Mathematical Proof that English Harbour is cheating

[caruso]

I also recall the Playcheck-style logs from when I last played OddsOn.

All players who participated should now email in requests for copies of their play logs. For the most part the play was very brief, so these should be easily provided.

 

[Linus]

I believe this fatal damage to reputation will prove much more expensive to them in the long term than the money they made from their unfair doubling game (although of course nobody knows how much that is - it could be substatnial).

There's no reason to think doubling was the only fixed game at English Harbour.

How long would it take to show Royal flushes weren't being dealt fairly?

I'm guessing you'd go broke long before you could determine it.

 

[wizard]

I'm on the case

Yesterday, May 2, I was made aware of the issue possible irregularities in the doubling feature of Odds On software. Let me assure you that the Odds On management and myself are taking this very seriously. As soon as possible I plan to conduct an analysis of all double or nothing bets made since January 1. If necessary, Odds On will hire a third party to conduct the same study. Until we have had a chance to review the log files we can neither confirm nor deny the accusations.

I will say now that indeed I have auditing most of the Odds On casinos, including the English Harbour. Also, I agree with the original post that the probability of 84 or fewer wins in 235 resolved bets is 1 in 135929.

 

[Flavio4321]

just a question, is the user "wizard", michael shackelford aka "the wizard of odds"?

 

[spearmaster]

My apologies, I guess the Wiz is still on the job At least now we know that a proper review will take place by one or more parties.

I would suggest that until this has taken place, members of the forum kindly refrain from any accusations or insinuations as to what happened.

Flavio - yes, that's the Wizard of Odds.

 

[gfkostas]

My apologies, I guess the Wiz is still on the job At least now we know that a proper review will take place by one or more parties.

I would suggest that until this has taken place, members of the forum kindly refrain from any accusations or insinuations as to what happened.

Though it might be already late for an investigation if they have 'fixed' their game back to normal should that be possible.

 

[JohnGalt]

Though it might be already late for an investigation if they have 'fixed' their game back to normal should that be possible.

He's retroactively checking all the doubling since January 1. He should be able to tell if they've altered the logs by comparing the results on this site against the results they've recorded, if that becomes necessary (He might have to collect some usernames to do this but hopefully it won't come to that).

 

[thelawnet]

Yesterday, May 2, I was made aware of the issue possible irregularities in the doubling feature of Odds On software. Let me assure you that the Odds On management and myself are taking this very seriously. As soon as possible I plan to conduct an analysis of all double or nothing bets made since January 1. If necessary, Odds On will hire a third party to conduct the same study. Until we have had a chance to review the log files we can neither confirm nor deny the accusations.

I guess in the event that the software was at one time fair, you will be able to identify when it stopped being fair by analysing data from each day (and depending on the amount of data that you actually have), as although it is clear that the software was behaving unfairly for a period on and before April 30 2006, it was not necessarily unfair on say January 1: otherwise it might have been noticed then. I think it reasonable to assume that this has been going on for a few weeks but not necessarily months.

I'm also interested that the Hot Pepper and Fire and Ice casinos, outside of the EH group of casinos, althought not as thoroughly tested as EH, did not appear to have aberrant (and indeed abhorrent) doubling results.

I am somewhat surprised at English Harbour's early statement about the results being consistent with a fair game as it is quite clear that the game was not fair (not withstanding that results indicate that they clearly changed the game early on 1 May, although their carefully worded statement did not explicitly denied that they had not) , and that statement is now looking premature with analysis to be conducted here.

 

[thelawnet]

Lawnet, you should contact the Wizard of Odds with your results. If he agrees, this problem will get more attention.

Cheating by Odds On does not surprise me.

And yes, if the double down feature is not 50/50, the game is cheating.

I did email him on 29/04

 

[caruso]

It's extremely important that the figures are not just grouped together from some distant date. I doubt the gaff was in place on January 1st since, as already noted, it'd have been picked up on most likely. Assuming it was in place no earlier than (say) early April, those earlier fair results will dilute the the effect of the cheating results, with the extent dependent on how recently the gaff was put in place.

In order for this not to happen, the results need to be looked at on a daily basis at least.

Even then, there is still risk of diluting the effect, since we don't know how the gaff was set up - overall, individual accounts or whatever.

The Casino Bar experiment was conducted over less hands, if I recall correctly, and the probability was MUCH, much less - I think somewhere in the trillions. As such, I'm curious to know exactly why these results, far worse in terms of probability, are not immediate evidence of rigging, where the Casino Bar experiment results were. The only difference is that the Casino Bar test was set up in the correct manner, whereas here we're selectively looking backwards at data which fits the contention. Other than that, the results here are far, far worse than anything Casino Bar ever did.

 

[caruso]

The only difference is that the Casino Bar test was set up in the correct manner, whereas here we're selectively looking backwards at data which fits the contention.

