Biased Dice: The science begins
http://www.youtube.com/watch?v=pbGz1njqhxU
You never know what will happen if you look.
Quote: MathExtremistAveraging biases across multiple dice is a flawed methodology.
This is true assuming that there is no bias as a result of the difference in rigidity of the dice as a result of a differing amount of holes drilled and filled with a more rigid material than the transparent portion of each of the casino dice.
It is possible that there is an inherent bias to each cube as a result of the cubes not being symmetrical. The lack of symmetry of the two materials used to create the dice can cause some non-uniform bounce characteristics that are dependent on which corner is making contact with the felt or other surface.
Many people want to close off possibilities because they cannot conceive of them. The conclusion that there is no bias present as a result of the similarities of how casino dice are manufactured has yet to be proven, or even considered, to the best of my knowledge. Therefore the methodology is not flawed.
Talking about a methodology that should, in fact, reveal fairness for fair dice is flawed, is in fact a flawed statement.
If the results come out as fair, and there is no evidence of bias, where is the flaw in doing a study that concludes that the dice being investigated are fair according to samples made by impartial individuals and inspected by experts using well-known statistical analysis?
To be fair to the conspiracy theorists, though, it's important to get the details of the casino, the dice brand, and other details as well.
Quote: WizardSo far I've seen no evidence to support the loaded dice conspiracy.
Nor I. But I have seen some interesting things that have got the brain churning. Perhaps I'll comment once I can sort those thoughts.
Quote: WizardAlso, just out of curiosity, if a casino were inclined to introduce loaded dice to the tables, what would be the revenue-maximizing way to bias them?
That's above my pay grade ;)
Quote: AhighIf the results come out as fair, and there is no evidence of bias, where is the flaw in doing a study that concludes that the dice being investigated are fair according to samples made by impartial individuals and inspected by experts using well-known statistical analysis?
No harm at all. Keep those samples coming. It takes me just seconds to do a chi-squared test.
Quote: AhighMany people want to close off possibilities because they cannot conceive of them. The conclusion that there is no bias present as a result of the similarities of how casino dice are manufactured has yet to be proven, or even considered, to the best of my knowledge. Therefore the methodology is not flawed.
You play with dice all the time, right? Why don't you load a die well past the casino's allowable standards and test it?
I've been thinking of doing it myself for a long time now. Seems like you could easily heat up some fishing split shot and sink it into the acrylic just below the surface and seal it back up again. Perhaps that could give a sort of baseline and idea of just how much bias would be needed to have any real effect.
I'm starting to think even something as severe as that load might be surprising in how little difference it makes...
Quote: WizardAlso, just out of curiosity, if a casino were inclined to introduce loaded dice to the tables, what would be the revenue-maximizing way to bias them?
This was the primary reason for creating the spreadsheet. The answer depends on the common frequency of bets in the casino.
If the field plays triple, and you have big and bold field betters, the best biased dice to use are (12,11,11,11,11,12) dice.
Double/12 Field Edge: -3.2872%
Triple/12 Field Edge: -0.1730%
Passline: -2.4689%
Don't: -0.6453%
Free odds (4,5,6,8,9,10) -0.0433% -0.2811% -0.4239% -0.4239% -0.2811% -0.0433%
If the field pays double on the 12, then (7,6,6,6,6,7) dice produce these edges: (but watch out for big don't players and people hopping the hi/lo)
Double/12 Field Edge: -1.5235%
Triple/12 Field Edge: +1.8698%
Passline: -3.3615%
Don't: -0.0318%
Free odds: -0.1385% -0.5540% -0.8033% -0.8033% -0.5540% -0.1385%
Boxcars: +5.19%
Aces: +5.19%
You can also risk that there are no big field betters with (7,6,6,6,6,7) dice
But in general, the 9,8,8,8,8,9 dice are the best way to go. If you have an intelligent player, they won't bet the field or the hi/lo and you could gun for the 7,6,6,6,6,7 dice.
Also there are a few casinos using 6 cubes instead of 5. If you were to intentionally bias every other die to the ace or six face, it's a little more complex for the face, but you're more likely to pair up a cube heavy on the 6 with another die heavy on the ace compared to using sticks of 5.
The Gold Coast and California Casino are two casinos using 6 dice instead of 5.
Also, Wizard, I encourage you to answer that question. I think it's interesting that you're asking the question. My thought in response to the conspiracy guys going back years and years is "you know you're going to create exposure somewhere no matter how you bias them, right?"
The more I have looked into this, though, the more I realize that there is opportunity to exploit the less frequently made bet (going against the grain).
