict chaos.
Too many people with OCD about beating the game.
;)
Now it may also include rolling more 4 and 10s per X distributions.
He does not want to risk being made to look like a "fool" on some one else's camera to prove his exceptional DI talent, unless he makes the videos himself.
Hey, I believe all craps DI claims and want to see this filmed in a real casino.
Too many DIs I have seen crash and burn in a casino with all the right number of excuses why they failed.
Even a real TV guy, AlanM has even offered for free to film it, AHigh's hands- his dice rolls-
I want to play at that table and cash in on this DIs talents.
But, it looks like he has too much fear in failure to proceed with the filming
Too bad.
That leaves me to be at the casino when he is.
That has to wait until next month, and who knows what his new talents will be by then.
I do not claim to have his talent or even want it.
I just want the extra money in winnings, if any, to be in my pocket.
Quote: 7craps...$50 Buy 4 AND $50 Buy 10.
Total at risk = $100
(too powerful a bet for most craps players)
4 or 10 hits, the bets come down
Player either nets $98 ($2 vig) or $0
Handle is either $50 or $100
"A bet which reaches a winning or losing decision is said to have had action.
The amount of money that is utilized when the bet wins or loses is the amount of action.
This amount can also be called the handle or turnover."
= (0.5*98 - 0.5*100) / (0.5*50 + 0.5*100) = 1.33% HE
EV / action (or average total resolved bet)
Many mistakenly calculate this way
= (0.5*98 - 0.5*100) / (100) = 1.0% HE
ev/total bets at risk
It is not really a bad mistake, it just mixes apples and oranges and treats them as one and the same.
| EDIT: Thanks to Guido's other thread, 'Multiple Wagers, House Edge', I now understand why the Action is not simply $100.
Since 7craps has me blocked, I cannot ask him to explain this, so perhaps guido you can step in please?
I don't understand why the handle should be (.5*50+.5*100).
We have 2 bets on the table of $50 each.
If a 7 rolls, we lose $100.
If a 4 or a 10 rolls, we win $98, and both bets come down.
So, even though there were 2 bets on the table, only 1 bet will ever get a win resolution. The other bet is taken down with NO ACTION.
So, why should that 'no action' bet need to be included in the Action total ?
The $50 buy is still applicable to gaining advantage due to vig rounding down. I have always been charged $2 for a $50 Buy bet in 100% of the places I have played craps.
Hollywood did charge me $5 for the combined $50 Buy 4 and $50 Buy 10, but they charge the Vig up front, so they are already dicks, and don't count. First time that had ever happened.
Quote: AlanMendelsonI once proposed a mechanical, robotic arm, and if one was built, yes, using the same set and targeting the same spot on a table with the same force and tragectory would result in duplicate throws. That's just basic physics: for every action there is a reaction.
We are severely Off Topic, but seeing as that is not an offense I can be suspended for............
Let's say that a mechanical device could be built that would cause the dice to land EXACTLY the same every time. After the initial contact with the table, is there any guarantee that the dice would bounce the exact same way every time? After all, each time the dice hit an object their shape is changed just a bit due to wear. And even a slight change will be magnified every time the dice bounce, almost to the point that I would think a bouncing die is about as random as can be.
I think it is possible to build a device to make a die land exactly the same way every time, but if the die is allowed to bounce and roll, it will behave differently every time.
Have you read Mission 146 latest addendum : Paragraph 26 Subsection 14 Line 6 ?
Quote: buzzpaff" but seeing as that is not an offense I can be suspended for............"
Have you read Mission 146 latest addendum : Paragraph 26 Subsection 14 Line 6 ?
If there's no link, does that mean it's not real ?
AHigh,
I don't want to get into the Dice setting or the Dice Influence debate. All I care about is results. I have not taken the time to analyze your data, as I don't have enough math background to see anything other than lots of pretty numbers.
HOWEVER, I do care if you can prove that you are able to roll more hardways than what a normal distribution says you should be able to roll. How you achieve this result I could not care less, as long as the casino would not object to your method. Actually, I am hoping you have some kind of Uri Geller telepathic influence on the dice, as that would be the most awesome way of controlling the outcome.
According to the WOO the HE on the Hard 6 and 8 is 9.09%.
The HE on the Hard 4 and 10 is 11.11%.
A normal distribution says you should roll one each of all of these Hardways every 36 rolls.
To make the Hardways bet a +EV, how often would you need to roll just [1] extra Hardway?
In other words, we know we need to overcome roughly a 10% HE on all of the bets.
Would you need to roll 5 hardways in 36 rolls, instead of the expected 4?
Would you need to roll 9 hardways in 72 rolls, instead of the expected 8?
Would you need to roll 13 hardways in 108 rolls, instead of the expected 12?
I don't know how to calculate this. But once we know what the minimum number of hardways it would take to make the hardways a +EV bet, then we could look to your data to see if you are exceeding that requirement. If your data proves that you consistently can exceed the requirement, we can make a boatload of money, and we would never have to worry about heat! Who would expect an AP playing some of the worst bets on the craps table?
