Mathematical Proof that English Harbour is cheating

[spearmaster]

I can see that in this casino a casino might choose to make a quick denial that there was anything wrong, and that's not necessarily indicative of guilt, merely undue haste and poor behaviour in making a statement not backed by the appropriate actions to confirm its veracity.

HOWEVER, the fact remains that given that the problem entered the software for whatever reason (and of course we cannot say for sure what the intent was), it was identified and removed from the software in the short period between it being publicly proven that English Harbour's video poker was ripping off the players and their statement being issued.

Quick denial would be accurate. Poorly-thought out denial would also be accurate.

I didn't go back and check to see if this denial came after they fixed the bug or not, I will assume you were right. But it is also possible that a denial was prepared based on some test - perhaps even an incorrect test - and sent for delivery before the actual bug was discovered.

Example: "Dewey Wins!" For non-Americans, or those of you who haven't dug that deep into American history, the New York Post printed and delivered an edition with this headline referring to the winner of the 1948 presidential election - or at least who they thought had won. Unfortunately for them, Harry Truman won the election and served as President.

Shit happens. I am not going to make any excuses for EH or why they even made that statement - but it happened, and we all knew it couldn't be true, and if they had wanted to cover that up they would not have gone so far as to have an audit done, with the result that they admitted an error and refunded bets at 120%...

The fact they fixed the code so it played fair, something they MUST have known they had done, and immediately after that code was fixed issued a statement saying nothing was wrong does not look good to me.

Sure you or they can claim that the left hand did not know what the right hand was doing: that someone had found out it was broken (and that's assuming we actually believe that somehow the code could accidentally come to shortchange the player in the first place) and quietly fixed it, while at exactly the same time this was being fixed, the PR/legal/fairplay department issued a pseudo-mathematical statement saying there was nothing wrong, while not knowing what the coders were doing.

Believe it or not, I am reasonably sure this is what happened. But again, that does not excuse EH.

So for us to accept that the first statement should just be ignored and is not an indicator of malfeasance, we have to accept the following postulates:

1. That English Harbour could cause a game based on a player card being lower than the dealer card to deal lower cards to the player more often than is random, by some coding error (i.e. not deliberately trying to cheat the player)
2. That the error having been introduced into a game that does not appear outwardly to have changed and in theory should not have had its internal code changed at all, someone would go back to that game, check the code, identify the problem, and fix the bug
3. That the bug fix would be scheduled to be deployed in the tiny period between the public proof of foul play and the English Harbour statement, and the even tinier period where English Harbour could have 'reviewed their game play'
4. That having "concluded our review of the game play and randomness for all Video Poker games" English Harbour had no idea that the 'bug' had just been fixed, even though they had just fixed it, and must have had code check in logs, etc.

I can't go into much detail here. I did know that a statement was to be released, I did not know what it contained (and was very unpleasantly surprised). I do know what they believed at the time, even though I told them that this was clearly an erroneous belief and that they had better make certain their mathematicians - NOT their lawyers - checked the statement before it was issued.

Suffice it to say, I was probably more shocked than the rest of you when that first statement was issued, as you can ascertain from my response at the time.

After my reply, I went to bed - and when I woke up the next morning, I discovered that a full audit would be taking place because it was essential to get the facts. By this time, however, the problem had been fixed.

So yes, I basically believe that your 4 postulates are correct. It does not take long for a bug to be fixed if you are the coder - and of course it does not take long for you to change code back to what it was if you were the coder. Time is not an issue in modifying small pieces of code, though discovering the problem may take a bit more time if it was not deliberate.

Sure if you think all that's plausible, but to me given the circumstances, I don't think that it makes sense to take what they say at face value. There is prima facie evidence of cheating, and that should be enough to make any statement coming from them suspicious at best.

Prima facie evidence of a non-random game, if you don't mind. But I certainly wouldn't blame you or anyone else if you were suspicious of the statements... I was sure as hell suspicious of that first statement myself...

And we still haven't heard why this 'bug' didn't seem to affect Hot Pepper and Fire and Ice casinos.

We'll try to get to the bottom of that. But as you know, EH is the main client of Odds On, and sometimes they get stuff in advance. And sometimes they get stuff that other operators don't get.

None of the other software manufacturers are any different - nothing is instantly deployed at all operations at the same time precisely because they must be sure there are not any unusual problems - and even then problems do slip by - again, at any software manufacturer. I've personally spotted and reported problems with MGS software way back in 2000 and 2001 - mind you, they were rules-based problems (place bets not off by default on come-out roll, aces not being counted as high card in pai gow poker) - and to be honest these were just small slip-ups or different interpretations of the rules. Even MGS autoplay didn't always use correct strategy for their games.

Shit happens. Just let us figure out what the hell went wrong in this particular case - I must say I'm not all that hopeful but again, as a moderator, and for that matter as a middleman, it is incumbent on me to give every side a fair hearing.

 

[Linus]

Gah, this whole thing pisses me off.