Silverton: Bee (reported not to use Paulson any longer)
Wynn: Paulson, Midwest
Boyd Casinos: TK, Midwest
Stations Casinos: Bee
The M: Bud Jones, Bee
Lucky Club: Bee
Quote: Wizard... takes me just seconds...
=p
This demonstrates how many samples are necessary to get to sufficiently small enough p-values to demonstrate that the dice are fair for comparison to theoretically biased dice.
When you consider that biased dice (assuming that they exist) can be put into and out of play selectively (according to the biased dice conspiracy theorists), the conspiracy theorists may never accept that the dice are (always) fair, I'm afraid. Even the most ardent theorists believe that fair dice are out there and used. If you sample fair dice 9 out of 10 times, and biased dice the rest of the time, it will take a whole lot more samples to prove that there are biased dice out there .. EVEN IF they are all biased the same way.
Nevertheless, if nothing else, this demonstrates that we need more than just a few samples to come to any conclusions.
If a very heavily biased set of dice were used for ten hours and 1000 throws, that's 2000 samples, and it's still conceivable that the same outcomes resulting from biased dice could be from fair dice (due to the small number of samples).
Quote: MathExtremistIt relies on a pretty big assumption, namely that all dice have the same bias. If not true, which is exceptionally likely, the conclusions are invalid. Averaging biases across multiple dice is a flawed methodology.
I would suggest that if there is a bias in dice, it could vary from individual die to individual die in the case of worn dice.
It would be easier to demonstrate if the dice were manufactured that way -- so certainly if the "bias dice folks" really want to investigate they should start with the "new dice" fresh out of the box. At least that would offer a control, of sorts. There can be no control with "used dice."
Let me just add this: Even if the dice were heavily loaded so that one face always shows on top, on every subsequent roll you'd still have a 1/6 chance of rolling a seven..
I would like to see the "load experiment" made. I think we have all taken the five dice that are presented to us, and put them in our hand and spilled them out of the table to see which faces showed 'up." Has anyone done that several times to find the same face up on multiple dice? I never have. They all appear random enough to me from the get-go. But carry on if you think you are liable to find something.
Quote: AhighIf you knew for fact that the dice were biased enough to profit from the bias, the best way to do it would be on the free bets.
This spreadsheet allows you to effectively combat such theoretically biased dice on the fly. All of this is just theory, and as has been pointed out the chance that enough bias is present to take advantage of is impractical unless the bias is very strong.
Sounds a little bit like a strategy......... Taking past history of the dice to make a decision on future betting situations. Maybe I read it wrong
Quote: AhighA simulation could even use the equations to determine exactly how strong a bias would be necessary for a procedural method to be effective at battling such theoretically biased dice.
IE: it should be obvious that 100% bias to the aces would cause aces every roll, and after you rolled aces 12 times in a row, anyone could hop the aces and parlay into retirement.
Nope, I read it right.... It's a strategy..
I hope this meta-analysis does not end up in the wrong hands of a writer.. It has just enough math to con a novice player into ACTUALLY believing it is true.
Quote: AhighThis is true assuming that there is no bias as a result of the difference in rigidity of the dice as a result of a differing amount of holes drilled and filled with a more rigid material than the transparent portion of each of the casino dice.
No, it's true always. Aggregating data from multiple sources destroys information unless you already know that all of your sources are equivalent. If you have six dice and you don't know whether or how they're biased, just rolling two at a time and counting total faces is not a valid exercise. Consider the reductio ad absurdum: each of the six dice is 100% loaded toward one face, but only one die is loaded for each face. Thus, the first die always shows 1, the second die always shows 2, etc. Each roll, pick two dice at random and roll them. Do this 300 times and tally results according to your method. Because you do not distinguish your six dice and evaluate them separately, your face-counting method reveals that your six-dice set is perfectly fair: you will observe roughly 100 occurrences of each face. But the dice aren't fair, so the data collection method is flawed.
.jpeg "Bohemian"){kind=small}
Quote: JimboBUT--here--the actual dice combinations were NOT recorded.
And--here--Ahigh utilized his "Excel spreadsheet" to arrive at a number of conclusions beyond whether there is simply suspected bias towards future rolls.
My issue is that Ahigh has NOT relied on the dice combinations--but only the individual die faces--in order to determine the house edge for various dice combinations.
Ahigh also determined the results of Pass Line and Don't Pass Line bets, based ONLY on the recording of the individual die faces.
Nothing about dice combinations. Nothing about the order or sequence of the individual die faces. Nothing about the order or sequence of dice combinations.
I still don't understand how that is possible.
Jimbo, I agree.
