single deck and double deck automatic shuffler. Is it random or does the shuffler actually adjust or stacked cards to decrease ap advantage?
Short answer: Both kinds of shufflers are pseudo-random but not purposely cheating. There is no need, since the majority of customers will not be nearly good enough to overcome house advantage or inadequate capitalization to necessitate illegal measures against them.Originally Posted by Nick12345
Long answer: Both continuous and automatic shufflers are designed to quickly and pseudo-randomly shuffle decks of cards using mechanical means. They do not stack the deck against the player on purpose. Whereas continuous shufflers continuously shuffle cards used during each round together in a 3-to-5-deck shoe before the new mixed-up shoe is used to deal the next round, automatic shufflers simply shuffle the cards after the cut card comes out, but leave the cards used during a given round set aside to not be replaced into the shoe before the next round starts. Thus, AP advantage is decreased in the first case since it is impossible to count such a game, but it is not decreased in the second case because the shuffler merely takes on the role of the dealer in shuffling cards to begin a new shoe. CSMs are expensive to purchase and to maintain relative to any profits that could be gleaned from tables using them, and ploppies don't trust CSMs anyway because they perceive the machine as making it far too easy for the casino to stack the deck against them. Thus, they're not that common and can be easily avoided. I strongly suspect that ASMs are much in the same boat, in that they don't stack any cards against the player because that's not what they're designed for. Skillful mechanics can do that much more reliably, invisibly, tactfully, and cheaply than ASMs and CSMs can. Even if they're countable because they don't mix recently used cards back into the shoe for the next round, you should still avoid ASMs if you can't recognize the difference between ASMs and CSMs on sight and/or if you want to help discourage their use.
I'm a little confused by the descriptions and terminology. I would think an "automatic shuffler" is simply the mechanical device into which you place one of more decks of cards, and it shuffles them and returns the cards in a random fashion.
I believe the ASM's can recognize the faces of cards, and uses that information to flag a deck (here using the term "deck" to mean one or more 52-cards decks) that does not have 1/52 of the cards matching each rank/suit.
I had asked my shift manager for a couple of hours alone with an ASM and 8 decks of cards to conduct an experiment. I was going to arrange the cards in sequence (i.e. all 8 spade Aces first, then the 8 spade Kings, etc.) and then see if the resulting shuffle had consecutive cards of the same rank/suit/etc. more or less often than would be indicated at random. He said no (he always says no to my experiments) but it might well be that there's a license issue with the shufflers. I had tried to get information from the manufacturer on how they work, but no luck there either. I thought that there must be some promotional material they give to prospective buyers that wouldn't reveal trade secrets.
I feel confident that the shufflers (ASM or CSM) aren't hacked. That would be killing the goose laying the golden eggs. Swapping a card mechanic into a game with a high-roller - well, you can't get a search warrant to seize someone's hands the way you can to seize a mechanical/electronic device.
Remember this: Readers of this forum mostly include card counters and those who play perfect or near-perfect basic strategy and would like to count but don't feel it's worth the time/effort/risk. Players at the table include counters (well under 1% in my experience), those who play perfect basic strategy (maybe 10%), and an overwhelming majority who either don't know basic strategy or feel it's not playable because it basically doesn't work. If I had a dollar for every time a player has stood on 15 against an 8 or 9 and said "I'll see what you've got" I'd have my new house paid for by now. That, plus the panoply of side bets, means the casinos have no need to cheat.
Remember too that the manufacturers would not only have to engage in the considerable expense of rigging the shuffling algorithm, but also have some kind of backdoor key to unlock the rigged algorithm and then to make it known only to customers who (a) pay for the access and (b) can be trusted not to rat them out.
When in dealer school, I once tried an experiment. We didn't have automatic shufflers in school, so when it was someone else's turn to deal I asked him to let me shuffle, and I basically took a bunch of small cards out and then shuffled them into the first half of the resulting deck, artificially creating a shoe that would have an ever-increasing true count. (As far as the dealer getting practice, it doesn't matter to the dealer if the shoe is random or not.) On that particular occasion I still lost (play money of course) on the shoe, and I wasn't sufficiently motivated to try it for another shoe or two.
It would probably be possible to hack a CSM. The machine could be programmed to observe the dealt cards and then would know how they were distributed based on the timing; i.e. if they dealer deals 8 cards and they are 10/8/5/6/8/9/4/10, then a pause, the machine would know there are three player hands of 10-8, 8-9, 5-4, and dealer 6-10. The machine will know that the third hand will take one card, and as the next card is always available, that card can't be controlled, but the machine might surmise that after the next card is dealt, the one after that will go to the dealer, and move a 5 into the proper position.
But as mentioned above, that's a huge risk to the casino and the manufacturer.
Last edited by redtop43; 09-15-2023 at 01:27 AM.
Let me address the question of "pseudo" random. It's probably been addressed elsewhere.
Purely in a computer context, random numbers are always generated by a seed, and then an algorithm is performed on that seed to get the next seed. They are always scaled between 0 and 0.99999...
Here is a simple formula. Take a seed of x. Add pi to x, raise it to the 5th power, and drop the portion to the left of the decimal point. Use that number, then perform the calculation again.
This is what a "pseudo" random number generator is. It's not perfectly random but it's as random as you can get and can practically use. The one I postulated is not a particularly good one as there are biases built into it, but there are way better algorithms that this one.
When I write code to deal a deck of cards, I assign a (pseudo-)random number to each card, then sort them by the random number.
Now, I'm not sure what this has to do with automatic shufflers. I have no idea how they work mechanically. I just put the cards in and press a button and a new deck pops up. (Why can't they make laundry this easy???) I don't know how the cards move. I do know that there are USB ports on at least some models, because our weekly casino operations memo said that when the pit manager opens a game with a shuffler they should check to make sure nothing is inserted in the USB port. How 'bout dem apples?
Many many years ago, when they first started using automatic shufflers, they actually had ones that glass windows in them and you could see the cards being shuffled. The interesting thing to me was those machines did not use a one-to-one interleave. It was more like five to one. Like five cards from one half stack and five cards from the other half stack being shuffled to one stack. This looked like it was controllable also. I could not relate this interleave to winning or loosing. I think the bottom line here is to speed up the game. Then, they go and ruin it with the silly side bets that take the dealer a lot of time trying to figure out the pay offs.
Say you create a deck with small cards removed so it has a TC of +5. That's a 2% edge. It is a lot for a card counter but you won't notice any difference over a short session of play.
Over time however that 2% edge will really stack up. (I'm not making any comment on the viability of manipulating the shufflers in some way, just the mathematics)
2 commentsSay you create a deck with small cards removed so it has a TC of +5. That's a 2% edge. It is a lot for a card counter but you won't notice any difference over a short session of play.
That should translate to a lot of difference in EV in a short session subject to comment 2.
If at TC+5, you still have no edge if the count continues to climb. That’s one of the reasons why opposition type betting can be so effective - it’s also one if the reasons you get hit with high negative variance - Think about it.
Posting Permissions
-*-
Reply With Quote
Bookmarks