Yesterday I had a bad run at BJ on Crypto. Of course these things happen, but before I would be satisfied that it was bad luck. Now this little voice is nagging in my head "What if they pulled an EH?".

This whole thing has really reduced my enjoyment and trust level for all online gambling, not just EH.

English Harbour got caught because the adjustment to the VP game was so dramatic.

If it had been a small adjustment, or if it had been applied to a big payoff game (where it might take thousands, or hundreds of thousands, of trials to prove it wasn't paying off correctly), it might have gone on indefinitely.

I don't think it's safe to assume EH is the only casino that does this, simply because it's the only one to be caught (recently).

Mr. Shackleton blacklisted Elka/Oyster after a 200-game roulette trial. MGM software was BL'd after 700 hands of Video Poker. And of course, Casino Bar was listed after 332 game trial of Black Jack. (That one was similar to the English Harbour case, in that the game became "fair" again after the results of the original trial were reported.)

In none of those cases was there any more "proof" that the adjustments to the odds were "intentional" - other than the mathematical proof that the games were in fact not fair.

To a player, I'm not sure it matters "why" a game is not fair.

Especially since it's hard to imagine how "proof" that cheating is intentional could ever be obtained.

Whether it's intentional, or just inexplicable coincidence the net result is the same for players.

If a casino can't deal a fair game - for whatever reason - it shouldn't be recommended.

Casinos have no reason to deal fair games, except the fear of getting caught. If they can get caught, and still not get punished, they have no reason at all.

Res Ipsa Loquitur - the thing speaks for itself.

 

[erp1]

Any software that has a win/loss parameter on a card game is by definition cheating software.

I agree 100%

I don't see why you are interested in the win/loss parameter hypothesis, as it doesn't help English Harbour at all.

I don't take the obvious for given. I ask for facts and based on a critical analyze of facts I draw my conclusions.

It's the same method I used when I posted in the TrueGambler affair as greekvik.

Based on the facts that is presented until now I dont see how a response of my 3 questions will not make my conclude that the software provider belong to the same category as Elka/Oyster, Gambling Software Systems (GSS), COA World Entertainment (Casino Bar) etc.

What is less clear to me at this time is the role and responsibility of the casinos that use the software. Just to mention some of the questions: Who knew that the software had an option to create biased results? Is it a new or an old feature? Who requested such a feature?

 

[dirk_dangerous]

Well said Gamemaster. There is no reason for anyone to ever trust 0dds-on (or whatever it's called these days) software again. Period.

Excuses are nothing but BS. No quarter should be given to any software or casino that screws up in this manner

I'm sure that Bryan, Brian and Ted, plus others here will remember the Gambling Software Systems (GSS) scandal of a few years ago. For those that don't, GSS software was shown to have a "switch" in it that could change the odds of their Video Poker games by making a non-random deal. The online gambling community quickly became aware of this situation and eventually GSS went out of business, although some casinos using that software may have survived by switching to other software platforms.

The evidence presented here shows that Odds On software (or whatever its official name is) contains a similar "switch", be it in the form of a glitch or a deliberate feature. In any event, it obviously caused non-random results to occur and was repaired, which implies that their programmers now know how it happened - which further implies it could happen again.

The gambling public has no choice - in my humble opinion - but to adopt a "zero tolerance" attitude towards software that fails to produce random results. If software screws up - for whatever reason - it should be boycotted, period, amen. Sure, glitches can happen to any program, but with all of the other choices out there, why should any producer of casino software be forgiven? Is the online gambler to demonstrate compassion toward those who are clearly shown to have non-random results at any time?

That's not to say English Harbor itself shouldn't survive; perhaps the glitch wasn't their fault and it's true that they've done the "right thing" by making retitution to the players involved. But it seems to me that for them to survive, they must switch software platforms asap and then work to develop their business by relying upon their fine reputation, which has taken years to establish.

Online gamblers control very few aspects of their chosen form of recreation, but choice is one of them and it's the most powerful. We are all now faced with a choice. As for me, I will never advertise or recommend or play at a casino that uses Odds On software from this point forward. I hope many others will join me in this.

GM

 

[jeremiahsjohnso]

Now THIS is interesting. It is mathematically plain wrong. #wins/#losses will converge to 100% given a fair game. It is a realistic mistake for someone not used to mathematics. And the data posted by English Harbour EXCACTLY match this 50% ratio (1/3 wins and 2/3 looses)...

And given the casino had a parameter to configure for this ratio, just 1 number (the ratio) is the obvious way to make it. It would be stupid to have two % fields where you type 50% in one and 50% in the other.

The more I think about this, the more I see this as an unintentional rigging. Again, if EH wanted to screw people over, switching the doubling to 48% success would be a much better way to do it. This was too damn obvious.

And the fact that it went to 1/3 points out exactly what Zoozie suggests above -- someone got into the software and punched in "wins/losses = 1/2", thinking that they were being fair but actually changing the odds dramatically.

Note:
1) This does not justify that there should be a switch where you can alter the win/loss ratio

2) I'm still not fully satisfied by the EH responses (and their first response should be pretty much disowned). Though the refunds and open auditing were obviously a good start.