My mistake. The initial test was specifically set up to test the hypothesis, and as such followed exactly the correct protocol for statistical analysis - and the exact same protocol as the Casino Bar experiment.

 

[Aindreas_Daoc]

I tried looking just now, but couldn't find a history log. Live chat said that there is no play-check type feature, but that you can request play history directly from them.

Edit: I just requested the logs via e-mail. However, it is unclear whether they even record play money games. The support person terminated chat before I had a chance to ask.

Update: EH apparently does not log play money hands. I recieved my game logs from the time I tested the VP doubling, but none of the play money games were included.

 

[Simmo!]

Update: EH apparently does not log play money hands. I recieved my game logs from the time I tested the VP doubling, but none of the play money games were included.

Thats pretty standard across the industry as far as I am aware but anyway AD.

 

[thelawnet]

I was suspicious of the fairness of the doubling on English Harbour's video poker. So I deposited and carefully recorded data for over an hour, doubling after every win, except I collected any large wins, and recording for each double the result, win, lose or push.

I played until I lost all my money (playing 4 line, single coin, recording wins, losses and ties on the doubling game (I did not record the video poker itself - I was only concerned with the double).

According to my data, it is 99.999% certain that English Harbour is not offering a fair doubling game in their Tens or Better video poker game.

Although nothing is completely certain, 99.999% would be good enough to convict a man and sentence him to death, so I think it is good enough for any reasonable person in the world to be satisfied that English Harbour is a cheating casino.

Here are my results:
84 wins
151 losses
19 ties

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.

 

[dirk_dangerous]

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.

Sounds like Odds-On threw the "cheat switch" and then turned it back to fair.

I expect the usual "oh there was a software mistake" excuse.

Frankly, I will always declare every Odds-on casino as cheaters. The evidence is already here IMO.

CM and the rest of you portal owners; you really should drop all Odds-On casinos without any delay.

 

[jeremiahsjohnso]

At least they are handing the logs over. This should end any possibility of a cover up assuming the Wizard gets access to the same files (or at least, the doubles records).

The next question: what does the casino do once the fix is proven (by the Wizard)?

 

[spearmaster]

I repeat again - insinuations and accusations should be avoided for the time being. Although the first time was a request, this time it's a warning - until such time as we hear from the auditing parties, no more speculation of this sort, otherwise you will be speculating on vacation.

The fact that logs are being handed over to players should indicate some sincerity, for crying out loud. What program do you know will damn themselves like that if they were trying to hide the problem?

 

[Linus]

They're our logs of our own play - which have already been recorded here. How does turning them over damn them?

Wouldn't refusing to turn them over be an admission of guilt?

 

[The Dude]

...CM and the rest of you portal owners; you really should drop all Odds-On casinos without any delay.

This was done yesterday. There were two Oddson casinos in Casinomeister's accreditted section (Fire and Ice, Hotpepper). They'll be placed back there if or when I feel comfortable with the results and resolution from this thread.

 

[dovmin]

I want to share my thoery of what happened with the English harbour group let me know what you all think.


Theory number one, I got response from my mathematician friend and he had to say and he is right. The chance of that kind of result happening is possible but the chance that this kind of result is happening exactly when this guy tested them is zero.

They absolutely cheated, you might say it was not deliberately but I doubt it.

Theory number 2: I see in their terms and condition that many games are excluded for bonuses.

I bet that many bonus whoring players came there and use the doubling figure to maximize their winnings building their balance up and only then play the games which carry a high house edge.

The casino understood they need to fix their doubling figure to eliminate that and that what he did.

I played there and I can say that according to my feeling only no calculations two games are rigged there.

One, their Roulette and second the doubling at any Videopoker.

I remember casino bar response to the Wizard of odds issue, I remember they mentioned the fact there is a lot of bonus abuse in the industry and this is very hard to manage a casino anyway. Casinobar implied that of the line that they cheated to overcome the bonus abusers and I believe this is the reason a casino want to cheat.

They don't need to cheat in order to win without bonuses.

They need the cheating figure to cope with the chargebacker and Bonus abuser.

 

[spearmaster]

They're our logs of our own play - which have already been recorded here. How does turning them over damn them?

Wouldn't refusing to turn them over be an admission of guilt?

Not turning them over would be very dubious at best.

Turning them over would "damn them" if they had intended to cheat players. What I'm trying to point out is that perhaps they are trying to be upfront and provide you with what you should be entitled to in the first place.

So if they give you the logs - knowing that the data in there may indeed expose an issue - then this is tantamount to acknowledging that there may be problems and that they will work with you to address them.

 

[spearmaster]

Dovmin - again, while I don't disagree with you - I have already asked twice that no more insinuations or accusations be thrown for the time being.

