October 31st, 2011 at 8:51:02 PM
permalink

Quote:TheNightflyWiz, did you make any attempt to bet both sides?

No. In my opinion, showing they cheat on the pass side was enough. There is already other evidence they cheat the other way as well.

It's not whether you win or lose; it's whether or not you had a good bet.

October 31st, 2011 at 8:56:31 PM
permalink

Quote:WizardNo. In my opinion, showing they cheat on the pass side was enough. There is already other evidence they cheat the other way as well.

I thought the real question was how their algorithm adjusts when it sees bets in more than one spot. I.e. is it exploitable or not?

"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice."
-- Girolamo Cardano, 1563

October 31st, 2011 at 9:05:47 PM
permalink

Quote:MathExtremistI thought the real question was how their algorithm adjusts when it sees bets in more than one spot. I.e. is it exploitable or not?

I'm more interested in warning others players before they waste their money playing at a rigged casino.

It's not whether you win or lose; it's whether or not you had a good bet.

October 31st, 2011 at 9:11:15 PM
permalink

Greetings all,

I have a cash account on a casino that runs the BLR Tech software.

I first ran 100 rolls on the DONT; my results were consistent that I was playing a rigged version.

I decided to create a simple test to force the software to exhibit an extreme bias. I exclusively played the pass line. If a point was established, I then bought every number 4, 5, 6, 8, 9, 10, meaning that a 7 was the killer number to get if the software was rogue.

In 100 rolls, I established a point on 74 of them. After establishing a point, the first roll was a 7 on 41 out of those 74 points. The expected number is 12.33 with a standard deviation of 3.21, putting this result 8.94 standard deviations above expectation, or about 1-in-5130000000000000000.

Out of the total of 135 rolls made once a point was established, 63 of them were 7. The expected number is 22.5 with a standard deviation of 4.33, putting this 9.35 standard deviations above expectation, or about 1-in-4500000000000000000000.

A chi-squared test on the distribution of rolls after the come out gave a result that was 1-in-142000000000000.

I did not record these results on video.

I strongly advise not playing at any casino that operates this software.

I would like to personally thank the OP Clem for all he did to bring this issue to light. He deserves much praise for his persistence and conviction.

[edit. I lost about $200 in 100 rolls playing this strategy, at $1 per wager]

I have a cash account on a casino that runs the BLR Tech software.

I first ran 100 rolls on the DONT; my results were consistent that I was playing a rigged version.

I decided to create a simple test to force the software to exhibit an extreme bias. I exclusively played the pass line. If a point was established, I then bought every number 4, 5, 6, 8, 9, 10, meaning that a 7 was the killer number to get if the software was rogue.

In 100 rolls, I established a point on 74 of them. After establishing a point, the first roll was a 7 on 41 out of those 74 points. The expected number is 12.33 with a standard deviation of 3.21, putting this result 8.94 standard deviations above expectation, or about 1-in-5130000000000000000.

Out of the total of 135 rolls made once a point was established, 63 of them were 7. The expected number is 22.5 with a standard deviation of 4.33, putting this 9.35 standard deviations above expectation, or about 1-in-4500000000000000000000.

A chi-squared test on the distribution of rolls after the come out gave a result that was 1-in-142000000000000.

I did not record these results on video.

I strongly advise not playing at any casino that operates this software.

I would like to personally thank the OP Clem for all he did to bring this issue to light. He deserves much praise for his persistence and conviction.

[edit. I lost about $200 in 100 rolls playing this strategy, at $1 per wager]

October 31st, 2011 at 9:28:05 PM
permalink

So, doesn't anyone who is a member here have an account with such a casino and feel up to testing whether the biases can be exploited? Perhaps try teliot's technique but lay the 4, 5, 6, 8, 9, and 10 instead of buying them. If the trigger to make the 7 come up after the point is set is just that you have bet the pass line, it shouldn't take long to clean up and recover past losses. Or you can observe whether the software adjusts. However, the fact that teliot saw the 7 on the very first post-come-out roll so often suggests that the software recognized the buy bets in addition to the pass bet; it might adjust the other way for too many lay bets.

October 31st, 2011 at 9:44:44 PM
permalink

Quote:teliotI would like to personally thank the OP Clem for all he did to bring this issue to light. He deserves much praise for his persistence and conviction.

Thanks Eliot for doing your own test and confirming that indeed the software is not playing a fair game of craps.

Agreed, OP Clem is deserving of praise for alerting the public about his very suspicious results. The videos spoke for themselves.

It's not whether you win or lose; it's whether or not you had a good bet.

October 31st, 2011 at 10:00:24 PM
permalink

I knew I was being cheated but my main concern was keeping others from being cheated and closing down the rogue BLR software sites. I certainly found the right site to do this and I appreciate what everyone has done. I also learned a few new terms that I had previously been unfamiliar with "raw data'" being one of them. Thanks to the Rodney King incident I knew there was nothing better than video taping which I believe had I not done this most readers would have been extremely skeptical of my claims.

P.S. I still don't understand what chi-squared is or how it works.

P.S. I still don't understand what chi-squared is or how it works.

October 31st, 2011 at 10:20:16 PM
permalink

So who owned this rigged casino and how much money were they making by running a rigged craps game?

November 1st, 2011 at 6:11:14 AM
permalink

Chi Squared is simply a statistic: It is the sum of ((Observed - Expected)^2)/Expected.

You combine this with the number of degrees of freedom (number of outcomes - 1) to get a measure of the probability (a p-value) of the independent events happening.

In the case of the dice, you add up all of the Chi-Squared values, stick it in a calculator to get the p-value. A fair game will have low Chi Squared Values (<2). An unfair game will look like yours.

Math: Given that the experiment went MORE extreme (7 on 41 rolls of 74) when Eliot placed the numbers, there must be software in place to detect this and act accordingly.

I agree that the rogue casinos should be shut down after some heavy advertisting against them. Given that the Wizard appears near the top of most Google searches (you need more revenue, Wizard!), I think we simply need to open up a thread with the names of the 5 BLR casinos and an announcement that they're rogue. That should get their attention. I really don't think there's any way to exploit this.

BLR's website hasn't been updated in more than 2 years and I wonder if they are still in active business.

You combine this with the number of degrees of freedom (number of outcomes - 1) to get a measure of the probability (a p-value) of the independent events happening.

In the case of the dice, you add up all of the Chi-Squared values, stick it in a calculator to get the p-value. A fair game will have low Chi Squared Values (<2). An unfair game will look like yours.

Math: Given that the experiment went MORE extreme (7 on 41 rolls of 74) when Eliot placed the numbers, there must be software in place to detect this and act accordingly.

I agree that the rogue casinos should be shut down after some heavy advertisting against them. Given that the Wizard appears near the top of most Google searches (you need more revenue, Wizard!), I think we simply need to open up a thread with the names of the 5 BLR casinos and an announcement that they're rogue. That should get their attention. I really don't think there's any way to exploit this.

BLR's website hasn't been updated in more than 2 years and I wonder if they are still in active business.

-----
You want the truth! You can't handle the truth!

November 1st, 2011 at 7:30:46 AM
permalink

Quote:boymimboChi Squared is simply a statistic: It is the sum of ((Observed - Expected)^2)/Expected.

That is correct. However, to dumb it down a bit more, the chi-squared statistic is useful in comparing actual results against expected results when there are multiple outcomes possible. For example, the outcome in roulette or video poker.

It's not whether you win or lose; it's whether or not you had a good bet.