Quote: RaleighCrapsAnd now for a change of pace, Back to the OP's topic.........
AHigh,
I don't want to get into the Dice setting or the Dice Influence debate. All I care about is results. I have not taken the time to analyze your data, as I don't have enough math background to see anything other than lots of pretty numbers.
HOWEVER, I do care if you can prove that you are able to roll more hardways than what a normal distribution says you should be able to roll. How you achieve this result I could not care less, as long as the casino would not object to your method. Actually, I am hoping you have some kind of Uri Geller telepathic influence on the dice, as that would be the most awesome way of controlling the outcome.
According to the WOO the HE on the Hard 6 and 8 is 9.09%.
The HE on the Hard 4 and 10 is 11.11%.
A normal distribution says you should roll one each of all of these Hardways every 36 rolls.
To make the Hardways bet a +EV, how often would you need to roll just [1] extra Hardway?
In other words, we know we need to overcome roughly a 10% HE on all of the bets.
Would you need to roll 5 hardways in 36 rolls, instead of the expected 4?
Would you need to roll 9 hardways in 72 rolls, instead of the expected 8?
Would you need to roll 13 hardways in 108 rolls, instead of the expected 12?
I don't know how to calculate this. But once we know what the minimum number of hardways it would take to make the hardways a +EV bet, then we could look to your data to see if you are exceeding that requirement. If your data proves that you consistently can exceed the requirement, we can make a boatload of money, and we would never have to worry about heat! Who would expect an AP playing some of the worst bets on the craps table?
The Math on it is actually fairly simple, which is why I will now do it. Let us say that we are going to bet $10 on all four Hardways every single bet for thirty-six rolls.
The HE on each bet made is 2.78%.
The Odds of rolling a Hardway on any of the numbers are 1/36.
The Odds of rolling any other 6 or 8 are 4/36 each while the odds of any other 4 or 10 each are 2/36.
The Odds of rolling a Seven are 6/36.
The Seven is going to clear all bets, so that should happen six times in a span of 36 rolls which would result in 6 * -$40 = -$240
The Easy Eight and Easy Six are going to come up four times each, so -$10 * 2 (to account for both) * 4 = -$80
The Easy Four and Easy Ten will come up twice each, so -$10 * 2 (to account for both) * 2 = -$40
The total amount lost on the Hardway Bets should be -$360.
The total amount won on the Hard Six and Eight which will come up 1/36 each is $90 * 2 (to account for both) = $180
The total amount won on the Hard Four and Hard Ten which will come up 1/36 each is $70 * 2 (to account for both) = $140
The total amount won on the Hardway Bets should be +$320.
The total EV of all bets is -$40.
2-3-5-9-11-12 will comprise the other fourteen rolls. 1/36, 2/36, 4/36, 4/36, 2/36, 1/36 = 14/36
Conclusion
If you assume an equal distribution of the dice with exception only to the extra Hardway, by throwing 5 Hardways in 36 rolls (as opposed to four) you would have an EV of $30 if it was a Hard Four or Hard Ten and did not replace a losing Hardway bet. If it was a Hard Six or Hard Eight and did not replace a losing Hardway bet, then the result would be an EV of $50. If we are not going for a particular Hardway, then the probability of any HardWay is equal so $50 + $30 /2 = $40 EV.
The EV is $50 in such event that the Hardway roll replaces a losing Easy Way roll. The EV is $80 in such event that the Hardway roll replaces a Seven.
Ultimately, you would only need one extra Hardway per 36 rolls to show a consistent profit. It's actually less than that, but the Math would be more complicated on determining an exact ratio. It is easy to apply the same numbers to 72 rolls, since that only requires doubling everything based on perfect distribution.
Losses: -$720 Wins: $640 Expectation: -$80
In this case, an extra Hard Four or Ten that does not at least replace an Easy 4/6/8/10 would still result in a negative expectation of -$10. If it replaced an Easy 4/10, then the expectation would be $0.00, overall, because that is an $80 swing. If the Hard 4/10 replaced a losing Seven, then it would be $30 because that is a $110 swing.
The extra Hard 6/8 that does not replace a losing roll is $10. If it replaces any Easy roll loss, then it is $20 and $50 if it replaces a Seven.
Generally speaking, then, an extra Hardway roll once every 72 rolls would result in a profit based on roll replaced. First of all, since we are assuming a Hardway, then the extra will be a 6/8 50% of the time. When it is not a 6/8, then it will be a 4/10 which replaces an unresolved roll 14/36 times, replaces a Seven 6/36 times and replaces an Easy 12/36 times.
In other words, you still end up down $10 14/36 times or 10 * 14/36 = -$3.89
You end up with $0 12/36 times, so $0.
You end up ahead $30 6/36 times or $30 * 6/36 = $5.00
In essence, given an otherwise equal distribution of the dice, but with a Hard 4 or Hard 10 replacing any roll, the EV would be $1.11.
IOW, you would basically need only one extra Hardway once every 72 rolls still, but it wouldn't cause a profit every time and could still result in a loss.