This is a final warning to everyone. Talk all you like but do NOT make any more accusations or insinuations because there are a number of parties already working on the issue. "Non-random" is okay. "Cheating" or "deliberate" is not.

 

[thelawnet]

I want to share my thoery of what happened with the English harbour group let me know what you all think.


Theory number one, I got response from my mathematician friend and he had to say and he is right. The chance of that kind of result happening is possible but the chance that this kind of result is happening exactly when this guy tested them is zero.

They absolutely cheated, you might say it was not deliberately but I doubt it.

Theory number 2: I see in their terms and condition that many games are excluded for bonuses.

I bet that many bonus whoring players came there and use the doubling figure to maximize their winnings building their balance up and only then play the games which carry a high house edge.

The casino understood they need to fix their doubling figure to eliminate that and that what he did.

I played there and I can say that according to my feeling only no calculations two games are rigged there.

One, their Roulette and second the doubling at any Videopoker.

I remember casino bar response to the Wizard of odds issue, I remember they mentioned the fact there is a lot of bonus abuse in the industry and this is very hard to manage a casino anyway. Casinobar implied that of the line that they cheated to overcome the bonus abusers and I believe this is the reason a casino want to cheat.

They don't need to cheat in order to win without bonuses.

They need the cheating figure to cope with the chargebacker and Bonus abuser.

I am not sure about the roulette. You present nothing to back this up, not even anecdotal 'evidence'.

However it does make some sense if you perceive that many players are making money from your bonus to attacke the game those players (whose business you do not want) will play. In that sense if I were operating a casino and wanted to attack the nonprofitable bonus players (given that other players are profitable because all casino games will offer a casino edge), I would fix the odds the game perceived most attractive by such players. As the expected loss on a video poker double is zero, this game is likely to be preferred by bonus-oriented players.

It is difficult to see how this has happened. It is clear that the game was not behaving fairly, and that the chance the game was fair is basically zero. The fact is that it appears the game suddenly now is playing fair where it wasn't before while the casino have said that the results were fair, even though that is statistically essentially impossible. I think there is only one conclusion that can be drawn from this evident change and early denial; the problem is for EH with logs available and being analysed, I believe that it will be extremely difficult for them to reconcile evidence showing bad odds changing to fair odds with the initial statement that effectively denied that they had changed anything

 

[MarcyW]

I want to share my thoery of what happened with the English harbour group let me know what you all think.


Theory number one, I got response from my mathematician friend and he had to say and he is right. The chance of that kind of result happening is possible but the chance that this kind of result is happening exactly when this guy tested them is zero.

They absolutely cheated, you might say it was not deliberately but I doubt it.

Theory number 2: I see in their terms and condition that many games are excluded for bonuses.

I bet that many bonus whoring players came there and use the doubling figure to maximize their winnings building their balance up and only then play the games which carry a high house edge.

The casino understood they need to fix their doubling figure to eliminate that and that what he did.

I played there and I can say that according to my feeling only no calculations two games are rigged there.

One, their Roulette and second the doubling at any Videopoker.

I remember casino bar response to the Wizard of odds issue, I remember they mentioned the fact there is a lot of bonus abuse in the industry and this is very hard to manage a casino anyway. Casinobar implied that of the line that they cheated to overcome the bonus abusers and I believe this is the reason a casino want to cheat.

They don't need to cheat in order to win without bonuses.

They need the cheating figure to cope with the chargebacker and Bonus abuser.

Unreal. A casino gives a bonus then needs to cheat so players can't win? Wow, I spit out my coffee on the keyboard I laughed so hard.

 

[The Dude]

Unreal. A casino gives a bonus then needs to cheat so players can't win? Wow, I spit out my coffee on the keyboard I laughed so hard.

Spearmaster gave a warning against making any more comments such as this. Account suspended one week.

 

[soflat]

I'm confused as to what role OddsOn plays. Wizard said that OddsOn turned over the logs. Does Oddson own and operate the casino, or are they a software company that licenses the product to third parties? Do the games run from OddOn's servers?

 

[spearmaster]

OddsOn is the software provider. In theory, the operator should own the servers, but in many cases the software provider may provide all the related services as well.

 

[Webzcas]

In theory, the operator should own the servers, but in many cases the software provider may provide all the related services as well

One good example is the Prima Poker Network. Where all prima operators utilise the servers in Canada.

 

[wizard]

To the Internet Gambling Community, from Michael Shackleford (the Wizard of Odds)

As a professional mathematician, one of the services I provide is evaluation of the logs of online casinos to ensure fair play, though I stopped taking new clients for this service a few years ago. One of my existing clients is Odds On, which provides the software for English Harbour casino among others. On April 29, 2006, a player posted at CasinoMeister.com that he lost more frequently than expected on the double-up feature when playing video poker at English Harbour. This was brought to my attention on May 2 by both Bryan Bailey of CasinoMeister and English Harbor Management. As soon as I found out about this I immediately requested detailed logs from English Harbor in order to analyze them. In preparing the log files for me, Odds On themselves discovered anomalies in the double-up records. When I received the logs and analyzed them I was able to confirm that the double-up on video poker did indeed pay out less than it should have between April 13 and May 2. According to English Harbor problem was caused by a buggy software update on April 13, which was automatically corrected by a subsequent update on May 2.

