Shuffling discussion

[donkeyradish]

This isn't one of those 'the shuffling at site X is biased' topics. I'm just wondering about the mechanics of shuffling on-line.

For example:

Can we be sure that an on-line shuffle is really a shuffle (the whole deck is randomly ordered before any cards are dealt), rather than - each card being randomly selected from the remaining pool of cards as the hand progresses? Would it make any difference?

Assuming on-line cardrooms do some kind of actual shuffle, then how could it ever be ensured/verified that all possible orderings of 52 cards are equally likely?

[Myrtle]

....I'd probably post this in the "Poker Theory" Forum where you might get some more responses.

There are many different thoughts surrounding this subject.....

It's a tricky one.

/images/graemlins/confused.gif

[Rudbaeck]

Of the sites that publish an outline of their shuffling procedure Ultimate Bet is the only one that can even generate all 52! possible decks.

The rng seed has to be 226 bits long to cover all possible decks. Many sites use significantly smaller seeds. /images/graemlins/mad.gif

What is worse is that many sites don't use a physical device to get their seeds, but rather some pseudorandom function. UB relies on thermal noise from a zener diode. Thermal noise is entirely random. It beats the snot out of any professional Vegas dealer for sure.

From a physics stand point thermal noise is demonstrably random. If it's not then the entire quantum physics model is wrong. And someone who could figure that out is probably better off in another career than cheating at online poker. /images/graemlins/grin.gif


As far as I know no site shuffles the deck and leaves it for the entire hand. They ALL just select a random card of the remaining cards. I can't think of any way this can be bad, but I can think of one way it's incredibly good - no hacker can tap into your poker room and sneak a peak at the deck, because it will be reshuffled for every card that is dealt.

[moondogg]

[ QUOTE ]
This isn't one of those 'the shuffling at site X is biased' topics. I'm just wondering about the mechanics of shuffling on-line.

For example:

Can we be sure that an on-line shuffle is really a shuffle (the whole deck is randomly ordered before any cards are dealt), rather than - each card being randomly selected from the remaining pool of cards as the hand progresses? Would it make any difference?


[/ QUOTE ]
Given the assumption that nobody at the table has any foreknowledge of the ordering the of cards, I don't see why it would make a difference. It's a matter of random fate. The next card off will be determined at some point, but will be random until that point. In the end, from the perception of the players, the card is completely random, regardless of when it was determined. If I pick a number out of a hat and tell you what it was next week, it makes no difference whether I picked it today or tomorrow. Assuming that the numbers where randomly arranged in to start with, if I pick two numbers from a hat, it does not make it any more or less random if I shake the hat again between between numbers.

The only argument I see here is that shuffling in the middle of the hand will throw off the "integrity" of the deck. However, this is just a superstition, along the same lines as Blackjack players claiming that someone messed up the shoe because they cut in a bad place or split tens.

I was playing with some friends last weekend, and the game was fairly fast paced, with the next dealer shuffling the deck for the next hand (I've known and played with the guy for years, it wasn't a trick, it was absentmindedness). In the confusion, a few times the current dealer started shuffling the stub of the deck for the current hand after the flop. I was amazad that some people thought it made a difference.
[ QUOTE ]

Assuming on-line cardrooms do some kind of actual shuffle, then how could it ever be ensured/verified that all possible orderings of 52 cards are equally likely?

[/ QUOTE ]
FWIW, most sites have their random deal approved by independent auditors to ensure that the results after an extended period of time match the expected patterns. For example, Paradise's deal is audited by PriceWaterhouse Cooper, and the results are published on Paradise's website. You may take whatever comfort your like from that.

[drewjustdrew]

[ QUOTE ]
As far as I know no site shuffles the deck and leaves it for the entire hand. They ALL just select a random card of the remaining cards.

[/ QUOTE ]

I was talking with Howard Lederer at a Fulltilt promotion party during the WSOP and he told me something similar. He said for their site, the cards are continually shuffled throughout the hand, which is essentially the same thing. I can't speak for the other sites.