Quote: Mission146If you assume an equal distribution of the dice with exception only to the extra Hardway, by throwing 5 Hardways in 36 rolls (as opposed to four) you would have an EV of $30 if it was a Hard Four or Hard Ten and did not replace a losing Hardway bet. If it was a Hard Six or Hard Eight and did not replace a losing Hardway bet, then the result would be an EV of $50. If we are not going for a particular Hardway, then the probability of any HardWay is equal so $50 + $30 /2 = $40 EV.
The EV is $50 in such event that the Hardway roll replaces a losing Easy Way roll. The EV is $80 in such event that the Hardway roll replaces a Seven.
Ultimately, you would only need one extra Hardway per 36 rolls to show a consistent profit.
"If we are not going for a particular Hardway, then the probability of any HardWay is equal so $50 + $30 /2 = $40 EV."
If the expected EV for 36 rolls is -$40, and the average EV for 1 extra hardway in 36 rolls is +$40, wouldn't that make it break even if the shooter could roll 9 hardways out of 72 rolls? Understanding of course, that if the 9th hardway was a 4 or 10, we could still possibly end up -$10, depending on what roll was replaced.
Therefore, rolling 5 hardways in 36 rolls is overachieving. If he rolls 9 hardways in 72 we should break even, so if he rolls 14 hardways in 106 throws, we come out ahead.
Average roll lasts 8.x throws? So the shooter would have to be the shooter about 11x or so in order for us to come out ahead. That could be a really long session to get a chance to shoot 11x...........
|EDIT: Mission146 was editing his post (adding more information) at the same time I was creating mine, and we managed to duplicate each other.........
No man should be afraid to make a bet, however..........
Quote: RaleighCrapsWe are severely Off Topic, but seeing as that is not an offense I can be suspended for............
Let's say that a mechanical device could be built that would cause the dice to land EXACTLY the same every time. After the initial contact with the table, is there any guarantee that the dice would bounce the exact same way every time? After all, each time the dice hit an object their shape is changed just a bit due to wear. And even a slight change will be magnified every time the dice bounce, almost to the point that I would think a bouncing die is about as random as can be.
I think it is possible to build a device to make a die land exactly the same way every time, but if the die is allowed to bounce and roll, it will behave differently every time.
Raleigh, you are absolutely right that every time there is a throw of the dice there are ever-so-slight changes made to the dice, to the felt, to the back wall, to the alligator bumps. There is also the risk of minor earth tremors, of someone hitting the table, of air currents.
But any test of a robotic arm would not involve one single "robotic throw" and I think these "changes" would not have a significant impact on the results of say a hundred throws or a thousand throws.
Of course the only way to know for sure is to build it and do it. In theory, there should be no change in results. And in theory, human dice control is also possible.
So the question becomes does a machine have a better chance than a human when you take into account the other variables, as slight as they might be??
Quote: odiousgambitI can understand being camera shy. Let's back off from that.
No man should be afraid to make a bet, however..........
You realize this thread is about a video I made with me being on camera right?
You realize that I play 3 hours of craps every day right?
Not sure I understand what you're thinking at all.
In a previous post we determined that if you can roll 14 hardways in 106 throws, instead of the expected 12, you can overcome the HE on the FOUR hardways bets. To make it simple, I am going to make four $100 Hardways bets active each time you throw the dice, including the come out rolls (I want to get paid when you set a point with a hardway roll). Can you do this consistently ?
Quote: AhighHere's the roll data for the entire session:
64`33`53`11`53`41`21`22`53`52`62`54`23`62`62`44`31`53`22`62`54`54`11`66`43`31`42`31`22`42
54`55`12`23`52`43`51`63`54`12`51`54`44`54`31`31`44`55`62`51`54`61`51`55`43`52`12`43`53`22
53`53`55`11`43`66`61`23`41`54`23`55`51`64`41`42`12`64`64`53`53`23`54`56`31`55`33`41`64`12
23`31`66`43`32`45`56`62`31`64`44`12`44`31`42`41`51`54`45`31`65`33`44`11`51`51`23`22`65`31
63`54`66`31`23`54`33`41`23
.
.
.
Even if all that comes out of this is fun and entertainment, I am absolutely enjoying all of this and I want to thank everyone for their continued interest in following along and helping me with the parts I don't understand like some of the more advanced math.
Okay, I was bored and decided to do some grunt work. I took these 129 rolls and made the following conditions:
- I was betting $100 on every hardway.
- I was active on every roll.
When I ran the above rolls with this criteria, I had $16,700 in winning bets, and I lost $8,800, for a net win of $7,900!
Pretty impressive.