I have been taking advantage of biased dice for several years now. I document 2 dice combinations and attack accordingly. I don't wait for 10,000 rolls before I attack.
Key issues to understand with most biased dice in play:
1. Results should be obtained from random rollers as good dice influencers can affect your database.
2. Dice that hit the glass/plexiglass, stack of chips or puck should also be eliminated from your database.
3. Hard wild throws also affect your database.
4. Softer the toss, the better the results for your database. THIS INCLUDES DICE INFLUENCERS IN TRAINING, that TRY to throw with a soft precision, but are better off not trying yet as their toss has very little if any influence.
5. As noted above, assuming that all dice in a stick are biased the same way is not always correct. Verify.
For example, The D in downtown Las Vegas is currently using biased dice with the following characteristics. These dice combinations are logged from Serial number 269 in May, 2013.
| Dice Result | Expected Count | Sample #1 | Tossed % | % Difference |
|---|---|---|---|---|
| 2 | 3 | 4 | 3.7% | 0.9% |
| 3 | 6 | 10 | 9.3% | 3.7% |
| 4 | 9 | 8 | -0.9% | -0.9% |
| 5 | 12 | 12 | 0.0% | 0.0% |
| 6 | 15 | 25 | 9.3% | 9.3% |
| 7 | 18 | 15 | -2.8% | -2.8% |
| 8 | 15 | 5 | -9.3% | -9.3% |
| 9 | 12 | 12 | 0.0% | 0.0% |
| 10 | 9 | 11 | 1.9% | 1.9% |
| 11 | 6 | 5 | -0.9% | -0.9% |
| 12 | 3 | 1 | -1.9% | -1.9% |
| Total | 108 | 108 |
Notes - this database is slightly tainted as it does include a good dice influencer that has somewhat skewed the results.
Also, there were no Hard 8s and all 8s came as a 6/2.
This type of biasness at the D is not the same at every casino that uses biased dice from time to time, but is consistent with several Las Vegas casinos (they may use the same dice factory or at least the same type of loads to get similar results?)
I have accounted for that in my betting strategy below. There may be many other ways to approach and win with biased dice, but this is how I do it given these facts at the D:
Bet the Aces and 3s every roll.
Lay the 8 on every roll except when a good dice influencer is tossing.
Bet the 6 on every roll except the Come-out.
Quote: BohemianI have been taking advantage of biased dice for several years now. I document 2 dice combinations and attack accordingly. I don't wait for 10,000 rolls before I attack.
108 rolls? How long did you wait before you attacked?
Quote: BohemianKey issues to understand with most biased dice in play:
1. Results should be obtained from random rollers as good dice influencers can affect your database.
2. Dice that hit the glass/plexiglass, stack of chips or puck should also be eliminated from your database.
3. Hard wild throws also affect your database.
4. Softer the toss, the better the results for your database. THIS INCLUDES DICE INFLUENCERS IN TRAINING, that TRY to throw with a soft precision, but are better off not trying yet as their toss has very little if any influence.
5. As noted above, assuming that all dice in a stick are biased the same way is not always correct. Verify.
For example, The D in downtown Las Vegas is currently using biased dice with the following characteristics. These dice combinations are logged from Serial number 269 in May, 2013.
this database is slightly tainted as it does include a good dice influencer that has somewhat skewed the results.
You can't set conditions such as "results should be obtained from random rollers" and that "dice influencers can affect your database" and then go ahead to quote results and add "this database is slightly tainted as it does include a good dice influencer that has somewhat skewed the results."
This is not scientific. You are giving us observations with perhaps a pinch of a hunch thrown in. Also, you can't make up the rules as you go along. And that also seems to be how this "investigation" is proceeding. There are a lot of problems here folks.
I suggest you go back to square one and come up with a goal, a method, and rules for the investigation and research. You have to limit the variables and there are just too many of them including different casinos, different brands, different throws, different dice, new vs used dice, dice that may be damaged, dice that have been in use for longer periods of time, dice that hit the back wall or mirror or chips vs those that don't hit anything and just roll.
What you will come up with is an alphabet soup of results and reasons and a "report" that will not hold water.
This investigation is already a 7-out.
Quote: AlanMendelsonYou can't set conditions such as "results should be obtained from random rollers" and that "dice influencers can affect your database" and then go ahead to quote results and add "this database is slightly tainted as it does include a good dice influencer that has somewhat skewed the results."
Certainly, the tosses must be random
Quote: AlanMendelsonYou have to limit the variables and there are just too many of them including different casinos, different brands, different throws, different dice, new vs used dice, dice that may be damaged, dice that have been in use for longer periods of time, dice that hit the back wall or mirror or chips vs those that don't hit anything and just roll.
