Mathematical Proof that English Harbour is cheating

[caruso]

At best you'll get a ruling that it was an unfair game, but even then, do we know that EH state the gamble is supposed to be 50/50? Or have we just assumed that? (no, I really don't know)

Yes, we do know, and no, we didn't just assume that. I posted it above, but I'll do it again for ya'.

The results of this ongoing audit indicate that the random number generator is truly random and accordingly, the game play results conform with accepted statistical norms.

The claim may be bogus, Shack may no longer audit anything, etc ad inf / nauseum. Nonetheless, the claim is thus. English Harbour claims to deal a fair game. English Harbour does not deal a fair game.

[thelawnet]

Soflat is right. You may feel "cheated", it looks *very* suspicious, and it wasn't a random game. But the fact is we can only speculate as to why this came about. Yes it's incredibly fishy, but if you took this to a court of law claiming it was a deliberate "cheat" with the "proof" presented in this thread, you'll be in there for about 10 minutes.

I hardly think so. There is absolute proof (nothing can proven, but this is more certain than any evidence under which people are found guilty of crimes - we are talking about trilllion trillion trillions to one against) that English Harbour did not provide a game where the odds provided to the player did not match the odds implied by the game rules. This would be illegal under Las Vegas law, and a casino operating in such a way would be fined thousands of dollars on the basis of evidence no more solid than this. The only thing which we don't know is how the unfair game came to be in place, but that's not as important as the fact of the cheating.

At best you'll get a ruling that it was an unfair game, but even then, do we know that EH state the gamble is supposed to be 50/50? Or have we just assumed that? (no, I really don't know)

The help says "if the card you have selected is greater value than the dealers then you will DOUBLE your winnings".

According to them, their software "is truly random and accordingly, the game play results conform with accepted statistical norms." So #1, under reputable jurisidictional law it is illegal to provide a game where the odds implied by the game (drawing a higher card) do not match the true odds, and #2, they say that their games are random and that clearly is untrue.

[spearmaster]

So it'd be fair to say this was no "freak run".

I did say "albeit remotely" Nevertheless, it still exists within the realm of possibility... barely...

No, that is not possible.

522 wins out of 1537 trials has according to Excel a zero chance of occurrence. Of course it is non-zero, but the number is so small, Excel cannot calculate it. The best it can do is for 617 wins, when the possibility is

0.0000000000000019

So the fact that the software was cheating is a fact as anything can be.

See above If you replace the words "cheating is a" with "non-random is as close to" I can't argue.

People do sometimes make errors in coding. But that being said, they're going to have to climb a very tall mountain to make that one stick in this case.

[thelawnet]

This is categorically NOT a freak run, remotely possible. If I've read those decimal places correctly

You didn't.

I said

522 wins out of 1537 trials has according to Excel a zero chance of occurrence. Of course it is non-zero, but the number is so small, Excel cannot calculate it. The best it can do is for 617 wins, when the possibility is

In other words, 522 wins is such a tiny number Excel is not remotely capable of calculating it. The number, 1.9*10^-12 is for 617 wins, being the smallest Excel can handle. 522 wins is orders of magnitude less likely than even that tiny number. At a guess you are looking at at least 20 zeros before the number.

I do not have any suitable software installed to calculate this number, but I daresay someone does and can oblige us with the tininess of the number.

[spearmaster]

The results of this ongoing audit indicate that the random number generator is truly random and accordingly, the game play results conform with accepted statistical norms.

Come to think of it, this statement is bogus. Just because a random number generator is truly random does not necessarily mean that game play results conform with accepted statistical norms.

I can still use the number generated by the RNG and apply it to a formula which says "If (RNG>.00000000001) {dealercard="Ace";}"... LOL...

[GrandMaster]

In other words, 522 wins is such a tiny number Excel is not remotely capable of calculating it. The number, 1.9*10^-12 is for 617 wins, being the smallest Excel can handle. 522 wins is orders of magnitude less likely than even that tiny number. At a guess you are looking at at least 20 zeros before the number.

I do not have any suitable software installed to calculate this number, but I daresay someone does and can oblige us with the tininess of the number.

The probability of winning 522 or fewer out of 1537 trials is 4.9*10^-37. It is the same order of magnitude as winning the lottery with a single ticket 5 weeks in a row.

[thelawnet]

The probability of winning 522 or fewer out of 1537 trials is 4.9*10^-37. It is the same order of magnitude as winning the lottery with a single ticket 5 weeks in a row.

For comparison, there are apparently 7.5x10^18 grains of sand on every beach in the world.

So this is as likely as randomly picking a grain of sand from every beach in the world, and getting the same one three times.

In other words, it ain't gonna happen - with 4.6*10^17 seconds since the big bang, the chances are still remote if you tried a billion times per second since the beginning of time.

[caruso]

You didn't.

I said...

My consecutive blackjacks comparison was based on the figure given as the most excel could calculate, though I noted that it couldn't calculate this particularly extreme one.

Anyway, whatever the figure, this is blatantly not a freak occurance. That much is certain at this point.

[Macgyver]

The help says "if the card you have selected is greater value than the dealers then you will DOUBLE your winnings".

This was in response to the "how can we be sure it's 50/50?" question, and I'd like to ask something about it. Bear with me/humor me/laugh at me, but it's something that I've been thinking about.

When you double, you have the dealer's card face up and four cards to pick from, correct? If the card you pick is higher than the dealer's, than you double your winnings (hence the name).

Doesn't that mean, though, that you have a 1 in 4 chance of picking a higher card (which isn't 50/50)? I know that's over-simplifying things a lot, but it's a start to this nagging question. Not to mention that you have to factor in the possibility that if the dealer has a King, you only have a 4 in 51 chance (deck is missing a King and only one of the four Aces can give you a win).

Does that make sense?

[JohnGalt]

Soflat is right. You may feel "cheated", it looks *very* suspicious, and it wasn't a random game. But the fact is we can only speculate as to why this came about. Yes it's incredibly fishy, but if you took this to a court of law claiming it was a deliberate "cheat" with the "proof" presented in this thread, you'll be in there for about 10 minutes. At best you'll get a ruling that it was an unfair game, but even then, do we know that EH state the gamble is supposed to be 50/50? Or have we just assumed that? (no, I really don't know)

Issues like this require a balanced view, not an emotional one, if they are to be taken anywhere. That's no sleight on you TLN, as I think this is one of the most interesting threads in a long time, and for me personally, I won't be playing here.

In America (and the UK) there is a doctrine known as res ipsa loquitor, or, "the thing speaks for itself". In fact it is statistical in nature. Given that the software is cheating, what is the probability that it is cheating because the casino deliberately manipulated it? If that is > 50%, you win on summary judgment.

I'm not sure it matter if they did it deliberately, since they re responsible for what their software does. But I'd say this evidence may well be enough to win a court case, on summary judgment. So you might be there for 10 minutes, at which point you will already have won!

It would depend to some extent on the nature of the programming... how likely is such an error to occur by accident? My guess is not likely at all. And in a court case you could have a few expert coders testify to that and that would be it, you'd win. No direct evidence of the casino's actions is needed.

But in this "court" of public opinion, I suppose we could use one or two knowledgeable people speaking to this likelihood before we can out and out brand them deliberate cheaters.