Then I looked at the distribution of the numbers as to what was rolled, against expected numbers:
| Number | # of hits | # Expected |
|---|---|---|
| 2 | 4 | 3.58 |
| 3 | 7 | 7.17 |
| Easy 4 | 12 | 7.17 |
| Hard 4 | 5 | 3.58 |
| 5 | 16 | 14.33 |
| Easy 6 | 12 | 14.33 |
| Hard 6 | 4 | 3.58 |
| 7 | 11 | 21.50 |
| Easy 8 | 15 | 14.33 |
| Hard 8 | 6 | 3.58 |
| 9 | 17 | 14.33 |
| Easy 10 | 6 | 7.17 |
| Hard 10 | 6 | 3.58 |
| 11 | 4 | 7.17 |
| 12 | 4 | 3.58 |
| Totals | 129 | 129 |
I believe you mention you use a couple of different pairs sets. One I know was the 66 faces up. My understanding of dice setting is pairs setting is dangerous because it can lead to more 7s than any other set. However, you rolled only 50% of the expected number of 7s! This defies everything I have read on dice sets. Although, usually the dice setters are trying to keep the dice on axis, which I guess you are not trying to do.
Ahigh, do you have any comments or any ideas why the numbers appear this way? Can you do it again?
Please record a series of rolls like you did above, in a multiple of 36 if you would, and I can plug your roll sequence into the spreadsheet I have built. Please have a video to back up the roll sequence you provide.
I will plug the new rolls into the spreadsheet, and can produce the win/loss and the roll distribution table for comparison again.
This is all very interesting.
Quote: RaleighCrapsOkay, I was bored and decided to do some grunt work. I took these 129 rolls and made the following conditions:
- I was betting $100 on every hardway.
- I was active on every roll.
When I ran the above rolls with this criteria, I had $16,700 in winning bets, and I lost $8,800, for a net win of $7,900!
Pretty impressive.
Then I looked at the distribution of the numbers as to what was rolled, against expected numbers:
Number # of hits # Expected 2 4 3.58 3 7 7.17 Easy 4 12 7.17 Hard 4 5 3.58 5 16 14.33 Easy 6 12 14.33 Hard 6 4 3.58 7 11 21.50 Easy 8 15 14.33 Hard 8 6 3.58 9 17 14.33 Easy 10 6 7.17 Hard 10 6 3.58 11 4 7.17 12 4 3.58 Totals 129 129
I believe you mention you use a couple of different pairs sets. One I know was the 66 faces up. My understanding of dice setting is pairs setting is dangerous because it can lead to more 7s than any other set. However, you rolled only 50% of the expected number of 7s! This defies everything I have read on dice sets. Although, usually the dice setters are trying to keep the dice on axis, which I guess you are not trying to do.
Ahigh, do you have any comments or any ideas why the numbers appear this way? Can you do it again?
Please record a series of rolls like you did above, in a multiple of 36 if you would, and I can plug your roll sequence into the spreadsheet I have built. Please have a video to back up the roll sequence you provide.
I will plug the new rolls into the spreadsheet, and can produce the win/loss and the roll distribution table for comparison again.
This is all very interesting.
I am getting ready for G2E at work and I will probably be working this weekend and every waking moment until G2E is done. I'm anxious to do what you're asking and I think it's awesome that you're interested. I had just about accepted the fact that all the work I had done to this point was for nothing.
I actually have something called an "AverMedia Gamer Live HD" card that broadcasts video live.
http://q.avermedia.com/en/recentral/live-gamer-hd/xsplit-for-live-gamer-hd
Maybe something like this would help verify the rolls as not be taken from the "best of" series if all the recorded rolls are broadcast live and they are all published without exception (even the worst ones).
I have published here every single roll I have recorded since August 1st so far with this particular throw. I did do an experimental session with another throw that had horrible results, but in the video I start out explaining that that session was experimental and a different throw. I can also make this video and the roll results available if there is interest to see every single roll that I have recorded. But it absolutely does not make sense to put that in the same database with the other throws as I was just experimenting with a new throw.
With this other throw, I have been able to get a few back-to-back pairs. Today in the casino I rolled hard 4 back-to-back with this other new throw. But in general, I haven't used this throw enough to really count on it. The hard-4 back-to-back was a point and winner and I won five bucks if that helps you understand how much I trust my own throw on this throw (I don't .. it's experimental).
But if you look at those 1428 rolls or however many, there are so HORRIBLE strings of seven after seven after seven even with my best effort and best throw. You gotta know I wasn't happy about that!
Anyway if you can wait until after G2E I will be able to focus more. But now I gotta get to work!!!
Inquiring minds still want to know...Quote: RaleighCrapsAhigh,
In a previous post we determined that if you can roll 14 hardways in 106 throws, instead of the expected 12, you can overcome the HE on the FOUR hardways bets. To make it simple, I am going to make four $100 Hardways bets active each time you throw the dice, including the come out rolls (I want to get paid when you set a point with a hardway roll). Can you do this consistently ?
Can you do this consistently... in a real casino?
Ahigh, Where is the data from all your actual casino dice rolls?
I assume there are some.
RaleighCraps and I do not mind crunching the numbers.
If they are not available because you have just recently changed your dice set and you can play Craps 3 hours each day, when can we expect some actual casino dice roll data?
Thanks
RaleighCraps, great question.Quote: RaleighCraps
To make the Hardways bet a +EV, how often would you need to roll just [1] extra Hardway?
In other words, we know we need to overcome roughly a 10% HE on all of the bets.