I think that most of the variables that you mentioned have been addressed by others.
As was discussed, the casino should be noted as should the particular brand of dice (if possible). Of course, damaged dice would not be reliable.
In regards to the other variables that you mention, they are all irrelevant as long as the tosses are random. A toss that bounces off of the stickman's nose and then cracks the mirror and knocks over a stack of dealer chips is just as random as the typical finger snapper toss that bounces off of the back wall and comes to rest two feet away.
Quote: AlanMendelsonWhat you will come up with is an alphabet soup of results and reasons and a "report" that will not hold water. This investigation is already a 7-out.
Are you jumping to conclusions again? Sounds like you are overreacting to a single poster who doesn't represent the entire "investigation."
Quote: tuppAre you jumping to conclusions again? Sounds like you are overreacting to a single poster who doesn't represent the entire "investigation."
the entire "investigation" is not scientific. And that's the bottom line.
Quote: AlanMendelsonthe entire "investigation" is not scientific. And that's the bottom line.
All of the variables that you suggest are uncontrolled have just been addressed and eliminated. Actually, those variables were also discussed in this forum before you mentioned them.
Please stop overreacting and jumping to conclusions.
Quote: Bohemian
Dice Result Expected Count Sample #1 Tossed % % Difference 2 3 4 3.7% 0.9% 3 6 10 9.3% 3.7% 4 9 8 -0.9% -0.9% 5 12 12 0.0% 0.0% 6 15 25 9.3% 9.3% 7 18 15 -2.8% -2.8% 8 15 5 -9.3% -9.3% 9 12 12 0.0% 0.0% 10 9 11 1.9% 1.9% 11 6 5 -0.9% -0.9% 12 3 1 -1.9% -1.9% Total 108 108
p value = 4.17%
As I've said too many times, your various counts of rolls and faces would only be circumstantial and not conclusive.
If there is a load in a die, the die is examined to find the load. It is not rolled to determine if there is a load. This counting of rolls is an elementrary school science experiment and nothing more.
Take a look at what one dice maker says of its process: http://midwestgamesupply.com/dice_manufacturing.htm
Perhaps those who believe dice are biased should be asking each manufacturer about the safeguards it uses to guarantee precision. By the way, there are several web posts by manufacturers who state that the weight of the pip material matches the weight of the die material that is removed for the pips. Did Harley bother to check with the manufacturers or did he just make his own assumption?
Quote: MathExtremistNo, it's true always. Aggregating data from multiple sources destroys information unless you already know that all of your sources are equivalent. If you have six dice and you don't know whether or how they're biased, just rolling two at a time and counting total faces is not a valid exercise. Consider the reductio ad absurdum: each of the six dice is 100% loaded toward one face, but only one die is loaded for each face. Thus, the first die always shows 1, the second die always shows 2, etc. Each roll, pick two dice at random and roll them. Do this 300 times and tally results according to your method. Because you do not distinguish your six dice and evaluate them separately, your face-counting method reveals that your six-dice set is perfectly fair: you will observe roughly 100 occurrences of each face. But the dice aren't fair, so the data collection method is flawed.
MathExtremist: If each and every dice ever manufactured is biased due to the difference in rigidity resulting from the asymmetry of the hardened epoxy resin relative to the non-epoxy transparent portion of the dice, you're wrong in saying this is a flawed method of study.
Additionally, there is a desire to look at actual face outcomes from people who are not partial on the topic of biased dice. I mean even if Bohemian aka Harley came in here with a bunch of data, would anyone even believe him? We already know what this guy has to say about the subject.
Now I am not saying that you don't know a better way to collect data. But if you do have a better way to collect data that is easily explained to Face and anyone else who is signing up to collect data, share it with us. But to come in here and say that what is being done is 100% flawed is just simply wrong, MathExtremist.
Also counter to your suggestions that this effort is futile, ANY deviation from random expectations, regardless of what the cause was, and even for a small sample is of relevance to the discussion, especially from someone who already is not predisposed to a particular answer. The interesting thing to me is that it seems that people who believe that the dice are all 100% fair are the most predisposed to a particular answer. So let the data continue to flow in.
Once we get to 10,000 samples, either the dice will be within the expected distribution or they won't. I will agree that if they are within the expected distribution it tells us very little at all. But if they are not in the expected distribution for fair dice, yet they are in the expected distribution of dice that have more ace and more six faces appearing, that will be another story entirely.