 

[nafanny29]

Just a thought, do they not fully test software for errors before releasing it, especially as theres huge amounts of cash at stake. I would have thought that they would test, test and test again just to make sure there was not an error in the PLAYERS favour.

And depending on what else was upgraded when the bug was put in the software maybe there are other errors still undetected.

Anyway I cannot believe that they didnt run the softwore through at least 100000 times in test mode to make sure there were no payout errors etc. Surely this is standard practice in this industry, isnt it

 

[Simmo!]

We however all hold our personal opinion but it seems we cannot express it here without facing suspension.

LOL. So what you're saying is that it's perfectly okay to slander someone else on someone else's website? If you want to slander someone, set up your own site and take some responsibility for your actions.

The problem is that you and a few others don't think about the ramifications for others before you state your opinions. Personally I think Meister is actually pretty lenient when it comes to allowing posters to provide subjective opinion but some people just aren't satisfied with that, they feel they have the right to mouth off without the slightest concern for who it affects.

Most posters here realise that there are ways to state opinions that can more than get a point across without resorting to slanderous comments, thankfully. You are far more likely to get snowballed for a lack or respect than for stating an opinion...and not just here, but in life in general.

 

[tronk]

I agree that there are many ways to get your point across. I also understand that this site has to play the delvils advocate and i do think that what can be, and is discussed, is liberal. What i said probably wasnt fair, but it wasnt meant as criticism so much a comment on caruso getting banned. I apologise if i came across direspectfully to casinomeister.

 

[Simmo!]

I agree that there are many ways to get your point across. I also understand that this site has to play the delvils advocate and i do think that what can be, and is discussed, is liberal. What i said probably wasnt fair, but it wasnt meant as criticism so much a comment on caruso getting banned. I apologise if i came across direspectfully to casinomeister.

No worries - and I apologise if i was on my soapbox for a moment there It's just that I think this board stands head and shoulders above many others because it manages to keep things in perspective and has a majority of respectful posters who do it, and themselves, credit.

 

[spearmaster]

Tronk -

I actually missed that comment - but as Simmo has said above, we respect everyone's opinions here, provided that you also respect other opinions and understand that there must be a point where a line needs to be drawn.

Also, I have already pointed out that we will be talking to the management shortly to find out what happened - and that any other slanging or accusations or whatever would not help - in fact, could even aggravate or compromise the situation.

Requests were made in advance numerous times - they were not heeded, which is the only reason that stronger action is being taken.

I'm not going to reiterate any more than is necessary - I will simply start giving out prizes.

 

[kengam]

Was this the first time that this "bug" could be found in this software? As a professional player for many years I can tell you that I have thought this software to be dealing biased double ups a couple times in the past. The double up feature, as most of you know, was heavily utilized by pros because it satisfied wagering requirements wihtout having a house edge. Myself and others collectively had data that suggested that the double up was not being dealt fairly. Our results weren't as bad as evidenced in April, but bad enough for a group of bright mathematically inclined individuals to abandon the practice of using the double up feature.

If I had to guess, the win/loss frequency can be set to unfair levels. This "switch" as most like to call it, is not always off and not always on. As a group we have noted periods of time where it was at an acceptable level and then periods of time when it seemed skewed. Had we known for sure that play mode and real mode were dealt the same we would have saved a whole bunch of money trying to determine if it was dealing properly.

I have not been to this site in quite a while, but I was drawn to this thread because I have extensive play using the double up and certainly have had my doubts at different times along the way. I am reasonably certain that April was not the first time that this "bug" was causing the play on double up to be skewed. My peers and I had discovered this a few other times over the last couple of years. I believe this "bug" to be their "solution" to the problem and profit taking of advantaged professional players. There are/were enough of us to take a decent chunk of their profits away.

 

[soflat]

Our results weren't as bad as evidenced in April, but bad enough for a group of bright mathematically inclined individuals to abandon the practice of using the double up feature.

That might be something that could be looked into. How bad were your results? and when?

 

[Linus]

Spearmaster - English Harbour only has three options.

They can say that the trials were invalid, that there was a "bug," or they can admit cheating.

#1 has been tried and rejected. They will never ever openly admit to intentional cheating.

They've settled on the "bug" defense, and now they have to stick with it. No amount of talking or meeting with them will change that.

You may have reasons for not wanting people to say they're "cheating," but it can't have anything to do with negotiations with English Harbour.

They've placed their bets (and rolled the dice). It's too late to choose another number now.

 

[soflat]

Spearmaster - English Harbour only has three options.

They can say that the trials were invalid, that there was a "bug," or they can admit cheating.

#1 has been tried and rejected. They will never ever openly admit to intentional cheating.

They've settled on the "bug" defense, and now they have to stick with it. No amount of talking or meeting with them will change that.

You may have reasons for not wanting people to say they're "cheating," but it can't have anything to do with negotiations with English Harbour.

They've placed their bets (and rolled the dice). It's too late to choose another number now.