Would you need to roll 5 hardways in 36 rolls, instead of the expected 4?
Would you need to roll 9 hardways in 72 rolls, instead of the expected 8?
Would you need to roll 13 hardways in 108 rolls, instead of the expected 12?
I don't know how to calculate this. But once we know what the minimum number of hardways it would take to make the hardways a +EV bet, then we could look to your data to see if you are exceeding that requirement. If your data proves that you consistently can exceed the requirement, we can make a boatload of money, and we would never have to worry about heat! Who would expect an AP playing some of the worst bets on the craps table?
I have to go over the math by M146 later, but I do remember Alan Shank talking about this.
For the Hard 6 (or Hard8) for every 100 bets resolved, only need one extra win to break even. 2 to show a profit.
I think the H4,H10 was for every 80 resolved bets, a little more difficult.
Flat betting the hardways I think would be difficult with more dice rolls.
I would think the average bet could increase per each distribution.
That brings us back to the distribution size.
In WinCraps we can easily alter the dice probabilities to match what AHigh says he can throw.
We could then easily simulate to get the results
AHigh's casino dice roll data would be interesting to compare against his home roll data.
Quote: AhighI am getting ready for G2E at work and I will probably be working this weekend and every waking moment until G2E is done. I'm anxious to do what you're asking and I think it's awesome that you're interested. I had just about accepted the fact that all the work I had done to this point was for nothing.
I actually have something called an "AverMedia Gamer Live HD" card that broadcasts video live.
http://q.avermedia.com/en/recentral/live-gamer-hd/xsplit-for-live-gamer-hd
Maybe something like this would help verify the rolls as not be taken from the "best of" series if all the recorded rolls are broadcast live and they are all published without exception (even the worst ones).
.
.
.
Anyway if you can wait until after G2E I will be able to focus more. But now I gotta get to work!!!
I can certainly wait until after G2E.
At this point, I do not need to witness your throws live. I just want to be sure you are giving me real, unedited data.
- Decide how many throws you are making for this session, before you begin. 36,72,108,144,180,or 216.
- Start the Camera, Show to the camera, the number of throws you are going to make, and then start chucking dem bones.
- No warm-up throws, no picking a point to start counting the rolls.
- Give me the roll sequence like you did in the post I quoted.
I don't care about looking at your history, I only want to see what you can do going forward. Use the throw/set you have the most confidence in.
I will plug the numbers into the spreadsheet, and we'll see how the $100 hardway bets would have done, and what your distribution looks like compared to expected.
If it tiruns out after we have done a dozen or so of these sessions, that the hardways bet would have been profitable, then we may want to try some of the live video feed demonstrations, as a sanity check.
We may also want to throw in a control subject as a test. It would be awesome if we could get a trained monkey :-D , but I guess any human would do. Have them throw on your set up, just to make sure there isn't anything that biases hardways in the set up.
Unlike most of the members here, I don't necessarily have to be convinced of the math or the science. I don't have an overwhelming need to be able to explain why something is happening. I only care if it is happening, and can it be consistently repeated. I used to play Parchessi with my wife a lot (until she got mad after a loss and broke the board over my head). She had an uncanny ability to throw double, double and send me back home time after time after time. I have no explanation for it, nor do I really need one. All I know is, it happened over and over. It was like she was able to will herself into throwing the damn things. I wish she would apply that mojo to craps for me. Alas, she hates the game.
If you can demonstrate a consistency with this, not only can we take hardways action in the casino, but I know SOOPOO would make a fair wager available against this ability too. I hope you can generate enough consistency to make this a possibility. It would be the first craps challenge to culminate in a bet that I am aware of, and should be exciting.
Start a new thread with your first session, whenever you get time to get back to it.
Repeat this mantra 100 times a day until you can substitute the word " have" for the word "had just about".
Only then will you have inner peace and happiness.
Quote: RaleighCrapsAwesome Mission146! (Don't let it go to your head, I am easily impressed.)
Thanks! I won't.
Quote:If the expected EV for 36 rolls is -$40, and the average EV for 1 extra hardway in 36 rolls is +$40, wouldn't that make it break even if the shooter could roll 9 hardways out of 72 rolls? Understanding of course, that if the 9th hardway was a 4 or 10, we could still possibly end up -$10, depending on what roll was replaced.
The $40 was not the average EV for any Hardway, the verbiage in my post above is kind of convoluted, but what I was expressing is that $40 is the average swing for a Hardway roll that does not replace a losing Hardway roll. (i.e. Easy 4/6/8/10 or Seven) That's given one additional Hardway in 36 rolls.
Quote:Therefore, rolling 5 hardways in 36 rolls is overachieving. If he rolls 9 hardways in 72 we should break even, so if he rolls 14 hardways in 106 throws, we come out ahead.
That was amended above. If he rolls 9/72, the expectation is still a profit. I intend to determine the exact number of rolls where we have a break-even point (or the closest possible thing) for Nine Hardways. I believe I have the formula for it (thought of it in the shower) just not the time to do it atm. Will get to it later.
