dice experiment

Hello everyone, I decided to do a dice control experiment beyond what I have seen done so far on the forums. I have a set up in my basement, not a table but as close as I can get with a minimal investment . I have a landing station with the rubber pyramid back wall and casino grade felt. I made the surface to resemble the tables I like to play on. I would call it medium, not rock hard but not super bouncy. It is wide and deep enough for the dice to come to rest without hitting the side walls. It is 28" off the floor as are most casino table surfaces. I have a similar throwing station. I positioned the throwing station 7 1/2' from the backwall which would put me next to the stickman on a standard 12' table which is where I like to shoot from. I used three sets of dice. The right was always blue and the left always green. this way we could track what each die did. I used the hardway set with the 5's on top and the 3's facing me on every throw. I threw in sessions with a minimum of 30 throws and a maximum of 90 per session over the course of two to three weeks. I made sure at least one die hit the backwall every throw but I tried to make sure they both did. I did not count them if one flew off the table. I made a total of 1024 rolls which I recorded and anyalyzed on bone tracker and Dice tool which was created by Stanford Wong and the Mad Professor.
I know 1024 rolls is a very small sample compared to the actual math of the game but I found the results interesting and the sample large enough so I thought I would share. Let me also add that I am not saying either way that dice control is possible or not, I have practiced for almost 2 years and I still do not think I can do it.
Here are my results, after all 1024 rolls my sevens to rolls ratio was 6.82, at one point after 725 rolls it was 7.22. This means I should have an edge on certain bets on the table. The CHi test showed after 725 rolls I was definitely influencing the dice but after all 1024 rolls the chi test said there was only possible influence. The dice tool also tracks what the results would have been had I used a different set. The throwing set came out with the highest SRR. Tied with it though was the paralell 6's with 6's on top and the 3's facing me. Box number to rolls ratio was 4.8 with random being 4. My dice stayed on axis 46.2 percent of the time with random being 44.44. So all this says I should have an edge on the pass or come with odds and the 6 and 8. Here is where it gets interesting. I then manually went through the sheets and tracked the bets as if I was actually making them on these rolls. I first did a $10 pass line with double odds. With this betting scenario I would have lost $530 on these 1024 rolls even though I should have an edge. I then added a come bet to the same scenario the come bet only lost $114. I then went through and looked at what would happen if I was the shooter but only bet the pass with no odds (so I could shoot) and placed the 6 and 8 for $12 each. That scenario would have produced a profit of $1292. At first I found this odd but it makes sense because the 6 and 8 are the numbers that should appear the most next to the 7. It also means that just avoiding the 7 is not enough you must be able to hit certain numbers more than called for, for dice control to work and obviously I was hitting other box numbers but not able to roll my point number. So if we do have a slight edge do we forget the pass with odds and just get some money on the inside? I also ran a few other scenarios. I took the intial $10 pass line bet with double odds and implemented the 5 count. This produced a profit of $350 instead of a loss of $530. I also ran $44 inside which produced a profit of $1170. Buying the 4 and 10 produced a profit of $550 but buying just the ten added $1000 to the cause. I also loooked at a couple superstitions. The first is the 7 follows the hard 10. I threw 31 hard 10's only 6 were followed by a 7. I also looked at how many point, sevens there were. I would have made an extra $160 by not laying odds until a box nuber was thrown after the point was established. I did not figure this scenario with place bets only the pass with odds. I believe that had I not had my place bets working after a point was established until a box number was thrown I would have been even further ahead. I found it interesting that in the end the numbers showed I had an edge in the game but lost money on the lowest house edge bet available. I am heading to Vegas in a week I am going to try the pass no odds with the 6 and 8 when I am shooting and see if the theoretical profit I can muster up in my basement can show up in real action if I can get the dice to behave. I will keep you posted with the return results. I know that this is a very small unscientific test but I stiil thought the results interesting enough to share with you. I know I do not have the excitement, distractions or atmosphere in my basement that we have in the casino but I am going to give it my best effort.

tl,dr.

Paragraphs...they are your friend.

Quote: Triplell

Paragraphs...they are your friend.

Yes, sorry I did ramble a bit

Maybe I am missing something here... but if you can acheive a 7 / non seven ration of 1 to 6.82, just roll 'regular' on the come out roll and you will have a HUGE advantage once you start 'controlling' once a point has been established. As stated many times, I am willing to bet you cannot acheive that level of dice control in a real casino which requires you to hit the back wall with both dice. Good luck in Vegas, and let us know how you do.

Thanks for that, Ralph. It is consistent with my anecdotal experiences. Keep us posted on your casino work to come!

MrRalph
I’m toying with a similar method as to your Pass Line w/ no odds , and placing the 6 and/or 8.

However, I’m making a come bet on the second roll, nothing else .( I hate those point Sevens.)
If the come bet goes to a 6 or 8, I will take odds, and place the other 6 or 8.
If the bet goes to another number besides the 6 or 8, I take no odds on my come bet, but now place the 6 and 8 .

