+ Register

[Turner]

Bayes, my cynicism is going to start to irritate you, I am sure


I see this as marketing. Clever marketing....but just marketing. Playing on the fears of gamblers...soft soap.


Its just my way of looking at things.


Probably read and watched too much Charlie Brooker.

[Xander]

"Crooks in the Online Casino Industry"

Posted on January 2, 2014 | 8 Comments -Written by Dr. Elliot Jacobson.


"In January, 2012, I was hired by an online casino to audit the online casino software company that provided their software. This casino wanted to get a Certified Fair Gambling seal. I began the audit in late March, 2012. The software company had previously obtained a TST certification for their RNG, which was sent to me together with other documentation, log files and code slices. TST wrote, in part:

"TST has verified, through mathematical and statistical analysis, within a 95.0% confidence interval, that the RNG outcomes exhibit sufficient non-predictability, fair distribution and lack of bias to particular outcomes.

TST's evaluation was limited to outcome-based testing in the laboratory environment, and was performed using a test version of the RNG. TST's evaluation was based on specific information and materials to be outlined within the forthcoming Final Report (including, but not necessarily limited to, source code, software, hardware, configurations, documentation and general correspondence), as submitted to TST throughout the duration of the evaluation."

What does TST certification really mean?

As I audited the log files from the casino, everything came back normal for Keno and Blackjack. I noted that this company's shuffle was non-standard and needed some repair. I also indicated to them that their data format was substandard. When it came to auditing craps, that's when things started to go wrong. The following are my notes as I audited craps:

Craps (log files received 03/24/2012)
◾RTP = 97.62%. Not clear if I computed it correctly.
◾Audit dice #1 (5 DOF, p-Value 0.10373)
◾Audit dice #2 (5 DOF, p-Value 0.32725)
◾Audit dice total (11 DOF, p-Value 0.33259)
◾Dice total correlation tests (11 DOF, p-Values inconsistent).  7 values under 0.05.
◾Dice total correlation tests other direction (11 DOF, p-Values also inconsistent) 6 values under 0.05.
◾Dice correlation for single dice (dice 1 test) for dice value 4 gives (5 DOF, p-value = 0.000035993). That's 35-in-100000. Other values are ok. (5 DOF, p-Values are inconsistent).
◾Dice correlation for single dice (dice 2 test) for dice value 2 gives (5 DOF, p-value = 0.0056).
◾REQUESTED – code that shows how dice values are being generated.

As you see, I was not happy with the results from the correlation tests. One type of correlation test considers the value of one dice if the value of the other dice is fixed. Another form of correlation test considers consecutive rolls. I wanted more details.

I contacted the software company and was sent a code slice that showed how the RNG was used to produce dice rolls. It was clear that the code was incomplete. As I was making additional requests, I received an e-mail from the lead programmer for the company asking if I had a few minutes to talk by phone. What happened next was truly remarkable. I spoke with the coder who confessed everything to me. Here are the notes I took during that call:

CFG_Audit_Rogue_01

Hopefully you get the general idea. The next day, I spoke with the owner of the software company on the phone. Here are my notes from this call:

CFG_Audit_Rogue_03

Later, I got this email, denying the confession:

CFG_Audit_Rogue_02

The owner had culled data from customers who played the version when it was not in "rogue-mode" and wanted me to look at that data and give craps a "pass." The owner wanted me to ignore the phone call from the software developer. The owner apologized for the bad data that had accidentally been sent to me. The owner expected me to go back, look at new clean logs, and continue as if nothing bad had happened.

I spoke with the programmer and owner several more times by phone. The programmer told me that he had asked the owner to stop lying to me. The programmer told me that the owner indicated that the lies would continue in order to protect the company in question.

I spoke with the owner of the software company again on April 13, 2012. The owner indicated that the software company still wanted CFG certification. The owner said that they had paid 50% in advance for it and expected to be certified. I told them no. No certification. No money back.

Later, I got a letter from someone involved who wanted the audit to continue in the face of everything that had come to pass:

"In case you're wondering why I am still going after this so hard, I really don't want this cheat and liar to get away with it.  XXX is a horrible human being and doesn't deserve to be rewarded for their dishonesty. "

In the end I concluded that the casino was totally surprised and caught off guard. They did not know the software was rogue or had a cheat mode. In my opinion, the software company accidentally left a cheat option in place on a  piece of craps code that was deployed to the online casino.  The casino wanted the CFG seal and expected to get it. Needless to say, they no longer use this casino software.

It is simply not the case that every casino software company produces software that is capable of cheating. I know many companies with integrity. I know many companies that are not crooked. There are good guys out there. Hopefully this account shows just how tough it can get as a game fairness auditor. Protecting the good guys means dealing with the bad guys.