Quote:Average roll lasts 8.x throws? So the shooter would have to be the shooter about 11x or so in order for us to come out ahead. That could be a really long session to get a chance to shoot 11x...........
Ideally, were someone a DI who could accomplish this successfully, he would be well-advised to play at a Table as the only player, if possible, and certainly to play with as few people as possible.
Quote:|EDIT: Mission146 was editing his post (adding more information) at the same time I was creating mine, and we managed to duplicate each other.........
Yes, hopefully that clarified the 72 rolls and +$40 thing.
Quote: AhighYou realize this thread is about a video I made with me being on camera right?
You realize that I play 3 hours of craps every day right?
Not sure I understand what you're thinking at all.
You don't want someone else to use a camera, it would seem.
And you can't seem to think of a bet that would answer the challenge, "back up your claims"
do you have Wincraps on your computer? You can tweak the probabilities of rolling this and that on it.
We have determined that based on perfect probabilistic distribution we can be expected to lose $40 over 36 rolls by betting all four Hardways each time.
This represents a loss of $1.11/roll, on average.
We will now determine the Hardway replacement value for each possible result. This will initially be based upon 32 rolls as a Hardway does not replace itself.
6/32 rolls will be Sevens and replace a loss of $40.
12/32 rolls replace an Easy Way loss of $10.
14/32 rolls replace rolls that would not resolve the bet.
We are going to look at the total, "Swing," multiplied by the probabilities of each event occurring. We are assuming the extra Hardway, so the 4/36 chance of a Hardway never comes into play, since we make no assumptions as to what specific Hardway we are rolling, they will be equally likely.
An extra Hard Six or Hard Eight would reflect a, "Swing," of $130 to replace a Seven, $100 to replace an Easy and $90 to replace a non-resolution roll.
Seven: $130 * 6/32 = $24.375
Easy: $100 * 12/32 = $37.5
Non: $90 * 14/32 = $39.375
Total: $101.25
An extra Hard Four or Hard Ten would reflect a, "Swing," of $110 to replace a Seven, $80 to replace an Easy and $70 to replace a non-resolution roll.
Seven: $110 * 6/32 = $20.625
Easy: $80 * 12/32 = $30
Non: $70 * 14/32 = $30.625
Total: $81.25
($101.25 + $81.25)/2 (Equal Liklihood) = $91.25
$91.25/1.11 (Expected Loss per Roll w/Perfect Distribution) = 82.207 rolls
PROOF:
EV of 82.207207207 rolls all with Hardaway Bet:
82.207207207 * 14/36 = 31.969469469 (Rolls that do nothing)
82.207207207 * 8/36 = 18.2682682682 (Easy 6, Easy 8) * -$10 (loss) = -$182.86828268
82.207207207 * 4/36 = 9.13413413 (Easy 4, Easy 10) * -$10 (loss) = -$91.3413413
82.207207207 * 6/36 = 13.701201201 (Sevens) * -$40 = -$548.048048
82.207207207 * 2/36 = 4.567067067 (Winning Hard Six/Eight) * $90 = $411.036036
82.207207207 * 2/36 = 4.567067067 (Winning Hard Four/Ten) * $70 = $319.694694
TOTAL: -$91.52
***NOTE***
The numbers do not come out exactly right because the $91.25 from above was divided by $1.11 which should have been $1.11111111111 had I not rounded down. That would then be $91.25/1.111111111111 or 82.125 rolls. Therefore, the expected loss would be less, I'm guessing -$91.25 with perfect distribution. I'll do this a bit later to find out.
82.125 * 14/36 = 31.9375 (Rolls that do nothing)
82.125* 8/36 = 18.25 (Easy 6, Easy 8) * -$10 (loss) = -$182.50
82.125 * 4/36 = 9.125 (Easy 4, Easy 10) * -$10 (loss) = -$91.25
82.125 * 6/36 = 13.6875 (Sevens) * -$40 = -$547.50
82.125 * 2/36 = 4.5625 (Winning Hard Six/Eight) * $90 = $410.625
82.125 * 2/36 = 4.5625 (Winning Hard Four/Ten) * $70 = $319.375
TOTAL: -$91.25
Conclusion
If you bet $10 on all Hardways every single roll and you get one extra Hardway for every 82.125 rolls equally distributed between 6/8 and 4/10, then you will break even. In the event that you roll in excess of one extra Hardway for every 82.125 rolls, you will profit based upon the Hardway replacement value and all else going according to equal distribution.
In 82.125 rolls you are normally expected to roll 82.125 * 4/36 = 9.125 Hardways, so you would simply need 10.125 per 82.125 rolls.