Good Luck on your experiment

I forgot that in my post. I figured taking odds if the point was 6 or 8 and not placing that number

Not surprised at all by the results. I have had similar short term results with my SRR as well. Keep up the practice and have fun with it.

Quote: SOOPOO

Maybe I am missing something here... but if you can acheive a 7 / non seven ration of 1 to 6.82, just roll 'regular' on the come out roll and you will have a HUGE advantage once you start 'controlling' once a point has been established. As stated many times, I am willing to bet you cannot acheive that level of dice control in a real casino which requires you to hit the back wall with both dice. Good luck in Vegas, and let us know how you do.

No my point was that having an SSR of 1 to 6.82 was not enough to show a profit as I could not repeat the point number. Even though on paper I had an edge in the physical I could not roll that point number before the 7 appeared. So just having a SSR above random in this small sampling was not enough to over come the lowest house edge bet on the table which I found interesting. Thanks for the good wishes and I will let you all know how it played out.

thanks for taking the time to do that.

but I would like to make this point:

For everyone who has ever practiced throwing dice at home, and trying to influence or control the dice with your practice at home, you are pretty much wasting your time using any kind of home made or privately bought table. Because every actual table in use in a casino is unique and you can never get the same results.

the only thing you can practice at home is your grip and release and throw -- and you can do that with the dice hitting your bed.

Once your perfect grip, release and throw hits a real craps table in a casino, there is nothing else that you can duplicate with your home practice.

Thanks for the recap, very interesting. Good luck and let us know how you do. Do you keep just as detailed reports on your rolls in a casino?

Thanks for the experiment and most importantly the analysis. I too have a better than 6:1 SRR, but most of the time, I don't win. Having a workable betting strategy is even more important than having a long roll.

I picked up a few pointers from you: Use 5 counts, bet 6 & 8, place bet inside if only I hit an inside number, come bet after the comeout roll, not always place odds at the pt., bet the "signature" number.

Good luck at Vegas. My suggestion is to find a very low min. table, bet only the min. on Pass Line, and just practice until you get the confidence.

Quote: AlanMendelson

For everyone who has ever practiced throwing dice at home, and trying to influence or control the dice with your practice at home, you are pretty much wasting your time using any kind of home made or privately bought table. Because every actual table in use in a casino is unique and you can never get the same results.


Once your perfect grip, release and throw hits a real craps table in a casino, there is nothing else that you can duplicate with your home practice.

I disagree. You can make one or two stations that simulate different tables at your local casinos. This isn't that hard. Than you have at least a feel for them, and in most cases a distinct advantage. It depends on how you practice.

It isn't that hard to re-create a casino table.

Edit: Double post, ignore. WTF

Quote: sam46810

Thanks for the experiment and most importantly the analysis. I too have a better than 6:1 SRR, but most of the time, I don't win. Having a workable betting strategy is even more important than having a long roll.

I picked up a few pointers from you: Use 5 counts, bet 6 & 8, place bet inside if only I hit an inside number, come bet after the comeout roll, not always place odds at the pt., bet the "signature" number.

Good luck at Vegas. My suggestion is to find a very low min. table, bet only the min. on Pass Line, and just practice until you get the confidence.

Bam, best advice you could use is right up here ^

I have thought about conducting a similar experiment, but because of job changes I haven't had time.

First, I'd try to quantify if you have control, and how much control. I would toss the dice 1000 (10k would be better) times. Have the dice in the same set (6's on top, 3/4 on the ends, 2/5 front/back). Don't bother with the number of sevens, or the resulting roll. Every roll is counted in one of 3 tallies. No throwing a roll out. If the dice go off the table, are not a legal roll, or touch the sides of the landing table; they go into tally number 1. If you want even more reality, add some chips on the side or front edge of the landing table. If the dice touch any of the chips, they also go into tally number 1. Tally number 2 is any controlled, legal roll with a 3 or 4 showing on top. Tally number 3 is any controlled, legal roll with no 3 or 4 showing on top.

The reason you don't throw any results out is it unfairly alters your results. If you were at a real casino table and you toss the dice and they land on someone's chips, or bounce to the side of the table, or hit the back wall so hard that they bounce way up the table (off the landing practice table); then these are real rolls that would count. They would be random rolls without any real control, but they still count. So you want to know what percentage of the time you can make a legal toss and have it land with the possibility of control. The ratio of the tally 2 to tally 3 will tell you how often you can keep them "on edge". That is, how often you can keep the side edge from showing. 20 of the 36 rolls will have a 3 or 4 side showing. So a random roll should have a 5 to 4 ratio between tally 2 and 3.

Personally, I think you will be surprised by how big a number is in tally number 1.

It's pretty easy in Laughlin, especially during week days very late at night, to find craps tables that are barren of other players. I'm surprised nobody has utilized this "perfect" scenario to try their dice controlling experiments.