If you're suggesting that after we get 10,000 to 100,000 samples and we have a tiny p-value for randomly distributed dice and another p-value for dice that would yield better holds for casinos, I want to then hear your comments about how the methodology was flawed.
I think 5-2 would be the best bias possible for the casino. Most ploppies will bet 6-8 and the pass line. Biasing a die towards a 5 and 2 would decrease the Aces and 12 wins. Very few people bet the 10 or 4.
Quote: AhighQuote: MathExtremistNo, it's true always. Aggregating data from multiple sources destroys information unless you already know that all of your sources are equivalent. If you have six dice and you don't know whether or how they're biased, just rolling two at a time and counting total faces is not a valid exercise. Consider the reductio ad absurdum: each of the six dice is 100% loaded toward one face, but only one die is loaded for each face. Thus, the first die always shows 1, the second die always shows 2, etc. Each roll, pick two dice at random and roll them. Do this 300 times and tally results according to your method. Because you do not distinguish your six dice and evaluate them separately, your face-counting method reveals that your six-dice set is perfectly fair: you will observe roughly 100 occurrences of each face. But the dice aren't fair, so the data collection method is flawed.
MathExtremist: If each and every dice ever manufactured is biased due to the difference in rigidity resulting from the asymmetry of the hardened epoxy resin relative to the non-epoxy transparent portion of the dice, you're wrong in saying this is a flawed method of study.
Ahigh, you just realized that you said exactly what MathExtremist said: "unless you already know that all of your sources are equivalent". You just acknowledged that.
If the bias is different within a set of die for each die, then aggregating results does destroy information. The data collection method is flawed.
But it's a start.
Quote: boymimboThe wizard asks a brilliant question. The best way to load dice for the casinos benefit is obviously a roll set that gets you to a seven out. You would have to then bias dice differently to do so. Is it possible to get opposite faces on all dices to bias? That is, a dice to bias TOWARDS say a 5 and 2?
I think 5-2 would be the best bias possible for the casino. Most ploppies will bet 6-8 and the pass line. Biasing a die towards a 5 and 2 would decrease the Aces and 12 wins. Very few people bet the 10 or 4.
Actually, a simulation would be required to determine the correct answer. And part of the input for the simulation would be what is the average bet amount for each bet on the felt at that casino.
In addition, having each die biased differently would be required to get the best possible results, and the number of dice being presented to the player (usually I see five, but sometimes six in the case of Boyd owned properties lately).
There are a lot of parameters to such a simulation with so many bets on the felt, and the theoretical possibility to weight each of the six sides of the dice differently an infinite number of ways.
Yet still, given that the edges on the horn bets are higher, biasing to produce more horn numbers is the most advantageous way to bias the dice as you are creating a bias that cannot easily be exploited. Specifically biasing towards the high and the low .. the only true check and balance is the field when paying triple on the 12 or on the 2. Otherwise, it's a clear and obvious answer that bias towards the six and one faces is the way the casino would like things to go in order to create a bias that cannot be easily exploited by the player.
If you're still wondering about this question, and you haven't yet downloaded the spreadsheet, you should. Otherwise we can talk about this quite a bit instead of just looking at the data very clearly that will easily allow you to model various theoretically biased outcomes to see which biases benefit the casino the most.
As far as the whole argument from ME, the thing is that there are two cases: casinos are benefiting from biased outcomes, or they aren't. If they are, we will see heavy ace faces and heavy six faces. If they aren't, we will see (what we already see) people making very stupid bets on the craps felt and losing and blaming the dice.
Quote: AlanMendelsonI think the "scientists" would be better off investigating the manufacturing process of casino dice than counting faces and rolls.
Somebody call Doc.
Quote: AhighQuote: MathExtremistNo, it's true always. Aggregating data from multiple sources destroys information unless you already know that all of your sources are equivalent.
If each and every dice ever manufactured is biased due to the difference in rigidity resulting from the asymmetry of the hardened epoxy resin relative to the non-epoxy transparent portion of the dice, you're wrong in saying this is a flawed method of study.
Even that unfounded leap of faith doesn't get you there: you're still not ruling out any other biases, nor are you accounting for variations in the bias you assume is there in all dice. The only way you can conclude that all dice are equivalent, and therefore justify your aggregation scheme, would be if each die were biased in *exactly the same way* due to the manufacturing process, and there is *no other bias* in any die from any other source.
The problem with that approach is that it assumes your conclusion. It's circular. Moreover, if you already know that the dice are only biased in one way, and that all dice are biased in exactly the same way, then you don't need to be doing this study at all. Just test one die and extrapolate the results to all the other dice on the planet which are necessarily, per your assumptions, equivalent.