Quote: Mission146EV of 82.125 rolls all with Hardaway Bet:
82.125 * 14/36 = 31.9375 (Rolls that do nothing)
82.125* 8/36 = 18.25 (Easy 6, Easy 8) * -$10 (loss) = -$182.50
82.125 * 4/36 = 9.125 (Easy 4, Easy 10) * -$10 (loss) = -$91.25
82.125 * 6/36 = 13.6875 (Sevens) * -$40 = -$547.50
82.125 * 2/36 = 4.5625 (Winning Hard Six/Eight) * $90 = $410.625
82.125 * 2/36 = 4.5625 (Winning Hard Four/Ten) * $70 = $319.375
TOTAL: -$91.25
Conclusion
If you bet $10 on all Hardways every single roll and you get one extra Hardway for every 82.125 rolls equally distributed between 6/8 and 4/10, then you will break even. In the event that you roll in excess of one extra Hardway for every 82.125 rolls, you will profit based upon the Hardway replacement value and all else going according to equal distribution.
In 82.125 rolls you are normally expected to roll 82.125 * 4/36 = 9.125 Hardways, so you would simply need 10.125 per 82.125 rolls.
This is pretty good stuff. Since I generally don't bet big hardways in the casino, I've been anxious to do my own analysis with my software at home. But this is absolutely awesome what you've done here. The edge at 2.78% per roll is only one way to understand what you have to overcome. But this is a much more intuitive way to understand how the math works out to how many more hardways you need per group of rolls.
What I do know and what I have witnessed over the years is that people win big on hardways betting on me. The dealers and many other people often ask me why I don't bet hardways, and the general answer I give them is that I don't want to lose my money that quickly. It seems like beating a 2.78 edge per roll would be a lot harder than beating a 0.42 edge per roll. And so I generally focus on line bets and free odds. But maybe I will change my strategy. I generally get my ass kicked on hi/lo and ace-deuce when I do get it kicked. I one bet $750 on the come and rolled the ace-deuce. (Not that this is my biggest one-roll loss, but that was a comical one).
Yesterday I rolled (at the Silverton) aces, aces, ace-deuce, twelve, yo in five consecutive rolls. I didn't win or lose anything as I didn't have any line bets or horn bets (I never bet any horn bets ever) for those rolls. But martingale on the come for that sequence would not have been nice if your theory about low edge on the come with a betting system were part of your method of grinding a profit.
I'm heavy on those from my charts. So I could be having a bigger negative edge per roll on myself than the house has with 0.42% if I do have a bias with my current betting scheme. I don't mind telling you that I am not making claims that I always win a bunch of money and I am not claiming to even be ahead.
Quote: Ahigh
This is pretty good stuff. Since I generally don't bet big hardways in the casino, I've been anxious to do my own analysis with my software at home. But this is absolutely awesome what you've done here. The edge at 2.78% per roll is only one way to understand what you have to overcome. But this is a much more intuitive way to understand how the math works out to how many more hardways you need per group of rolls.
Thanks for the compliment! It's useful for that purpose, but Variance would be a factor even if you were getting that extra Hardway. It's important to remember that averaging that extra Hardway per those 82.125 rolls still results in a loss if the extra Hardway is a 4/10 replacing a non-resolution roll or even if it replaces an Easy Way of any kind. In fact, the 6/8 still results in an overall loss in this scenario if it replaces a Non-Resolution roll. A not insignificant portion of the value comes from the extra HW replacing a Seven, but again, with otherwise perfect distribution, that's your break-even point.
We established earlier that an extra Hardway per 72 rolls results in a break-even on 4/10 replacing an Easy and still a loss on 4/10 replacing a non-resolution. The lowest Hardway swing is +$70 when a 4/10 replaces a non-resolution, so if you want, I can determine at what number of rolls you are looking at -$70 EV rolling the appropriate amount of Hardways. Your goal would then become to roll an extra Hardway per that number of rolls, in which event, you would do no worse than +/-$0.00 if you were successful.
Actually, that's easy, so I'll do it here. We know that -$1.111111111 is the expected loss per roll, so -$70/-1.111111111 = 63 Rolls. In 63 Rolls you are expected to get 63 * 4/36 = 7 Hardways, so you would hit a pretty good, "Safety point," by getting to the point where 8/63 of your rolls are Hardways. If you did that consistently, your, "Worst," Hardway of 4/10 replacing a Non-Resolution is a break-even.
How many times do you think you can fire off 63 rolls in an eight-hour period if you are the only one at the table?
Quote:What I do know and what I have witnessed over the years is that people win big on hardways betting on me. The dealers and many other people often ask me why I don't bet hardways, and the general answer I give them is that I don't want to lose my money that quickly. It seems like beating a 2.78 edge per roll would be a lot harder than beating a 0.42 edge per roll. And so I generally focus on line bets and free odds. But maybe I will change my strategy. I generally get my ass kicked on hi/lo and ace-deuce when I do get it kicked. I one bet $750 on the come and rolled the ace-deuce. (Not that this is my biggest one-roll loss, but that was a comical one).
Let's assume for a second that you will be capable of setting the dice successfully to exert at least a minor influence on them: Why not learn how to throw Sevens, play the DP w/Odds, let the first roll essentially be random, and then start throwing for Sevens? It seems that you are intimidated by the HE on HW's and do not want to lose your money quickly, so this method would put you against the lowest HE on the Table?