That would involve either driving or flying to Laughlin and staking real money. The combined cost of that would be, from all accounts, greater than the potential profits from anyone who's been posting details about their travails with dice setting. In other words, if anyone out there actually can influence the dice, they're not posting here about it.

Quote: MathExtremist

In other words, if anyone out there actually can influence the dice, they're not posting here about it.

shhh ;)

Quote: MathExtremist

That would involve either driving or flying to Laughlin and staking real money. The combined cost of that would be, from all accounts, greater than the potential profits from anyone who's been posting details about their travails with dice setting. In other words, if anyone out there actually can influence the dice, they're not posting here about it.

True. I guess what I'm saying is that I want to hear someone say they went to Laughlin, staked real money, and performed a dice "influencing" experiment. Even if they lose, I would give them credit for at least having the guts to put their money where their mouth is, and trying to influence the dice in a real world laboratory.

Also, I like the use of the word "influence", as though they are coaxing the dice into certain numbers. "Come on, honies, just three more hard 10s, and we'll all go get a nice steak dinner!"

Have there been any more public/quasi-public dice experiments since the Stanford Wong challenge? Personally I believe it's possible, but not for everyone. Like card counting is a mental exercise for which not everyone has the capacity, controlled throwing is an athletic endeavor that probably isn't for everyone either.

I think the trial was not valid unless a magician was present. Preferably unknown to witness the event!

I think the distinction between card counting and controlled throwing is one of variability. In a depleted blackjack deck, at any given point in time, there is exactly one EV. It is readily determined with a computer based on the cards seen thus far. The various card counting systems are simply shortcuts to knowing to a sufficient degree what that EV is so as to inform a betting decision. By and large, the methods to count cards all correlate with each other because of that single EV. Moreover, there's only one way to bet on a blackjack table -- in the circle. The only choice is how much to bet.

On the other hand, there seem to be as many potential way to set and/or throw and/or bet on the dice as there are dice shooters. I've seen stories about people doing sliding techniques, gentle soft throwing, controlled bouncing, stacked dice, crossed-sixes, parallel sixes, flying-V, you name it. I think controlled throwing is probably more of a mental endeavor than you think -- you have to know how you're altering the dice probabilities so you know how much to bet and where. The fact that the OP, according to his records, demonstrated a very high level of dice control but did not see a profit is a perfect example.

Thanks for posting MrRalph. I find it very interesting. Did you also track the effectiveness of the "hardway" set? Were you able to produce more hardways than random? If so, perhaps your betting strategy should include hardway bets.

I did not do the math so if one of the math guys wants to do the calculations as to wether the number of hardways thrown was above or below random then here are the numbers. 1024 rolls.
31 hard 10's
14 " 4's
24 " 6's
23 " 8's

the software showed using the paralell sixes set I would have had an edge on the hard 6 and 8 but those are not bets I would normally make nor is that a set I have used. I also would not think enough of my ability to duplicate this in a casino enviornment to make the hardway bets. The hardway set is really a seven avoidance set more than a set to actually hit the hardways. There is an explanation in one of the Wizards craps appendix's.
One of my expectations for this test was that using the all seven set would produce a higher SRR than the hardway set because it would take a primary hit to roll a seven and it takes away the dreaded double pitch which produces the seven in the hardway set. The SRR for the all sevens set was only 5.1. This is because my percentage of hardways was 13.28% with random being 11.11%. My double pitch percentage was only 9.61% with random being 11.11%. I am going to do this test again just to compare results after I get back from my trip.

I also agree that card counting and controlled throwing of the dice are not the same thing. In Black jack once the cards are dealt you either have an edge or you do not and it changes with every deal either to your advantage or disadvantage until the shuffle where it goes back to the starting point for that game with it's particular rules. In craps wether you control the dice or not every throw has the same built in house edge for the bets you are placing. I know in my case I am not skilled enough to think that I can throw the dice consistently enough to change the house edge to a player edge. But I can count well enough to know if I have an edge or not. Now having big enough kahuna's to bet your edge is a different story

FYI to all: The hardway set is primarily used to reduce the appearance of the 7. Its secondary objective is to make hardways and inside numbers. The hardway bets are always bad bets with House edges so high that even for a controlled shooter, they are terrible bets.

When you threw those hardways, how many easy numbers and how many sevens did you observe with the same technique?

Here's my theory. You were reducing the seven, but reducing it much more on the comeout than after the point is established, and also decreased the 11 and increased the 2,3,12 on comeouts. That would've changed the 2:1 advantage on the comeout.. if that was lowered significantly to say.. 1.5:1 than that might be enough to end up giving you a negative expectation. ( I haven't done the math).

However, this might influence me to forgoe the come bets and just do place bets.

Why are you combining the totals of both dice in your experiments?

Why don't you simply record the face number on each cube, and then calculate the theoretical based on your results?
It would be far more interesting to view the chi square and the standard deviation graph for the individual outcomes.

Also, if you're testing your throwing abilities, then please tell us that you're are using new dice and not the voided ones.

New dice. Very sharp.