Readers might understandably wonder how English Harbor can claim that I review their logs for fairness if a problem like this can arise without my knowing about it. The answer is that I do the analysis monthly, and though players noticed the problem in April, I would not normally see the April data until mid May. This underscores the important point that my auditing service can not guarantee that players will get a mathematically fair game at every given instant. I can only attest to the fairness and return percentages in the past.

Although both English Harbor and myself believe that April 13 to May 2 was the only period affected we will be scrutinizing earlier logs to make sure, and will be monitoring future play closely.

The following table shows the results of all double or nothing bets in real play mode from April 1 to May 3.


Apr 01: wins=1225, losses=1245, ties=177, win ratio=49.6%
Apr 02: wins=763, losses=794, ties=89, win ratio=49%
Apr 03: wins=588, losses=580, ties=74, win ratio=50.34%
Apr 04: wins=1293, losses=1232, ties=156, win ratio=51.21%
Apr 05: wins=951, losses=918, ties=108, win ratio=50.88%
Apr 06: wins=1015, losses=996, ties=119, win ratio=50.47%
Apr 07: wins=950, losses=989, ties=127, win ratio=48.99%
Apr 08: wins=759, losses=735, ties=90, win ratio=50.8%
Apr 09: wins=818, losses=862, ties=111, win ratio=48.69%
Apr 10: wins=1203, losses=1168, ties=152, win ratio=50.74%
Apr 11: wins=529, losses=524, ties=64, win ratio=50.24%
Apr 12: wins=1199, losses=1195, ties=121, win ratio=50.08%
Apr 13: wins=194, losses=266, ties=27, win ratio=42.17%
Apr 14: wins=212, losses=424, ties=43, win ratio=33.33%
Apr 15: wins=284, losses=607, ties=73, win ratio=31.87%
Apr 16: wins=314, losses=602, ties=98, win ratio=34.28%
Apr 17: wins=139, losses=304, ties=45, win ratio=31.38%
Apr 18: wins=143, losses=317, ties=29, win ratio=31.09%
Apr 19: wins=141, losses=307, ties=42, win ratio=31.47%
Apr 20: wins=89, losses=169, ties=28, win ratio=34.5%
Apr 21: wins=134, losses=295, ties=42, win ratio=31.24%
Apr 22: wins=72, losses=128, ties=23, win ratio=36%
Apr 23: wins=52, losses=112, ties=21, win ratio=31.71%
Apr 24: wins=138, losses=322, ties=41, win ratio=30%
Apr 25: wins=66, losses=121, ties=16, win ratio=35.29%
Apr 26: wins=75, losses=126, ties=18, win ratio=37.31%
Apr 27: wins=176, losses=399, ties=47, win ratio=30.61%
Apr 28: wins=93, losses=173, ties=24, win ratio=34.96%
Apr 29: wins=40, losses=81, ties=15, win ratio=33.06%
Apr 30: wins=96, losses=169, ties=25, win ratio=36.23%
May 01: wins=73, losses=86, ties=9, win ratio=45.91%
May 02: wins=95, losses=125, ties=17, win ratio=43.18%
May 03: wins=142, losses=158, ties=20, win ratio=47.33%

Win Ratio = ratio of wins to sum of wins and losses.

English Harbor management tells me that they plan to reimburse each and every player who made a double or nothing bet during the period affected with his/her net loss on those bets over affected period. This offer will apply to all casinos using Odds On software. In addition, as a show of their regret for this mistake, they will add 20% to each player's reimbursement.

Michael Shackleford, A.S.A.

WizardOfOdds.com

 

[The Dude]

English Harbour Statement

This just in from English Harbour:

First and foremost, the English Harbour Group would like to formally apologize for the error in the software release that caused this situation. We hold in high regard all our players and respect for the Internet Gaming industry. We feel it is important for players to understand what happened, what went on during the investigation, how it was resolved and our compensation to affected players.

The issue was caused by a bug in the software staging release on the later part of the day April 13 and went unnoticed. On May 2 we had a maintenance release that during the course of our investigation seemed to have corrected the problem. When we were alerted of the claim, we ran through several trials and also simulations. We checked randomness of the distribution of cards in the doubling game over different time periods and nothing seemed to be out of the ordinary. We then did an interim posting on Casinomeister. We continued the investigation and sought the audit services of Michael Shackleford. Michael asked for the log files and we fully complied to his request. As part of the delivery of the log files to Michael, we noticed an issue that seemed to have spanned the later two weeks of April. We mentioned this to Michael as we continued to work with him. Through Michael's investigation, he concurred with our observations and was able help quantify the variance in the odds.