On the other hand, if you're honest, you will admit that you have no idea whether dice are biased, and if so, how they are biased and to what degree those biases vary between dice manufacturers and even individual dice. By assuming that all dice are equivalent, you're setting yourself up for an easy refutation: all anyone has to do is demonstrate that two dice are not equivalent and your entire result set gets thrown out.
If you want to do this the right way, you need to collect data for individual dice. Not lumped together across different casinos and different manufacturers. Results for each individual die must be recorded separately.
Quote:If you're suggesting that after we get 10,000 to 100,000 samples and we have a tiny p-value for randomly distributed dice and another p-value for dice that would yield better holds for casinos, I want to then hear your comments about how the methodology was flawed.
We don't need to wait that long. There are an awful lot of possibilities for how an aggregated data set like the one you're collecting could show a bias toward the 1 and 6 faces. Suppose one casino used cheating loads heavily favoring the 1 and 6 faces, and every other casino used fair dice. By aggregating the results like you're doing, and including the cheating casino in your dataset, you might observe a slight overall bias toward 1 and 6 and incorrectly assume that all dice at all casinos are slightly biased. That wouldn't be the truth, but it would be the result you'd obtain with your methodology. A flawed methodology leads to invalid results.
Maybe if instead of telling me why what I was doing was wrong you could tell me how you would assert that, in general, the casinos are NOT using biased dice, how would you go about doing that?
As far as my conclusion, I fail to understand how I have a foregone conclusion. All that I have done is to understand what type of bias would even ALLOW for a greater hold. I personally think that just offering bets with higher edges is plenty sufficient to increase the holds.
I just watched a guy lose $5,000 at Mandalay Bay this morning, and he started off with $50 on each hardway. Then he did $300 across excluding. He was betting odds up the max odds, but only on the pass line. He was also betting $25 yo with a $5 dealer yo. I walked away when he lost most of his money, but the point is that this guy doesn't need biased dice to take more money from what should be free odds. He was having no problem losing his money on sucker bets.
If you think that I'm here as part of the biased dice crew and just trying to come up with some manufactured data to say "hey look the dice are biased" I think you just have the wrong idea.
I have ALWAYS wanted to come to a conclusion that they are NOT biased. Unfortunately, I don't have any data to assert that at all. The Wizard is participating only under the condition that the samples are recorded from casino environments. And I'm pretty sure that Face is impartial, but what do I know? I am just doing the best I can here.
If you want to come up with ways to make it better, do that! If you want to say it's not good enough, and even WHY it's not good enough, I think we got the point already. Coming up with a way to improve the process of inspecting possible bias from real casino dice in real casinos would be more valuable.
I've already made the point that associating the data for dice manufacturer, casino, date and time collected, and all that stuff is very helpful.
You got anything else to add for how to improve things, or just want to point out more problems?
Quote: AhighMaybe if instead of telling me why what I was doing was wrong you could tell me how you would assert that, in general, the casinos are NOT using biased dice, how would you go about doing that?
Quote: MathExtremistIf you want to do this the right way, you need to collect data for individual dice. Not lumped together across different casinos and different manufacturers. Results for each individual die must be recorded separately.
Quote: MathExtremistIf you want to do this the right way, you need to collect data for individual dice. Not lumped together across different casinos and different manufacturers. Results for each individual die must be recorded separately.
Unfortunately, the Wizard would not be participating if we proceeded this way. He wants real casino conditions. So that's not an option. Maybe if the Wizard could explain how this would not invalidate his requirements for participation. Otherwise that's not a useful suggestion.
Quote: AhighUnfortunately, the Wizard would not be participating if we proceeded this way. He wants real casino conditions. So that's not an option. Maybe if the Wizard could explain how this would not invalidate his requirements for participation. Otherwise that's not a useful suggestion.
Another consideration is to determine what you really want to test for. If you want to show that All dice from one manufacturer are flawed, there's a test for that. If you want to show that all dice at a casino are flawed, there's a test for that. If you want to show that all dice everywhere are flawed, there's a test for that, too. But they're all slightly different tests, with their own flaws and burdens. It's probably possible to build one data set that would satisfy all three. You'd need to:
Note the casino where each trial occurred
Note the manufacturer of the dice in play
track each die separately (this will be very hard to do accurately, but it's important)
Then go record hundreds of sessions of thousands of rolls at each of a few dozen casinos and cross-tabulate your data across casinos and across dice vendors.
Not for nothing: ME's method is superior to the current plan. ME and the Wiz are equally capable of analyzing your data.