If it was just an isolated incident that only happened two weeks during the casino's entire existence, and the casino is open about sharing its logs with the WoO for any period of time, and they explained what the bug was, then I would tend to believe it was just a bug.

Maybe they are meeting and negotiating with EH to get full audits done or something like that.

In other words, maybe they are waiting for all the evidence to be examined.

 

[spearmaster]

I am waiting for a technical explanation of what happened, plain and simple. If it is not forthcoming soon, there won't be much doubt in my mind as to what happened. But keep in mind that they were honest enough to give you the logs and to admit the problem even if they did muck up that first statement.

 

[kengam]

That might be something that could be looked into. How bad were your results? and when?

bad enough to be 95% confident that the results were not within reasonable range. (3.5 standard deviations)


I cannot release any statistics as I have confidential agreements with the other players involved. I am posting merely to state that this isn't the first time that the double up game was out of whack. This was also the case on no less than 4 other occasions besides this past April. If they were to release all their information regarding double up in the past 3 years, many more episodes would come to light. I do not expect this to happen.

 

[spearmaster]

I guess my only question is - if you had a reasonable belief (95%) that the doubling game was not fair, why didn't you come out with this data in the first place? I would've thought 95% should have been plenty good enough to raise some doubts...

 

[soflat]

How hard would it be to go back through the last two years of double-up stats and see if the claims of this professional bonus hunter are valid?

 

[kengam]

I guess my only question is - if you had a reasonable belief (95%) that the doubling game was not fair, why didn't you come out with this data in the first place? I would've thought 95% should have been plenty good enough to raise some doubts...

The old kengam would have tried as I am sure you know. The new kengam doesn't have the time or desire to fight against thieves in this industry anymore. While this event does not surprise me, a big part of me is sad and irate about it. I have been a long time supporter of the fairness of the big software providers. It has come to that time that even the mighty are tempted to increase profits and strike back at the thorn that has been in their sides since the beginning. I feel like this "bug" is a personal attack on me and people like me who have been making money off of advertised promotions (see bonus hunter). Unfortunately other "innocent" players get trapped in this snare as well.

 

[spearmaster]

LOL... I kinda liked the old kengam, even if we didn't always agree...

I sure hope that it doesn't turn out that the fix was deliberate... there is absolutely no need for this in the industry, and I sure as hell won't support it.

Anyhow, let's see what pans out. All I can do is try my best... as I'm sure others like Meister and Jetset will continue to do.

 

[Threeler]

Blast from the past.

I played on their site in March for a total of about a thousand hands. I'm not entirely certain what the statistics were, but lets just say I busted. I have a lot of friends who play around at different online casinos and they've all noticed issues with their software in regards to doubling. The major complaint was with range of cards. For example, aside from the regular ace or paint as an up card, any mid card seemed to leave you drawing dead and the low ones were rarely better than a push. I lost about $700 (bonus included) and know at least eight other people who can say the same, while I know a couple people who about broke even with the bonus (down just al ittle, but that's the point of bonuses) and ONE person who was up (via a royal which I've still never hit online... ah well). With the positive expected value of the bonus, this seems a bit off, to a point where me and my group of gambler friends here at college quickly put it on the "do not play" list along with some other sites which can be found in the rogue gallery here. There's about ten of us who know to avoid this site, but some insisted it wasn't rigged and in fact instantly assume no online casino is. After reading this, I've e-mailed them to audit their stats from March and I guess I'll tool around with them. The Wizard says they're legit, but I'm a bit curious anyway, hell, even if they fix it once a week per month, well, that's still bogus (word of a nice fat bonus spreads quickly so I wouldn't be surprised to hear we all did it within a period of two or three days).

 

[dovmin]

Should I send an email asking for my money back if I played their doubling Videopoker in the past

How much is past, if t was half a year ago, can I do that ?

Should they return the deposit ?

 

[GrandMaster]

English Harbour sent me the logs.

My tally was very accurate except that I missed one of the losses. This makes things slightly worse for English Harbour.

The exact results were 19 ties, 84 wins, 152 losses.

I attach the results in CSV format.

As it seems that English Harbour have full logs and are handing them over, this looks very bad for them.

Has anyone else got their logs from EH? I would like to do some statistical analysis on them.

 

[Tony_Ross]

Apples & Oranges

However my point is, where does it say that doubling has to be a fair game at online casinos?

Because at BEST you get a fair game, but there is no guarantee. There are strict rules that VP games/Dices has to be fair, but are there similar rules for doubling?
Zoozie

To address this, lets just say I am well read enough in Mathematics to offer a solution to your question and the original poster. "are there similar rules for doubling?"

Yes and no. Yes as far as any gaming commission or governing body is concerned. In other words randomization standards are met and certified every 6 months.

No in the fact that you stand to loose almost every time. To quote,"the fluctuations go mostly in the favor of the house. The house has to win above the standard deviation, always! Or especially when player's bet is high! That's anything but fairness! Static progression betting has virtually NO mathematical validity."

You do not have permission to view link Log in or register now.