[kem]

[ QUOTE ]
The rng seed has to be 226 bits long to cover all possible decks. Many sites use significantly smaller seeds. /images/graemlins/mad.gif

<snip><snip>

As far as I know no site shuffles the deck and leaves it for the entire hand. They ALL just select a random card of the remaining cards. I can't think of any way this can be bad, but I can think of one way it's incredibly good - no hacker can tap into your poker room and sneak a peak at the deck, because it will be reshuffled for every card that is dealt.

[/ QUOTE ]

If your second point is true (that cards are randomly chosen from the deck), then w/ regard to your first point (initial order of deck) it doesn't make one bit of difference that the random seed isn't large enough to map 1:1 to every possible ordering of the cards.

If the procedure to pick a card "randomly" out of the deck is truely random, sites should be able to not even shuffle the deck -- when they need a card, they just randomly pick one from the remaining cards. It shouldn't make one bit of difference what the initial ordering of the cards is. I suspect that sites shuffle the deck because (1) it takes virtually 0 time, and (2) to keep clueless people from saying "I'm not playing on that site, they don't even shuffle the deck!"

[carlo]

About 2 weeks ago I sent a e-mail to p-stars about this very question. Support answered that the cards are shuffled prior to each hand and the deck is set They also use thermal noise and mouse movements to adjust their random number generator which is an INTEL device(hardware or software?). The stars site discusses their shuffle(random noise,mouse movements,etc.) but doesn't specifically discuss the time of the shuffle and thats why I asked the question. In a more practical sense I have a habit of seeing what I would have had if I had stayed in--the point is moot if the shuffle is time dependent and continuous for if I had played, this would set the shuffle to a different time pattern and I therefore would never know what I would have had. But if the deck is set as in a live game I would be able to second guess myself.Yuk!LOL--If anyone could confirm the timing of the shuffle I would certainly like to know. Thanks-carlo

[moondogg]

[ QUOTE ]
In a more practical sense I have a habit of seeing what I would have had if I had stayed in

[/ QUOTE ]
This is a bad habit. It makes no difference what you would have had. If you should fold, fold. Whether you would have caught your miracle card is not relevant. You can't predict the future, so stop trying.

[ QUOTE ]

--the point is moot if the shuffle is time dependent and continuous for if I had played, this would set the shuffle to a different time pattern and I therefore would never know what I would have had.

[/ QUOTE ]
The point is moot either way.

[ QUOTE ]
But if the deck is set as in a live game I would be able to second guess myself.Yuk!LOL--If anyone could confirm the timing of the shuffle I would certainly like to know. Thanks-carlo

[/ QUOTE ]
I can confirm that it makes no difference one way or the other. If it makes you feel better, just tell yourself that the cards are all shuffled before hand.

[carlo]

You're too harsh,pompous and self righteous.I'm way ahead of you in this matter but the original post is thoughtful and asks a pertinent question. It always helps to know what you're dealing with(no pun intended) and if you believe in this artificial randomness to the extent of your post then I suggest you have been captured by gypsy's but don't worry,we'll come to your rescue in spite of you.-carlo

[moondogg]

[ QUOTE ]
You're too harsh,pompous and self righteous.I'm way ahead of you in this matter but the original post is thoughtful and asks a pertinent question. It always helps to know what you're dealing with(no pun intended) and if you believe in this artificial randomness to the extent of your post then I suggest you have been captured by gypsy's but don't worry,we'll come to your rescue in spite of you.-carlo

[/ QUOTE ]

I apologize if my post came off as harsh or pompous, as that was not my intent. My response may have been self-rightous, but I make no apologies for that.

Regardless, the issue makes no impact on the results of the game in any that can be predicted or accounted for. I do not believe it actually helps you at all, and can mislead you into what would have been the correct play would be in a given situation based on "what would have happened." This is a common mistake that many newer players make, as demonstrated by some people's insistance on "rabbit hunting."

Bring on the gypsys!

[carlo]

You're right,it doesn't help to second guess yourself but there is a non linear pattern arrangement which manifests itself and I plan to imbibe it(yes,the gypsy's have me)--I really don't second guess myself and in fact if calling or folding pre flop were all there is to the game I'm batting in the 99th percentile. But I do always relate the flop to my discards much like one does not think about how to read--could be fixed,but I really don't respond emotionally to this or in any abberational manner. Thank you for your kind response to my confrontational post. Good Luck,--carlo