But if they are following the expected distribution for dice that would lead to an increased hold for a casino, it will take a lot of samples to see that.
Quote: AhighUnfortunately, the Wizard would not be participating if we proceeded this way. He wants real casino conditions. So that's not an option. Maybe if the Wizard could explain how this would not invalidate his requirements for participation. Otherwise that's not a useful suggestion.
I didn't say anything about not collecting data in a casino. You just need to distinguish between each die.
Quote: MathExtremistYou just need to distinguish between each die.
just, huh? "Just" tell Face!
1. If the dice are biased, is it deliberate or accidental in nature?
2. If the dice are deliberately biased, is each die equally biased?
3. Assuming the biased dice are weighted differently to produce bias in the total (perhaps with more sevens), how does the casino get around the fact that the player can choose any two dice from a set of 5 or 6?
Also, lest anybody jump into the middle of this thread and think I've become a conspiracy-theory believing quack, I am posing these questions mainly in the interests that we might learn something about the scientific method of proving/disproving a theory.
I still don't understand how just taking someones word for it makes zero sense.
Quote: AhighMaybe if instead of telling me why what I was doing was wrong you could tell me how you would assert that, in general, the casinos are NOT using biased dice, how would you go about doing that?
How about this: just say "I've never seen any evidence of biased dice."
Quote: AhighAs far as my conclusion, I fail to understand how I have a foregone conclusion.
Well, you've certainly talked about "biased dice" affecting your rolls, haven't you? Didn't you do an entire TV show about why the dice were biased and you were throwing so many sevens? It seemed that you were pinning your lack of "control" on the dice because of the bias.
Quote: AhighIf you think that I'm here as part of the biased dice crew and just trying to come up with some manufactured data to say "hey look the dice are biased" I think you just have the wrong idea.
This is a new one. I think we've all had the idea you thought the dice were biased. Didn't you report on allegedly biased dice at a certain casino?
Quote: AhighI have ALWAYS wanted to come to a conclusion that they are NOT biased. Unfortunately, I don't have any data to assert that at all.
How about this: how about coming up with some proof that dice are biased before indicting and libeling an entire industry including the dice manufacturers? Frankly, I wouldn't be surprised if their lawyers aren't already monitoring this site. Just google "dice bias" and this thread comes up in the results along with some specific pages.
Quote: AhighI am just doing the best I can here.
If you want to come up with ways to make it better, do that! If you want to say it's not good enough, and even WHY it's not good enough, I think we got the point already. Coming up with a way to improve the process of inspecting possible bias from real casino dice in real casinos would be more valuable.
I've already made the point that associating the data for dice manufacturer, casino, date and time collected, and all that stuff is very helpful.
You got anything else to add for how to improve things, or just want to point out more problems?
Unfortunately casino conditions are not really laboratory conditions. For one thing there are no controls on the quality of the dice, the normal wear on the dice, and especially how the dice are thrown. There are no controls on the record keeping, or the particular dice used. Was there a bad batch? Was a die worn or did a die become sticky because some gal kissed them with her lipstick?
Casino conditions are probably the worst place to gather test data.
Quote: WizardI think what this topic is crying out for is a more finely-tuned hypothesis. It would make confirming/denying the biased dice theory easier if we could narrow down how the dice are allegedly biased. Here are some questions to help in the formation of a more specific hypothesis.
1. If the dice are biased, is it deliberate or accidental in nature?
2. If the dice are deliberately biased, is each die equally biased?
3. Assuming the biased dice are weighted differently to produce bias in the total (perhaps with more sevens), how does the casino get around the fact that the player can choose any two dice from a set of 5 or 6?
Also, lest anybody jump into the middle of this thread and think I've become a conspiracy-theory believing quack, I am posing these questions mainly in the interests that we might learn something about the scientific method of proving/disproving a theory.
Wizard, I'm sorry, but recording rolls in a casino is not going to prove any of this.
Quote: AhighAlan, you need to get your facts straight. Wow.
List my errors, one by one, please.
Quote: AlanMendelsonList my errors, one by one, please.
How about research your statements before you tell lies.
I don't have time.
Quote: AhighHow about research your statements before you tell lies.
I don't have time.
Which "lie" was it?
--- That you didn't allege that a particular casino had biased dice?
--- That you didn't do a TV show blaming biased dice for throwing too many 7s?
I'm waiting. And at this point, after you named several manufacturers in addition to at least one casino, it wouldn't surprise me if their lawyers are also waiting.
Quote: AlanMendelsonWhich "lie" was it?
--- That you didn't allege that a particular casino had biased dice?
--- That you didn't do a TV show blaming biased dice for throwing too many 7s?