Quote:Yesterday I rolled (at the Silverton) aces, aces, ace-deuce, twelve, yo in five consecutive rolls. I didn't win or lose anything as I didn't have any line bets or horn bets (I never bet any horn bets ever) for those rolls. But martingale on the come for that sequence would not have been nice if your theory about low edge on the come with a betting system were part of your method of grinding a profit.
If you're interested, the same, "Replacement Roll Metric," could be applied to all Doubles, Doubles + Yo or just on the Yo. Based on your previous statements, however, I doubt if you want to get after that with a bet on every single roll, and understandably not.
The long-term HE does not change, and all betting systems are useless from a standpoint of actually trying to win long-term, but I should prefer the Labouchere to the Martingale for lower potential of having your bankroll drained (or hitting Table Max) in only a few rolls.
Quote:I'm heavy on those from my charts. So I could be having a bigger negative edge per roll on myself than the house has with 0.42% if I do have a bias with my current betting scheme. I don't mind telling you that I am not making claims that I always win a bunch of money and I am not claiming to even be ahead.
It's cool, I'm just discussing the theoretical possibility that someone could actually influence the dice to roll more Hardways, anyway. It was an interesting, though ultimately pretty easy, problem.
And that is not likely to change in the immediate future !
Quote: buzzpaff" I don't mind telling you that I am not making claims that I always win a bunch of money and I am NOT claiming to even be ahead. "
And that is not likely to change in the immediate future !
Right. Now why don't you do something for yourself. You've helped me enough. And thank you! You're awesome.
Good Variance in your next visit to a casino.
Nice work.Quote: Mission146... It's useful for that purpose, but Variance would be a factor even if you were getting that extra Hardway.
Variance is always the main factor.
Lots of nice math work and no mention of actual variance anywhere and not even sim data to back up the math claims.
To each his own.
There should be no surprise to many, the Wizard, MathExtremist to name a few,
that in any set of 82 dice rolls while betting $100 on every hardway, every roll, working every roll,
about 37% of all the sessions played would and do show a $2100 average net win ($100 bets).
The median looks to be $1700. Interesting.
No DI skills at all are needed for these actual events.
Point is, the DI who wants to conquer the hardways really can have his cake and eat it too.
It should be very easy to accomplish, since you only have to do better for 63% of the sessions.
Randomness takes care of the rest. Just do not mess that part up.
With a -9.15 unit mean (rounded)
29unit standard deviation (rounded)
And only 20% of the time the 82 roll sessions ended in the black because of more hardways.
The other 80%?
Yep, you guessed it. Less losses. Especially from less 7s rolled.
Nothing new.
Gambling is not all about just winning, or winning the most.
Every gambler should also be concerned about losing the least.
a+b=$
Thank you for the compliments on the Math work, I believe. I do not see where I would need Sim data to back up the Math claims as they all consist of pretty rudimentary stuff, hopefully I didn't screw up, though I don't see how I could have.
The results are theoretical and based on otherwise perfect distribution of the dice, which I did state on multiple occasions. I don't know if it is possible to program a simulation to roll perfectly, with exception to an extra Hardway once per 82 rolls, but if so, please feel free to do that to verify my results as I have no idea how to program a simulation, or anything else.
I have also stated that control over 7's, if, "Dice Control," can actually be had would be far more prudent than the ability to occasionally roll the extra Hardway.
Quote: RaleighCrapsOkay, I was bored and decided to do some grunt work. I took these 129 rolls and made the following conditions:
- I was betting $100 on every hardway.
- I was active on every roll.
When I ran the above rolls with this criteria, I had $16,700 in winning bets, and I lost $8,800, for a net win of $7,900!
Pretty impressive.
Then I looked at the distribution of the numbers as to what was rolled, against expected numbers:
Number # of hits # Expected 2 4 3.58 3 7 7.17 Easy 4 12 7.17 Hard 4 5 3.58 5 16 14.33 Easy 6 12 14.33 Hard 6 4 3.58 7 11 21.50 Easy 8 15 14.33 Hard 8 6 3.58 9 17 14.33 Easy 10 6 7.17 Hard 10 6 3.58 11 4 7.17 12 4 3.58 Totals 129 129
I believe you mention you use a couple of different pairs sets. One I know was the 66 faces up. My understanding of dice setting is pairs setting is dangerous because it can lead to more 7s than any other set. However, you rolled only 50% of the expected number of 7s! This defies everything I have read on dice sets. Although, usually the dice setters are trying to keep the dice on axis, which I guess you are not trying to do.
Ahigh, do you have any comments or any ideas why the numbers appear this way? Can you do it again?
Please record a series of rolls like you did above, in a multiple of 36 if you would, and I can plug your roll sequence into the spreadsheet I have built. Please have a video to back up the roll sequence you provide.
I will plug the new rolls into the spreadsheet, and can produce the win/loss and the roll distribution table for comparison again.
This is all very interesting.
Now that G2E is over, can we expect to see some hardways roll sequences? I think you posted about some Field sequences though. Have you given up on the hardways experiment?
Quote: buzzpaffI would post my video of me having 163 rolls without a 7, but I left the lens cap on the camera. SIGH
I don't believe you. Please post a picture as proof