First, the casino considers a push as a win for the player, because the casino did not win. This outcome is never neutral for the casino software.

Secondly with a little math, one can see the odds just in a game of 50/50, that a double win is reduced to 1/4. However the degree of certainty just in a 50/50 game can take up to 10 bets just to get to a 99.99% degree of certainty(DC) to produce a win.

Care to estimate how many bets it takes to get a double win ( probability of an event occurring twice in a row)? I highly recommend the following link for free analytical software and education in gambling strategies.

You do not have permission to view link Log in or register now.

Finally, know this: Random outcomes are computed over a million iterations and are still considered "legal" by gaming commissions. These bodies also profit from the casino's revenues, so you can be damn sure they are in place for industry to profit.

<the rest has been snipped as it is not relevant and more like shilling>.


Sincerely, Tony Ross

 

[jetset]

Grandmaster posted:

QUOTE: Has anyone else got their logs from EH? I would like to do some statistical analysis on them. UNQUOTE

I had forgotten momentarily that Grandmaster is an ideal member to involve in this issue; he is totally independent, always straight with his views, appears to have expert qualifications and knowledge in this field and is respected by all the posters I know.

EH would be well-advised to contact him if they are sincere about clearing their reputation from the allegations and speculation that have been so much a part of this mess.

I would very much like to see this in addition to Spear's (programming) expertise in order to have a balanced and fair picture.

 

[thelawnet]

We were expecting a proper explanation as to how this happened. At what point are we to assume that none will come and add thse guys to the rogue section?

It can't be too hardd if this realy did happen by accident to find out why and post it here. They need too come forward and come forward quickly.

 

[nafanny29]

It can't be too hardd if this realy did happen by accident to find out why and post it here. They need too come forward and come forward quickly.

If they managed to fix it they must know what was wrong to be able to correct the error

thats fix as in "mend it" everyone, honest

 

[spearmaster]

I am studying the source code right now, but not likely to have any answers shortly as I am heading for the airport in three hours.

 

[thelawnet]

I am studying the source code right now, but not likely to have any answers shortly as I am heading for the airport in three hours.

I would be interested to see the broken code as well.

 

[chalupa]

I have a background in software engineering as well and I can't fathom how the software could be accidentally broken to cause the results experienced.

The obvious way to program the doubling game would be to deal out 5 random cards using the exact same code found in the main game, which was presumably working fine.

The nearly perfect 2:1 ratio of player losses:wins is also a bit too coincidental.

Finally, I can't help wonder why the delay in complete detailed disclosure. They obviously found the accidental problem some time ago, and should have been able to explain it fully and satisfactorily immediately to limit damage to their credibility.


So call me cynical but... if I was (hypothetically) a casino/software provider with millions of dollars at stake and got busted for a non-random game, what would I do? Well, after making sure my island hideaway was stocked with beer and ammo , I'd probably stall a few days to allow my programmers to crank out some plausible-looking source code to explain the "accidental" bug.

Which brings me to a question, when did the source code get released? And are they going to release it for full public disclosure? I know that I personally would never touch any casino running this software again without personally examining it or a number of experts examining it (no disrespect to Spearmaster's experience intended.)

Regardless, I probably will never touch any of those casinos again anyway, being the cynic that I am, because again even if the source code fully explains the problem, we have no assurance that it was the actually the code that was running at the time of the skewed results.

In fact even if I could somehow be 100% convinced there was NO intentional misbehavior, I would still not have confidence in their technical ability to run a fair and correct game in the future.

If they can make a MAJOR screwup that goes on for days and has to be caught by the PLAYERS, not the casino, what about a minor screwup? How would players ever know that a slot jackpot didn't pay off quite as often as it should? How do we know such "minor" screwups don't exist right now?

 

[chalupa]

As an aside... this is just one more reason the pending US legislation trying to outlaw online wagering is ridiculous.

How much better would it be to allow an online casino to set up in the US where it could be regulated (and taxed)?

Where would YOU rather wager serious money... an unregulated offshore casino with at least one incident of a major error in the casino's favor... or one run in Nevada where if a similar thing happened the gaming commission would immediately shut them down and descend on them armed with subpoenas and geeks?

In fact what the US should do if they were REALLY enlightened (haha)... is to revamp the ridiculously complex tax laws for consumer gambling, and in fact remove all consumer taxes on US based gambling, while keeping the tax (and penalties for non-compliance) on offshore gambling winnings. Then everyone would play US casinos tax-free, and the gov't could tax the easily-collectable revenues on the casinos, and everybody wins.

Well... the US gov't wins, anyway. The players still lose (as usual).

 

[sirius]

It is anyone's guess how the software managed to do what it did to deal an unfair game but it is very unlikely to impossible that it was accidental.
Someone posted earlier some probabilities for 'dealing seconds' that would match but I believe the probabilities they calculated are completely incorrect.

One thing I noticed that nobody has picked up on was that the Wiz and English Harbour claim it was corrected in a software update on 2nd May but when you look at the stats given it looks like it was only in April that the game was unfair. The May 1st stats look fine and are actually a few thousand to one chance of being 'unfair' but apparently they were supposedly still using the unfair code at the time. It could have been a lucky day though.