Michael continues to be a key part of our governance and fair game play strategy.

We discussed player compensation with Michael and we will be refunding affected players on the net loss on the doubling game plus an additional 20% of their net loss. Players will be receiving an email as well as a message within the casino itself shortly.

The English Harbour Group would like to thank the Casinomeister forum members for bringing this issue to our attention and their patience while we dealt with this issue.

The English Harbour Group Management Team

 

[AceMan76]

We take your comments seriously and as such have spent time reviewing our Video Poker game play and also await some feedback on an independent review from a 3rd party.

We have concluded our review of the game play and randomness for all Video Poker games.

We have found, taking several sample sets over different and varying lengths of time, that they yield in our opinion, a non biased distribution of the cards.

Should small sample sets be used to measure this metric, then results will vary as seems to be the case tracing through this thread.

LMAO on that one.

We trust, that we have responded adequately and any doubt in peoples' mind are put to rest.

That one is a laugher too.

Perhaps English Harbour should reconsider the language they use in their initial responses . . . Maybe "we don't claim sh** until the independent 3rd party review is done".

 

[kengam]

why was there significantly less play on the double up during the period in question versus the days before April 13th.

There are some serious questions that need to be answered mainly what exactly was the error causing the win % to be out of whack?

 

[kengam]

So if this can happen to the double up part of video poker is it safe to assume it can happen at any given time with any other odds on game?

 

[spearmaster]

Hey kengam... welcome back!

I think the possible answers will be obvious -

1. Any player who suddenly feels he is having bad luck at doubling will stop soon enough - especially if the results aren't what he/she is used to seeing from the past.

2. Obviously the issue being raised in the forums also contributed to a slowdown in real-mode doubling-down.

I too would like to see exactly what error was discovered... but if it's what I think it is, it is an obvious boo-boo. But better I not speculate - let's see if they're willing to tell us.

 

[Zoozie]

1.
I am pleased that this obvious 'bug/malfuntion' was admitted by the casino/software provider. And also their willingness to provide logs, this was the best they could do.

2.
I am also pleased with the solution which is refund of money due to 'malfunction' + small compensation bonus. Again this was the best and only solution.

3.
The software provider has taken reponsibility for this 'malfunction'. So the EH group are innocents in this gigantic misfortune.


4.
I am NOT pleased with the missing explanation of how it could happen. I will get into this part.

4a.
WoO does montly reviews so this does explain why he has not noticed anything unusual.
However is this the first time this happens? If it has happened before I have never heard of any compensation for the players or any news flash about it. What is standard procedure for this?

4b.
I need a fully (maybe deeply technical) explation for this bug. I am a software developer (mission critical software) and we have a 'unit-test' which is common practice even in non-critical software development. This 'unit-test' test the software and an error of this magnitude could never exists in new releases.

4c.
I do not believe such a software error could happen in a doubling game with cards by mistake. This would mean the cards are not dealth random, which means the only explanation is in the RNG(but then ALL GAMES was 'rigged'). To introduce such a software bug you would really have to change SEVERAL LINES of code which could not happen by mistake. Also the new code would look fundamentally different from the old 'symmetric' code.

4d.
This 'malfunction' was present in the software for weeks. It is 'interesting' that it
disappears right after the postings on this forum. Was the new release/patch of the
software only to fix this 'malfunction'. If that is the case the EH group could have
informed the player community about this allready when they contacted CM. As I see
it the only other explanation was that it was fixed by 'mistake' again, just as it had
occoured in the first place. This is not likely.

Maybe... there is a simpler explanation, but then the players deserve to hear it.



Zoozie

 

[boogaooga]

Hey kengam... welcome back!

I think the possible answers will be obvious -

1. Any player who suddenly feels he is having bad luck at doubling will stop soon enough - especially if the results aren't what he/she is used to seeing from the past.

2. Obviously the issue being raised in the forums also contributed to a slowdown in real-mode doubling-down.

I was prepared to make a post about how the data looked suspicious to me, too, but then a thought occurred to me. How many times does a player take a small win and try to double it repeatedly into a bigger win? Often, I imagine. But it's pretty hard to get a string of doubles going when you're losing close to 70% of the time, as was occurring for the last half of April. I suspect that accounts for most of the big slowdown in doubling that occurred once the, ahem, 'bug' manifested itself.

 

[caruso]

"Bug", I love that one bad.

Was the Casino Bar seconds dealing also due to a "bug"?

Obviously, precise details, which will be acceptable to those players here with programming skills, are needed before any of this is any further on.