This software company continues to have its product in dozens of online casinos worldwide. I am not sure how many of those casinos use its "cheat mode" or are even aware such a mode exists. I am not free to disclose the name of this casino or software company; both were clients and are protected by confidentiality.

I stand by the CFG seal and the casinos and software companies it represents. I'm also relieved to be done with the industry as a game fairness auditor.

[added 1.10.2014]

There has been quite a bit of discussion on some gambling message boards about why I am not willing to disclose this software provider. I was hired by a casino to audit their software provider so that the casino could post the CFG seal. This relationship was not a public audit and was never intended to be public.

I was not hired by a government agency to conduct a public audit. I was not hired by some third-party to publicly investigate a suspected rogue software product.  I was hired as a private business by a private business to audit another private business.  CFG is not affiliated with any jurisdiction or regulatory agency and has no public disclosure obligation. This relationship and all findings were and are confidential. The only public information would have been permission to use the CFG seal, if it was awarded.  It wasn't. My recourse under the terms of the contract did not provide for a public remedy.

As far as the number of casinos that run this software, I did a quick Google search and found 14 casinos. I am not sure of the exact number of casinos that currently use this software.

One of the greatest sources of security and safety are the players themselves. Every instance I know of a casino software company having their products exposed as rogue in the last few years has been initiated from a player complaint. Complaints by players may be taken lightly by the gaming community without significant evidence to back up their claims. Players who suspect rogue behavior need to keep careful records and be prepared to take proactive steps to back up their case. It is important that players continue to bring their issues forward through established portals, like Casinomeister, thePogg and Wizardofvegas.

No audit is perfect. No audit can be guaranteed to catch malfeasance. But, a rigorous fairness audit is better than no audit at all. It is important that players only choose eCogra, TST and CFG certified casinos. CFG certification, in particular, focuses purely on game fairness. These certifications may not be perfect, but a casino software product that fails to have any of these should be played with caution."  -Dr. Elliot Jacobson

It continues at http://apheat.net/2014/01/02/crooks-in-the-online-casino-industry/


-------------------------------


In short, some online casinos have some cheating code within various gaming software RNG that's designed to cheat you in order to ensure that the casino shows a profit each month.

If you don't believe me, then please take the time to read and research the link that I've provided above.   

-Xander

[Mr J]

Years ago at my casino here, we did not yet have roulette. The closest drive was just under two hours to the next casino. I went there 2-3 times per week, a four hour drive per visit. Keep in mind, this was BEFORE the laws were changed.....at that time, we could bet on-line and I still chose to make the drive.

Great question for you on-line guys >> why would you choose RNG over DublinBet, I don't get it?

Ken

[Tomla]

Live is THE only way! plus you get to watch your cash grow!

[FLAT_IN_O]

Years ago at my casino here, we did not yet have roulette. The closest drive was just under two hours to the next casino. I went there 2-3 times per week, a four hour drive per visit. Keep in mind, this was BEFORE the laws were changed.....at that time, we could bet on-line and I still chose to make the drive.

Great question for you on-line guys >> why would you choose RNG over DublinBet, I don't get it?

Ken

--If you think DB is different,then you wrong.At least I experienced negative adventure.....played for almost
a month,gradually rised my BR 7800 in plus....then one morning couldn't log in,next day and so for about 10
days......somehow one day succeeded,after trillions trys,withdraw my money.....and today can hardly log in
on DB........in other words they just don't let you play when you continously taking em on daily basis.

[Bayes]

I see this as marketing. Clever marketing....but just marketing. Playing on the fears of gamblers...soft soap.

Turner, that's ok, I'm cynical too. But with respect, I don't think you've understood how the hash function works. You probably just glanced over the page and thought to yourself "oh, they're trying to blind me with science - it's a marketing tactic". But it's not the case. The key point to understand is that the number(s) are generated BEFORE you place a bet, not after.

This is how crooked RNG's work; you place your bet, then the software generates a number which makes you lose. Note the order - it has to know first where your bet is before it can make you lose. But in the BV system, the spin is generated before you bet (though obviously not shown to you!), then a hash is taken of the spin and displayed. Next, you place a bet, and BV displays the spin (which was actually generated before you placed your bet), then you have the option to check whether the hash which was generated really does correspond to the spin which was given to you. If it doesn't, it means that the spin was changed after you placed a bet - see how it works? I posted about this a couple of years ago on rouletteforum.cc -

http://www.rouletteforum.cc/index.php?topic=4030.0

Note that the hash algorithm is in the public domain so nothing is hidden from you. I think some people have a problem getting their head around how this works, so they just dismiss it.

[Turner]

Bayes...
Ok..and if I pursue this further.....I have left my remit of proof and have entered the insane world of conspiracy theory...which I am fully aware of.
Not going there.
I accept what you say. After all...I trust random.org