 

[kengam]

I am suggesting that we give EH until the end of today before we start an all out war against them on all the boards. IMO they shouldn't be granted the opportunity to come up with an excuse. This is merely out of respect for those working on this. Spearmaster if you have the source code, that will not do you any good unless you have the source code for the buggy version.

Cheats in the industry are bad for the industry altogether. This is also very bad for the US's chances of ever legalizing and possibly regulating online gambling. At what point can we declare them to be cheats? Reasonable time has passed and if moderators want to ban people for calling what they see as cheating then it certainly is not going to look good for them. I encourage everyone to begin posting their thoughts on this matter on every message board they have access to after the day has passed unless we have a reasonable answer to what has happened.

 

[spearmaster]

Slow down, guys...

What I have is the buggy version which was introduced. It's a bit more complicated than I expected but I can also see what may have happened - the only thing really is to determine whether everything adds up or not. But I'll need a little time - I've just arrived in LA and obviously jetlagged... and those damn in-seat power systems didn't like my computer so I couldn't power up during the flight either.

I have obtained this code obviously on a confidential basis - but I am urging them to release at least one more copy to someone else experienced in debugging code since I would feel more comfortable having someone else to corroborate my findings (or correcting my mistakes LOL).

 

[The Dude]

...I have obtained this code obviously on a confidential basis - but I am urging them to release at least one more copy to someone else experienced in debugging code since I would feel more comfortable having someone else to corroborate my findings (or correcting my mistakes LOL).

Grandmaster - are you game?

 

[Vesuvio]

Slow down, guys...

Ok, I'll suspend my disbelief just a little longer - if they can find a way out of this one Houdini would be proud.

In light of some of the other posts above I hope English Harbour will be quick to post equivalent figures to the ones given for doubling going back over at least the last year or two.

 

[chalupa]

I have obtained this code obviously on a confidential basis - but I am urging them to release at least one more copy to someone else experienced in debugging code...

They should be able to release the relevant section of the code publicly. Video poker game logic is very straightforward, any decent programmer could create it easily from scratch, so they aren't given away proprietary secrets (well, unless maybe a secret cheating method, haha). And assuming they did their homework at all in properly designing a client/server there would be no security risk in releasing it.

But again, unfortunately source code released by them on their own terms and time schedule doesn't actually prove anything about what was actually running on their server when players were being shorted.

I also still don't understand why they couldn't have LONG ago explained the bug in plain language understandable by non-engineers. It's not rocket science here.

 

[erp1]

I have obtained this code obviously on a confidential basis - but I am urging them to release at least one more copy to someone else experienced in debugging code since I would feel more comfortable having someone else to corroborate my findings (or correcting my mistakes LOL).

It would be of great value to have an evaluation of possible explanations of the character of the so called bug that is based on facts and analyzed by knowledgeable players.

I think a good way to valuate the bug would be to form an independent committee of players with a professional background in software engineering to study a detailed technical explanation from English Harbour and Odds On including the source code.

 

[spearmaster]

They should be able to release the relevant section of the code publicly. Video poker game logic is very straightforward, any decent programmer could create it easily from scratch, so they aren't given away proprietary secrets (well, unless maybe a secret cheating method, haha). And assuming they did their homework at all in properly designing a client/server there would be no security risk in releasing it.

I agree. I think it ought to be that way with every software package. Unfortunately, there are always going to be some bits that are proprietary, from how they handle and generate the RNG, to the method of displaying graphics... whatever...

But again, unfortunately source code released by them on their own terms and time schedule doesn't actually prove anything about what was actually running on their server when players were being shorted.

I also still don't understand why they couldn't have LONG ago explained the bug in plain language understandable by non-engineers. It's not rocket science here.

No one can say for 100% certainty that this was the code that they were running at the time, though it does look like it to me. Just that my concentration at the moment is still lacking... but in any case I'd still prefer to have a second person looking at it.

The bug itself appears to be straightforward, but the sequence of events that sets it off is quite complex. This is what I am trying to sort through now. It is something that can be explained in simple terms but would be difficult to believe without going into great detail - this is why I have to retrace the whole set of procedures.

 

[spearmaster]

It would be of great value to have an evaluation of possible explanations of the character of the so called bug that is based on facts and analyzed by knowledgeable players.

I think a good way to valuate the bug would be to form an independent committee of players with a professional background in software engineering to study a detailed technical explanation from English Harbour and Odds On including the source code.

For all intents and purposes, in this case I would meet your description. And yes I would like at least one other. But I think an entire committee would be a bit hard for them to swallow... and I listened to two phone calls with detailed descriptions probably totalling about 90 minutes, with lots of questions interspersed. The explanation, in my case, was not sufficient without the code because it was "slightly" incredible because the error is too, for lack of a better term, "idiotic" - and entirely the type of error I make in my own programming as well. You know, like bonehead coding errors... the ones which make you slap your head when you discover them...