 

[kengam]

Apr 01: wins=1225, losses=1245, ties=177, win ratio=49.6%
Apr 02: wins=763, losses=794, ties=89, win ratio=49%
Apr 03: wins=588, losses=580, ties=74, win ratio=50.34%
Apr 04: wins=1293, losses=1232, ties=156, win ratio=51.21%
Apr 05: wins=951, losses=918, ties=108, win ratio=50.88%
Apr 06: wins=1015, losses=996, ties=119, win ratio=50.47%
Apr 07: wins=950, losses=989, ties=127, win ratio=48.99%
Apr 08: wins=759, losses=735, ties=90, win ratio=50.8%
Apr 09: wins=818, losses=862, ties=111, win ratio=48.69%
Apr 10: wins=1203, losses=1168, ties=152, win ratio=50.74%
Apr 11: wins=529, losses=524, ties=64, win ratio=50.24%
Apr 12: wins=1199, losses=1195, ties=121, win ratio=50.08%
Apr 13: wins=194, losses=266, ties=27, win ratio=42.17%
Apr 14: wins=212, losses=424, ties=43, win ratio=33.33%
Apr 15: wins=284, losses=607, ties=73, win ratio=31.87%
Apr 16: wins=314, losses=602, ties=98, win ratio=34.28%
Apr 17: wins=139, losses=304, ties=45, win ratio=31.38%
Apr 18: wins=143, losses=317, ties=29, win ratio=31.09%
Apr 19: wins=141, losses=307, ties=42, win ratio=31.47%
Apr 20: wins=89, losses=169, ties=28, win ratio=34.5%
Apr 21: wins=134, losses=295, ties=42, win ratio=31.24%
Apr 22: wins=72, losses=128, ties=23, win ratio=36%
Apr 23: wins=52, losses=112, ties=21, win ratio=31.71%
Apr 24: wins=138, losses=322, ties=41, win ratio=30%
Apr 25: wins=66, losses=121, ties=16, win ratio=35.29%
Apr 26: wins=75, losses=126, ties=18, win ratio=37.31%
Apr 27: wins=176, losses=399, ties=47, win ratio=30.61%
Apr 28: wins=93, losses=173, ties=24, win ratio=34.96%
Apr 29: wins=40, losses=81, ties=15, win ratio=33.06%
Apr 30: wins=96, losses=169, ties=25, win ratio=36.23%
May 01: wins=73, losses=86, ties=9, win ratio=45.91%
May 02: wins=95, losses=125, ties=17, win ratio=43.18%
May 03: wins=142, losses=158, ties=20, win ratio=47.33%

Win Ratio = ratio of wins to sum of wins and losses.


Does this list include play for fun mode? The only positive thing we have learned from this is that the fun mode works off the same program as in real mode. I know that is not the case for some other software types. This group better be prepared to open up about what they know because this issue is not likely to go away peacefully.

 

[thelawnet]

As a professional mathematician, one of the services I provide is evaluation of the logs of online casinos to ensure fair play, though I stopped taking new clients for this service a few years ago. One of my existing clients is Odds On, which provides the software for English Harbour casino among others. On April 29, 2006, a player posted at CasinoMeister.com that he lost more frequently than expected on the double-up feature when playing video poker at English Harbour. This was brought to my attention on May 2 by both Bryan Bailey of CasinoMeister and English Harbor Management. As soon as I found out about this I immediately requested detailed logs from English Harbor in order to analyze them. In preparing the log files for me, Odds On themselves discovered anomalies in the double-up records. When I received the logs and analyzed them I was able to confirm that the double-up on video poker did indeed pay out less than it should have between April 13 and May 2. According to English Harbor problem was caused by a buggy software update on April 13, which was automatically corrected by a subsequent update on May 2.

Readers might understandably wonder how English Harbor can claim that I review their logs for fairness if a problem like this can arise without my knowing about it. The answer is that I do the analysis monthly, and though players noticed the problem in April, I would not normally see the April data until mid May. This underscores the important point that my auditing service can not guarantee that players will get a mathematically fair game at every given instant. I can only attest to the fairness and return percentages in the past.

Although both English Harbor and myself believe that April 13 to May 2 was the only period affected we will be scrutinizing earlier logs to make sure, and will be monitoring future play closely.

The following table shows the results of all double or nothing bets in real play mode from April 1 to May 3.


