Online Casino Question

TrendYourFriend · #1 05-13-2006, 04:18 PM

Since I have been logging my blackjack and pontoon sessions, I have wagered $8936 with a total number of trials at 3837 (hands dealt) - $2 per unit flat betting.

I am now down 244 units during this sample. The house edge in Pontoon is 0.17% and Blackjack is 0.66%. Pontoon comprises 60% of this sample size. I play PERFECT basic strategy in each game to ensure that the house edge is a reliable indicator for what I can expect to lose over the long run.

I believe the probability of doing this poorly or worse over the size of this sample is 0.37%. Now I know that some unlucky bastard (me) has to be a data point on the far end of the distribution chart, but I find this to be almost inconceivable given that this is my first shot at online casino gambling.

The casinos I have played at in this sample are very reputable. I believe the software is very reputable (Cryptologic) as well.

So my question is whether or not my sample size is large enough to make any conclusions about the fairness of the game? If not, how large should the sample be?

Phil153 · #2 05-13-2006, 04:44 PM

I can't help you out directly, but you may be interested to know this wouldn't be the first time an online casino dealt an unfair Blackjack game:

http://wizardofodds.com/casinos/casinobar.html

Keepitsimple · #3 05-14-2006, 11:34 AM

[ QUOTE ]
I can't help you out directly, but you may be interested to know this wouldn't be the first time an online casino dealt an unfair Blackjack game:

http://wizardofodds.com/casinos/casinobar.html

[/ QUOTE ]
That is scary.

AaronBrown · #4 05-14-2006, 05:55 PM

It's good there are people like Shakleford with the skill and courage to compute and publish things like this. I have a few thoughts:

(1) Why did the casino go to an expert in game theory for its defense? This is a question for a statistician. I think they were misled by the word "game."

(2) Why would a casino cheat like this? They get your money eventually, whether the edge is big or small. If they make it too big, people lose too fast and don't come back.

(3) Yair Tauman, the casino's expert, is a really smart guy. But I hate that people like that will work for anyone who pays them. Professor Tauman makes a living at public expense. Even if he didn't, he should use his god-given talents to help the truth and settle disputes, not hide the truth and exacerbate disputes.

If I had been hired by the casino, I would have written that Shakleford's data clearly showed cheating and the only issue is whether or not his data are correct. I would have taken half an hour or less to conclude that, and probably would not have been paid.

Writing an expert report and billing a large amount (I'm guessing 40 grand) that just obscure the issues should be beneath Taubman. I know they can find some "expert" to defend them, but a real expert should have more pride.

Webster · #5 05-13-2006, 04:46 PM

To give you a non-answer.
If it is Crypto I've done a LOT of play there and TO ME - it seem freakishly streaky.

I have lost $100 in 40 hands and won $400 flat betting $2. It seems it is backwads cheating at times and just plan cheatng other times.

I have to assume you are playing perfect BlackJack?

I have had your type of streaks at Crypro trying to clear a bonus which is the ONLY !reason to ever play BK online.

BruceZ · #6 05-13-2006, 05:48 PM

[ QUOTE ]
To give you a non-answer.
If it is Crypto I've done a LOT of play there and TO ME - it seem freakishly streaky.

I have lost $100 in 40 hands and won $400 flat betting $2. It seems it is backwads cheating at times and just plan cheatng other times.

[/ QUOTE ]

For BJ that is impossible. You didn't win $10 per hand flat betting $2. Even to lose $100 you'd have to lose over 1 bet per hand. Unless you are playing multiple hands.

Webster · #7 05-14-2006, 10:32 AM

[ QUOTE ]
I have lost $100 in 40 hands and won $400 flat betting $2. It seems it is backwads cheating at times and just plan cheatng other times.

[/ QUOTE ]

well it FELT that way lol it was more like a couple hundred!

Speaking of odd things - last week I was a woman FLOP a Royal Flush and the very next hand flop Quad Aces!

BTW - you COULD lose that much - double downs and so forth.

BruceZ · #8 05-13-2006, 05:35 PM

[ QUOTE ]
Since I have been logging my blackjack and pontoon sessions, I have wagered $8936 with a total number of trials at 3837 (hands dealt) - $2 per unit flat betting.

I am now down 244 units during this sample. The house edge in Pontoon is 0.17% and Blackjack is 0.66%. Pontoon comprises 60% of this sample size. I play PERFECT basic strategy in each game to ensure that the house edge is a reliable indicator for what I can expect to lose over the long run.

I believe the probability of doing this poorly or worse over the size of this sample is 0.37%.

[/ QUOTE ]

It's actually 0.037% or about 4 in 10,000. This is how often it would be expected to happen by chance in a sample this size. You are 3.37 standard deviations below the mean.

SD =~ 1.1*sqrt(3837)

NORMDIST(x,mean,SD,true)

=NORMDIST(-244,(0.6*-0.17%+0.4*-0.66%)*3837,1.1*SQRT(3837),TRUE) =~ 0.037%.

EDIT: I have assumed the same standard deviatoin in pontoon as blackjack, but I'm not sure if this is true. If you find the SD for pontoon, add SD^2 to the SD^2 for BJ (computed for the number of hands of each game played) to get the combined variance, take square root to get the combined SD, and then repeat the above calculation.

TrendYourFriend · #9 05-13-2006, 08:44 PM

i'm going to have to chalk these results up to a legitimately horrible run based on the crazy [censored] that has happened to me the past two days.

In addition to this sample that BruceZ indicated had about a 1 in 2500 chance of occuring, in the past two days, I have:

- been dealt a Royal Flush @ Stars (hit it on the turn)
- dealt pocket 99's three times in a row @ Party
- dealt KTs in diamonds at the same table twice in a row @ Stars (this happened at the EXACT same table as the Royal Flush about 45 minutes earlier)

there's been other things too, but these are the most notable.

but i digress. I thought casino bonus hunting would be a good way to make an extra bucks, which it is based on statistics, but personally, the overlay is not worth the time, especially now that I have some doubts on the fairness of the games.