 

[nektar4d]

couldn't resist to repeat! so funny

Ok, I'll suspend my disbelief just a little longer - if they can find a way out of this one Houdini would be proud.

 

[chalupa]

Unfortunately, there are always going to be some bits that are proprietary, from how they handle and generate the RNG, to the method of displaying graphics... whatever...

Sorry, I don't buy this. It's just a video poker game, and not even one of the nicer ones. Overall it should be a pretty pedestrian piece of code... nothing that would be worth stealing by a competitor.

There is/was a HUGE problem in their software and they should be going over the top with disclosure, and they haven't.

Yes, the guts of the RNG may be proprietary, but they don't need to (and presumably did not) release the code for the RNG itself. Rather, the code for the game would just have calls to some RNG functions.

In my humble opinion, they are already too little too late in fully addressing this. Personally, I won't be visiting them again regardless of the eventual outcome.

 

[spearmaster]

Sorry, I don't buy this. It's just a video poker game, and not even one of the nicer ones. Overall it should be a pretty pedestrian piece of code... nothing that would be worth stealing by a competitor.

For the most part, it's exactly as you would imagine - standard code. And I fully agree that it isn't something I would really think is worth protecting to the extent that most manufacturers do.

Nevertheless, they do it... every single manufacturer out there - so you can hardly blame EH for doing what everyone else does.

There is/was a HUGE problem in their software and they should be going over the top with disclosure, and they haven't.

Yes, the guts of the RNG may be proprietary, but they don't need to (and presumably did not) release the code for the RNG itself. Rather, the code for the game would just have calls to some RNG functions.

In my humble opinion, they are already too little too late in fully addressing this. Personally, I won't be visiting them again regardless of the eventual outcome.

I understand that something is forthcoming shortly from another person. My brain is still in scrambled egg mode so forgive me for being a bit on the slow side - but it makes no sense for me to release any findings on my part until I understand exactly what happened - and though the picture I am seeing appears to be quite clear, I need to ensure that everything adds up, naturally. So now I accept part of the blame for being a bit slow...

 

[erp1]

For all intents and purposes, in this case I would meet your description. And yes I would like at least one other.

You are certainly one of the persons I want to join such a committee.

The explanation, in my case, was not sufficient without the code because it was "slightly" incredible because the error is too, for lack of a better term, "idiotic"

To me there are two aspects:
a) an explanation of the error that caused the not fair results in the second half of April.
b) The fairness of a software that might produce the kind of outcome we have witnessed - whether the bias was intentionally or not.

 

[Vesuvio]

To me there are two aspects:
a) an explanation of the error that caused the not fair results in the second half of April.
b) The fairness of a software that might produce the kind of outcome we have witnessed - whether the bias was intentionally or not.

Some other things to look at:
1) Why were they altering the code for this aspect of video poker in the first place?

2) How plausible would it have been to make the mistake claimed?

3) Is this more or less plausible than the other obvious explanation - that they've made up some code or an excuse of a bug after the fact? We need to bear in mind how unlikely it is that a software update would change a perfectly functioning aspect of the casino - and coincidentally the precise one which bonus hunters were using to profit from bonuses (if other games were also affected it would be more believable).

I trust Spearmaster to give a fair verdict on the code he's been provided with, but establishing the possibility of a bug doesn't in itself get us very far in deciding whether or not the casino deliberately manipulated results in their favour.

Two other hypotheses to have a look at: 1) Could EH have intended to weight doubling in their favour but have accidentally made it much more significant than planned? 2) That this has happened in the past as well (which would almost certainly invalidate any defence based on the code). We need to demand figures for doubling results going back over a year or two and not just for April/May of this year.

 

[spearmaster]

Answer to 1) - They appear to have been testing code for a new bonus game that would make the doubling a bit more exciting - this code was not supposed to have been released or live, but they forgot to comment out these lines and subsequently the routines went live while still incomplete and thus threw errors which were also not properly trapped.

2) and 3) - this is exactly what I am looking for, whether or not this was plausible - and this is also what is taking me a bit of time to uncover.

Hypotheses - 1) possible but far from definite. 2) I'm not testing for the past, and as no evidence has been put forward it would be a waste of time - though I can certainly understand why some people are concerned about the past.

One more thing I should point out - this does not appear to be a Houdini - it appears more likely to be the work of a sloppy programmer (like me, for example) plus very poor control over software handling and updating procedures. Despite all the noise you are hearing from other quarters, I have no interest in misleading anyone, and have no vested interest in a result either way. I do, however, want everyone to know, once and for all, whether this was an unintended bug, or whether EH intended to cheat players in which case I will personally go out of my way to ensure that justice is done.

 

[Zoozie]

Answer to 1) - They appear to have been testing code for a new bonus game that would make the doubling a bit more exciting - this code was not supposed to have been released or live, but they forgot to comment out these lines and subsequently the routines went live while still incomplete and thus threw errors which were also not properly trapped.