Apr 01: wins=1225, losses=1245, ties=177, win ratio=49.6%
Apr 02: wins=763, losses=794, ties=89, win ratio=49%
Apr 03: wins=588, losses=580, ties=74, win ratio=50.34%
Apr 04: wins=1293, losses=1232, ties=156, win ratio=51.21%
Apr 05: wins=951, losses=918, ties=108, win ratio=50.88%
Apr 06: wins=1015, losses=996, ties=119, win ratio=50.47%
Apr 07: wins=950, losses=989, ties=127, win ratio=48.99%
Apr 08: wins=759, losses=735, ties=90, win ratio=50.8%
Apr 09: wins=818, losses=862, ties=111, win ratio=48.69%
Apr 10: wins=1203, losses=1168, ties=152, win ratio=50.74%
Apr 11: wins=529, losses=524, ties=64, win ratio=50.24%
Apr 12: wins=1199, losses=1195, ties=121, win ratio=50.08%
Apr 13: wins=194, losses=266, ties=27, win ratio=42.17%
Apr 14: wins=212, losses=424, ties=43, win ratio=33.33%
Apr 15: wins=284, losses=607, ties=73, win ratio=31.87%
Apr 16: wins=314, losses=602, ties=98, win ratio=34.28%
Apr 17: wins=139, losses=304, ties=45, win ratio=31.38%
Apr 18: wins=143, losses=317, ties=29, win ratio=31.09%
Apr 19: wins=141, losses=307, ties=42, win ratio=31.47%
Apr 20: wins=89, losses=169, ties=28, win ratio=34.5%
Apr 21: wins=134, losses=295, ties=42, win ratio=31.24%
Apr 22: wins=72, losses=128, ties=23, win ratio=36%
Apr 23: wins=52, losses=112, ties=21, win ratio=31.71%
Apr 24: wins=138, losses=322, ties=41, win ratio=30%
Apr 25: wins=66, losses=121, ties=16, win ratio=35.29%
Apr 26: wins=75, losses=126, ties=18, win ratio=37.31%
Apr 27: wins=176, losses=399, ties=47, win ratio=30.61%
Apr 28: wins=93, losses=173, ties=24, win ratio=34.96%
Apr 29: wins=40, losses=81, ties=15, win ratio=33.06%
Apr 30: wins=96, losses=169, ties=25, win ratio=36.23%
May 01: wins=73, losses=86, ties=9, win ratio=45.91%
May 02: wins=95, losses=125, ties=17, win ratio=43.18%
May 03: wins=142, losses=158, ties=20, win ratio=47.33%

Win Ratio = ratio of wins to sum of wins and losses.

English Harbor management tells me that they plan to reimburse each and every player who made a double or nothing bet during the period affected with his/her net loss on those bets over affected period. This offer will apply to all casinos using Odds On software. In addition, as a show of their regret for this mistake, they will add 20% to each player's reimbursement.

Michael Shackleford, A.S.A.

WizardOfOdds.com

The explanation from EH is clearly unsound. They claim that there was a bug in the software that was fixed by an automatic update, and previously denied that there was anything wrong with their software.

There is no such thing as an 'automatic' update. Computer code does not write itself, and cannot fix itself either. The code can only have been changed knowingly, and the fact that it was changed almost immediately after te problem was posted here is evidence that they were aware that the code was not fair; despite this they released a statement claiming that the game was fair, even after they had just changed the code that was rigged against the playe .

Although not ideal, it is not the end of the world for a casino to have a software 'problem' that they fix and compensate the player, but having denied that there was a probem having stealth fixed it just before claiming that with a sufficiently large sample the game would be fair, before it became clear that this explanation would not wash, is very difficult for them. Why change your software from biased to fair, and then deny that anything was wrong, pretending that nothing has been changed?

They used the honest mistake defence, which was their best option, but only after dishonestly first changing their software and then telling the world that it was fair, a claim they knew was false because they had just changed it. This fatally undermines their latest statement.

And let's not forget the question of why the game came to be rigged. They claim this was an accident. Given the circumstances it seems a duty that they provide a fuller explanation as to how this 'accidentally' came to pass. WHat was the code check in that led to it supposed to have done? They must give a more plausible explanation, although following the initial denial and sly change, things don't look too good for them.

Regarding the detail of the analysis, isnthis just for the English Harbour casino, or for all casinos? Does it include Hot Pepper and Fire and Ice which appeared tobe fair? How was it that these casinos played fair but EH did not?

 

[nektar4d]

Forgive me to repeat it but I can't resist:

What an amazing thread!
:notworthy

 

[Simmo!]

Does it include Hot Pepper and Fire and Ice which appeared tobe fair? How was it that these casinos played fair but EH did not?

It would be interesting to see the Hot Pepper logs for the same period, but if it was a software error then they will reflect the same results.

Thought the Wizard tackled that one very quickly. Didn't expect to see results this quick

 

[AceMan76]

According to English Harbor problem was caused by a buggy software update on April 13, which was automatically corrected by a subsequent update on May 2.

I am a complete idiot when it comes to software and writing code, but why would you need to update the code that runs the RNG? I guess my question is, if the problem was caused by a buggy update, why the heck was the RNG code being updated in the first place?