I'm waiting. And at this point, after you named several manufacturers in addition to at least one casino, it wouldn't surprise me if their lawyers are also waiting.
You back up your statements. Not my job to tell you what you wrote that's bullshit. As best as I can tell, ignorance is an ally of yours that enables you to tell lies without remorse.
Listen for a minute on this you tube link where I give my explanation for why there were so many sevens...
http://youtu.be/a3Hodn7fNUY?t=9m48s
And how about citing where I alleged anything? Your claiming I alleged any casino was doing anything illegal is completely without merit. The only thing that I did was count faces right in front of everyone at the table, dealers, everyone. And I presented the data and I did say the data supported the conspiracy theorists. Nothing at all any different from what Face is now doing except his counts don't support the conspiracy theorists. If you go back to the thread titled, "Biased Dice: The Saga Continues" go ahead and quote where I alleged anything. Maybe I'm wrong! But personally I think you are just full of it. I also think you're a trouble-maker with these ignorant comments implying things that are patently false from my view.
Quote: AhighFace does deserve a big round of applause. Thanks for doing the heavy lifting and keeping the data represented by someone who is impartial towards the outcome.
You're welcome. But, just so we have full disclosure, I think the whole "biased dice" idea holds no merit, and the idea that casinos intentionally switch them in as laughable at best. But there are certain things inside this debate that have piqued my curiosity, so I figured I might as well help push it along. I find it easy to be objective and have no fears of being proved wrong.
Quote: AxelWolfNever mind if it is a good experiment, scientific or not. I would never trust random people reporting results. It only takes one person with an agenda to skew the results.
The only thing I can say to your repeated posts of apprehension is to trust to my integrity. I’ve been a member here for a while and have had plenty of interactions with most of the senior members. While I’ve certainly argued passionately with many and could provide several examples where I’ve been proved wrong, I’ve never intentionally misled or failed to admit when I was wrong. You have my permission either publicly or by IM to question anyone here on my integrity.
The only way these results would not be 100% true to life as they happened would be due to simple human error. I am doing my best to control that, and I imagine only a fraction of a percent might be affected.
I don't have to tell you about your own program which appears to be the start of this campaign of yours -- to blame biased dice on your inability to control the dice.
You even made allegations that at the Mandalay Bay that one brand of dice appeared to be "fair" when removed from the table, and was that an allegation that the new brand of dice brought to the table because of "advantage players" was not fair? See: https://wizardofvegas.com/forum/gambling/craps/13148-bad-dice-the-saga-continues/3/
Are you investigating or are you indicting? Do you have an open mind or are you trying to prove your allegations? Do you have proof or is this libel?
Quote: AlanMendelsonWhich "lie" was it?
Shawn Hannity on the radio is always being called
a liar and his response is always "Yeah? Name just
one lie I told." And they never do, they can't. They
respond like Ahigh did, they stumble and stammer
and say they don't have time to list all of them. First
Ahigh said we were all wrong all the time, now we
lie all the time. Its baloney, so of course he never
backs anything up.
Quote: AlanMendelsonLet's start with your allegation about the Fiesta. You will find that here: https://wizardofvegas.com/forum/gambling/craps/13148-bad-dice-the-saga-continues/
I don't have to tell you about your own program which appears to be the start of this campaign of yours -- to blame biased dice on your inability to control the dice.
You even made allegations that at the Mandalay Bay that one brand of dice appeared to be "fair" when removed from the table, and was that an allegation that the new brand of dice brought to the table because of "advantage players" was not fair? See: https://wizardofvegas.com/forum/gambling/craps/13148-bad-dice-the-saga-continues/3/
Are you investigating or are you indicting? Do you have an open mind or are you trying to prove your allegations? Do you have proof or is this libel?
You have a comprehension problem. All I said about Mandalay Bay counts was that they all looked fair. Where are the posts where I alleged anything at all was happening that was illegal? Where is it? I just don't see it. Allegation is a very strong word, Mr. Mendelson.
Quote: EvenBobShawn Hannity on the radio is always being called
a liar and his response is always "Yeah? Name just
one lie I told." And they never do, they can't. They
respond like Ahigh did, they stumble and stammer
and say they don't have time to list all of them. First
Ahigh said we were all wrong all the time, now we
lie all the time. Its baloney, so of course he never
backs anything up.
Yeah, you have all kinds of credibility. It should be pretty obvious Alan is just making up stuff the way that he remembers it to support how he feels about me. I don't blame him for disliking me. The feeling is mutual for sure. For someone who reports, he sure has a problem with getting his facts straight.