Thanks for reporting the 'progress' in this case. However I believe the community are getting very impatient for the fully explanation of this. When will there be a press release and by whom? Are WoD still on the case?

For your explanation I know you are just reporting partial findings, but lets say a certain line of code throws an error. In C++/Java such an error
would have to be 'catched' or the program would terminate. Btw, can you reveal if the server code is indeed programmed in C++?

Normally programmers would log this or have some kind of notification, so if this goes unoticed for so long it normally means some messy code or framework. Lets hope this dev team are not getting hired by the military for developement of an intercontinental ballistic missile defense system...

I belive the exception is 'caught' then. This normally means that some variables might not have been set correct. A good example would be
that you lost every double even though you picked a higher card. The server would never set the 'win=true' variable. However the client
are dealt the cards and the server handles the lose/win correct according to the selected card. So I can still not see how this can happen if a
line of code threw an error. I think you would need an error that is capable of (seems very unlikely) one of the following:

1) The 5 cards are not random. (can you confirm the server does generate such 5 cards btw?)
2) The server can switch the card you picked by another of the 4. (this is what a real life cheating dealer would do )

Zoozie

 

[spearmaster]

Thanks for reporting the 'progress' in this case. However I believe the community are getting very impatient for the fully explanation of this. When will there be a press release and by whom? Are WoD still on the case?

I don't know of any press releases, but there will certainly be more information coming out.

For your explanation I know you are just reporting partial findings, but lets say a certain line of code through an error. In C++/Java such an error
would have to be 'catched' or the program would terminate. Btw, can you reveal if the server code is indeed programmed in C++?

Well, my word was "trapped" while yours was "catched" - functionally equivalent. Errors were not properly trapped, invalid result returned and again at that stage not trapped.

Normally programmers would log this or have some kind of notification, so if this goes unoticed for so long it normally means some messy code or framework. Lets hope this dev team are not getting hired by the military for developement of an intercontinental ballistic missile defense system...

I don't think messy is the case here - sloppy and incomplete is probably a better description. Another appropriate description would be "comedy of errors".

I belive the exception is 'caught' then. This normally means that some variables might not have been set correct. A good example would be
that you lost every double even though you picked a higher card. The server would never set the 'win=true' variable.

As strange as it seems - you are very close to what appears to have happened.

However the client are dealt the cards and the server handles the lose/win correct according to the selected card. So I can still not see how this can happen if a line of code threw an error. I think you would need an error that is capable of (seems very unlikely) one of the following:

In at least two instances - without being specific - errors should have been thrown and execution terminated. It's a bit late at the moment so I can't recall exactly how many.

1) The 5 cards are not random. (can you confirm the server does generate such 5 cards btw?)

Yes.

2) The server can switch the card you picked by another of the 4. (this is what a real life cheating dealer would do )

There appears to have been something which did in fact change the result of the draw - not deliberately, but through the incompleteness and sloppiness described above. And the change was not necessarily bad - and in fact not necessarily predictable, because the result of the change could still result in a win.

I will be elaborating on this once I have finished studying what I need to. One might say that I need to be able to "tell the story" in such a way that it is easy for everyone to understand. And I have to be sure that the story "told" to me by the code matches up with the "story" told by the data - otherwise I will determine that the code is not what was in place at the time the problem occurred.

So again I ask for a little patience. The truth must, and will, come out.

 

[Vesuvio]

Thanks for the detailed response.

Answer to 1) - They appear to have been testing code for a new bonus game that would make the doubling a bit more exciting

Similarly to Zoozie's post I'd just be curious to know why making doubling "more exciting" should involve changing the code for the way cards are dealt. Surely the only changes would be to graphics, the delay before revealing the result and similar visual features? How likely is it in general that a casino suddenly decides to change one particular aspect of their video poker game to make it "more exciting". A complete aesthetic overhaul of the game or a change to the pay table would be hugely more plausible than a change only to a feature that was being exploited by bonus hunters. (I suppose it's possible they were planning to get rid of doubling, which was costing them money, and implement another game with a house edge in its place. That in itself would raise questions about the software company unless they were going to make it very clear that they'd actually removed rather than changed doubling)

2) I'm not testing for the past, and as no evidence has been put forward it would be a waste of time - though I can certainly understand why some people are concerned about the past.

Of course you can't test for past results. All I'm saying is that in the light of people commenting that they suspected something was wrong with the game in the past we should be given the figures going back over a year or two. If they show a similar pattern of sub 50% returns for the player we'd have conclusive proof of cheating without having to argue about the ins and outs of the code. If the casino won't produce these figures that would be extremely suspicious.

 

[spearmaster]

How likely is it in general that a casino suddenly decides to change one particular aspect of their video poker game to make it "more exciting". A complete aesthetic overhaul of the game or a change to the pay table would be hugely more plausible than a change only to a feature that was being exploited by bonus hunters. (I suppose it's possible they were planning to get rid of doubling, which was costing them money, and implement another game with a house edge in its place.

Actually, doubling would still be in place. However, upon winning a double-up, another bonus game could randomly be triggered - at least this is what they were planning at the time.