 

[nafanny29]

The explanation from EH is clearly unsound. They claim that there was a bug in the software that was fixed by an automatic update, and previously denied that there was anything wrong with their software.

There is no such thing as an 'automatic' update. Computer code does not write itself, and cannot fix itself either. The code can only have been changed knowingly, and the fact that it was changed almost immediately after te problem was posted here is evidence that they were aware that the code was not fair; despite this they released a statement claiming that the game was fair, even after they had just changed the code that was rigged against the playe .

Although not ideal, it is not the end of the world for a casino to have a software 'problem' that they fix and compensate the player, but having denied that there was a probem having stealth fixed it just before claiming that with a sufficiently large sample the game would be fair, before it became clear that this explanation would not wash, is very difficult for them. Why change your software from biased to fair, and then deny that anything was wrong, pretending that nothing has been changed?

They used the honest mistake defence, which was their best option, but only after dishonestly first changing their software and then telling the world that it was fair, a claim they knew was false because they had just changed it. This fatally undermines their latest statement.

And let's not forget the question of why the game came to be rigged. They claim this was an accident. Given the circumstances it seems a duty that they provide a fuller explanation as to how this 'accidentally' came to pass. WHat was the code check in that led to it supposed to have done? They must give a more plausible explanation, although following the initial denial and sly change, things don't look too good for them.

Regarding the detail of the analysis, isnthis just for the English Harbour casino, or for all casinos? Does it include Hot Pepper and Fire and Ice which appeared tobe fair? How was it that these casinos played fair but EH did not?

I also think that there is MUCH more to this than we yet know.

One thing about this is that the double up is pretty straighfowward isnt it? 4 cards are dealt face down and you pick one and the program examines it, turns it over and you either win or lose. To misprogram that by accident I would think is impossable. You would have thought if it was a genuine error the program would either be letting you win every double up or none.

And another thing, was the software (when it was faulty ) picking lower cards from the available player choices to lose to the dealers card, or was it selecting a higher dealers card than average. I forgot to look at that when I ran my 200 doubles.

Anyway im not convinced it was an error at the moment, I cant see how it could have possably been given that the cards should be all random. I smell a rat.

 

[nektar4d]

A poll please

Regarding the detail of the analysis, isnthis just for the English Harbour casino, or for all casinos? Does it include Hot Pepper and Fire and Ice which appeared tobe fair? How was it that these casinos played fair but EH did not?

OddsOn software is great.
As a software, it may have some "bugs" from time to time but at least gives "superior control" !


Who will start a poll pls with what readers and players believe about the situation?

 

[soflat]

April 13th, buggy software update at EH

April 28th, player thelawnet reports suspicious results from video poker at EH.

May 2nd Wizard contacted by CM and EH. Automatic software update at EH (update unrelated to VP doubling?).

May 3rd, EH tested the game and reported to CM that they found game fair.

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

In OddOn's and EH's favor, they knew Wizard was going to audit their logs monthly, so why would they cheat?

 

[Linus]

It's unfortunate (for Odds On) that the bug affected doubling so dramatically.

If the malfunction had stopped royal flushes from being dealt, I doubt lawnet would ever have discovered it.

If it had busted double-downs a little more than normal, they probably would have been fine, as well.

Hopefully Odds On won't have more "bugs" in the future.

 

[spearmaster]

thelawnet, by "automatic" I think they meant "previously scheduled". And btw, they said "automatically corrected by an update", not "automatically updated".

kengam - I don't believe their software will record free-play results, so the results above should only be from real money play.

Everyone get RNG out of your head. RNGs do not get updated, nor can they be programmed to give out bad results. Algorithms which interpret the random number provided by the RNG can be misprogrammed, either unintentionally or deliberately.

It is clear that an algorithm was updated in both situations.

I have also identified one possible scenario where a bug could accidentally have entered the system, and I am awaiting the results of a further analysis, or better yet, an explanation from the software provider, preferably here in the forums.

However, this is the only possible scenario where I personally would believe that an unintentional error occurred. Because of this, I have not identified this scenario to the provider nor will I identify it here for the time being. Once I hear an explanation from the provider, I will say more about what I think.

 

[tim5ny]

OddsOn software is great.
As a software, it may have some "bugs" from time to time but at least gives "superior control" !

Superior control?

 

[AceMan76]

Everyone get RNG out of your head. RNGs do not get updated, nor can they be programmed to give out bad results. Algorithms which interpret the random number provided by the RNG can be misprogrammed, either unintentionally or deliberately.

Thank you for clearing this up for me. My next question would then be, why would the algorithm need to be updated for an existing game that is presumably running fine?

 

[spearmaster]

Thank you for clearing this up for me. My next question would then be, why would the algorithm need to be updated for an existing game that is presumably running fine?

Normally, it wouldn't get updated.

I will answer in more detail later but I don't want to compromise anything just yet, pending an answer from the developer.