Stats = Casino Cheating?

Hey all,

Over quite a few months, my friends and I have been keeping detailed logs for an online blackjack game at a casino that does not shuffle every hand. Combined we have run over 200,000 hands. According the data we've compiled, for low counts (when the true count w/ the hi/lo system is < 3, but > 0 bc we wong out) we lose on average probably about $0.02 a hand. This loss is in stark contrast with the 0.005% casino advantage calculated according to the rules of the game when playing basic strategy. I got the basic strategy and casino advantage from www.blackjackinfo.com. Keep in mind that we bet the $1 minimum unless the true count is > 1.7 then we start scalling up by multiplying our calculated advantage (using 0.005% increase in advantage per true count) by our starting and arbitrarily-chosen bank roll of $2600. The most that would ever be bet for these games occurs when the true count = 3 in which case we bet 15.

At first we chalked up the results to human error so we moved from using the basic strategy charts to using computer programs to keep track of everything for us. To decide which move to make we used the BJ2021 program, which counts cards to determine the best possible play. Interestingly, we had greater winnings/hand when using the charts. Anyway, to keep track of the true count etc, one of my friends wrote a simple program. So it seems there was little source of human error the second time around and yet we still had a loss per hand much greater than the casino advantage.

Also interesting was that for when the true count was high (>3) we had negative net winnings as well, but I chalk this up more to standard deviation than anything, especially since we probably only have about 20,000 "high count" hands logged and were betting up to 50/hand before doubling and splitting.

In hindsight, it seems our results would be much clearer had we only bet $1 for all of the hands, but alas we also wanted to test out betting large on high counts.

Anyway, what do you all think. Is this basis to say the casino is cheating? Also, if so, what should I do? I have purposely not said the name of the casino yet because I don't want to be responsible for false acusations.

Much Thanks.
- John

What was your actual win/loss rate. How many wins, how many losses and how many pushes?

They might not shuffle every hand, but if they deal 3 decks from a 6 deck "electronic pack" counting isn't going to work...

The basic strategy chart you used was for the rules of your blackjack game you were playing. Its important to note that not all blackjack games are .5% with perfect basic strategy. So between not flat betting/possible human error/ and a possible less than optimum game, 2 percent loss on the dollar may not be that far from a possiblility.

If you want to possibly persue cheating then keep track of every card that is dealt, including suit, and see if more than the number of possiblilities comes up. If you played 200,000 hands, then surely you may have been unaware of maybe an extra couple of 5's or something of that nature.

Otherwise, you are at the mercy of their shuffling procedure and simply taking their word for it.

Hey guys,

I am not sure of my overall win/loss rate because that is not something we kept track of. We did however keep track of how many hands we had w/ the true count being > 3 and how many of these we won. The overall average for that was about 0.47 and the number of these high count hands was roughly 20,000. I can get the exact numbers if you really want... Basically 10% of the total hands had a true count >= 3.

The penetration was quite poor, probably about 40%-ish. I am not sure what you mean by dealing from an electronic deck. Do you mean an infinite deck? For all intents and purposes my impression from their website is that they are trying to simulate real casinos as realistically as possible. Also it is very obvious when the game shuffles as you see the discarded card pot empty. I have not read anything on their website about an electronic deck

I am quite sure we used the proper strategy cards as I used a respected strategy card engine to come up with them and they also matched what BJ2021 had us doing. I am also quite certain the casino advantage for that game is about 0.5% as I first heard of the game off a respected gambling website which specified it as about that and that value also matched the number given by the strategy engine.

As for human error, I believe it to be low because with BJ2021 program you simply put in the cards that were dealt and it tells you the proper move. The same was true of the program my friend made which is based on the basic strategy charts. We moved from using BJ2021 to using this program to double check that BJ2021 wasn't doing anything funny.

What do you think?

Anyone have anything else to add? Any conclusions? I'm fairly convinced they're cheating and would love to make them pay for it if I could report my findings and logs to some proper organization.