:\

I urge you to think about what you're saying

I suspect (no wait I'm positive) the first few people who mentioned a superuser on UB got the same treatment.

I commend OP for sticking to his guns and continuing to gather data. I'm going to do the same for when I'm at the casino (when they are using the shuffler). Even if it's a few hundred hands over the next while, at least it's another data source.

Anything specific I need to know (model number of shuffler etc) to make my results useful to you OP?

I'm only got this far before rolling my eyes.

That was not the point that Tom was making. In fact, the more hands that you do record that did reach the river, the more your results are skewed. That was his whole point, and the fact that you think he was saying something else entirely is proof enough that you should not be the one conducting this data gathering and analysis.

To say "I agree with you that not all hands that I recorded saw the river" is to say you didn't actually try to understand Tom's post. You have people in this thread who know what they are talking about and who are offering help, and you are not even giving their responses the respect they deserve by reading them. Come on.

I'm sure I'll have more to say once I finish reading what you wrote, but this is seriously ridiculous so far. This is a joke.

If you are seriously proposing that as a fair analogy, which it seems you are, then I really urge you to look for the differences between the UB revelation and this one.

That is to say, even assuming a worse case that Shufflemaster machines deal 100% predictable hands, the OP has gone about the investigation in just about the worst way imaginable.

If there is a problem with the machines, then the discovery of that problem has been delayed by the OP's sloppiness, ignorance, and unwillingness to listen to those who can assist him. It basically amounts to fear mongering as it is.

OP, if you go back to the same casino and sit at the same table and somehow the same machine is still there, sit there for 48 hours, watch a lot more hands and record everyone, you might see the data more evenly spread out. You would also have the start of some usable data since it all wouldbe from the same machine.

There will be variance, the dealers will each cut decks in different ways. Have you coinsidered that?

The Nevada Gaming Control Board has an Electronic Services Division that will pull machines, slots, video poker, shufflers etc, into their lab and verify they are operating the way they were when the NGCB approved their use.

Why doesn't someone who believes the shufflers are not random call the NGCB and ask them if they have discovered any malfunctioning shufflers?

It's my understanding that they fine casinos if they are using machines that are not operating as they should be.

The NGCB does not just approve then forget, they do in audits and inspections all the time.

I'm sure that someone will claim that the NGCB won't admit to such a thing, but what would it hurt to ask?

If a casino took the position that they can offer poker like other table games or a house advantage game who would know? Does poker have it's own special regulations? There has to be a way, for example, for a casino to introduce a new game. How does that work. The chief thing about a casino game is the odds or house edge. Is that the basis for a games regulatory profile? If so the casino could be "cheating" from a players perspective but not from a gaming commission perspective if they just said this game poker is a class whatever table game with these rules.

And don't forget we are not just talking about Nevada.

On the subject of suit clumping....you probably know that in hold'em when you are dealt a pocket pair it increases the odds that another player will be dealt a pocket pair. In the same way, if one suit is clumped then the chance that other suits are clumped increases. Let's say a truly random first flop comes out with, say, all spades and a few players stay in trying to catch, they fold and the A of spade semibluffer takes it down. They all muck one after the other, the dealer wipes up the board and throws it on the muck, clumping a bunch of spades. If the deck is not subsequently shuffled enough, then that deck will retain the random clumping that first appeared. The next board will be more likely to be all of one suit than if no suit clumping had ever occurred.

So your now claiming a conspiracy? The casinos are building in a house edge when they are not risking any money? Really? Why? What would they gain by not using shuflers that produced decks that have not been randomized?

Any gaming device used by a casino has to pass NGCB regulations. It has to produce random numbers if that what it claims to do. RNGs are not adjustable, they are hard wired into a chip.
Casinos do not buy the shufflers, they lease them and the lease includes support. If a shuffler stops working properly, it gets fixed. If a chip fails, the shuffler dies.

Before you had a nice little theory about how there might be a flaw in the machines. Improbable, but no harm in investigating to get some real data.

Now, on the other hand, now you've just gone down the rabbit hole.

If you are going to continue to collect data, you should count all hands that you see flops. There should be three "buckets:
bucket 1: # of hands that did not or could not have 4 to a suit. This includes rainbow flops and turns that have a max of 2 of one suit - and never made it to the river.

bucket 2: # of hands that did have 4 to a suit (even if they never made it to the river).

bucket 3: # of hands that might have had 4 to a suit but never made it to the river.
3A) 1-suited flops,
3B) 2-suited flops,
3C) 2-suited turns that had 3 of one suit.

I would be very very surprised if the data still tilted towards "suit clumping" after taking all of these hands into account.

I have seen enough of this thread.

I have peviously questioned the ability of the Deckmate to provide a truly random shuffle and have been shouted down on this forum for having insufficient data to back my allegations.

I will be the first to admit that my limited data sample over the past couple of years was insufficient and in hind sight flawed. My perception that the Deckmate was generating an inordinate number of boards with 3 or 4 or 5 to a flush was incorrect. The real situation is that the holdem boards seem to be coming with non-random clumps of cards - be it three or four to a flush or straight. After further analysis of my data, observing another 6 months of holdem boards and information provided from some of the responders to my query from last year, I realized that the situation is more easily explained if you accept the premise that the Deckmate does not randomize the deck. I know that this is a very hard pill to swallow but please hear me out.

If you accept that premise, then you can look at the standard dealing process for holdem at the end of a hand: the winning hand or hands and the board cards are mucked together and placed on the bottom of the discard pile. These cards (the board cards and the live cards at the showdown) are normally matching/connected - be it a staright or flush or trips. This means that the remaining portion of the deck -after the hand is over - is non-random because it is without the board cards and live cards.

At this point, nothing is abnormal with the deck after a hand is completed. Again, if you accept the premise that the shuffle will be flawed, the non-randomness of the deck will not be eliminated.

I have an excerpt from Shufflemaster which I will paste in at the bottom of this post. In this, they state that the deck is shuffled by grippers grabbing cards in a random manner and placing them in different locations.

So again, I submit that the Deckmate is doing a lousy job of randomizing the deck - based on my many years of experience and the limited data that I have recorded.

But wait, instead of demanding that Shufflemaster be damned, let's think about what can be done to randomize the deck to the benfit of the poker community.

I have no doubt that having the dealer scramble the deck before placing the deck into the Deckmate would eliminate the problem. However, this slows the game down, defeats the purpose of a shuffle machine and would not be acceptable to the poker room management because time is money.

I just returned from Montreal, Canada and witnessed the answer - a truly simple solution!! The dealers there are instructed that, before they place the deck into the Deckmate, they riffle-shuffle the deck one time. It only takes a few seconds.

I played there for twenty hours over three days and I was quite impressed with the normal boards that were produced.

Now before everyone starts shouting me down, I am going to ask the card room where I usually play to change their process to have the dealers do a quick shuffle (instead of cutting the deck) before placing the deck in the Deckmate. I will let you know if they agree.

I will get back to you in 6 months and let you know if what I observed in Montreal continues locally.

Best regards,

Shufflemate statement:
Although Shuffle master seldom comments on discussions involving its products via internet forums, we thought that it was important to correct some misconceptions regarding our Deck Mate shuffler as discussed here. The Deck Mate shuffler offers poker players several advantages over hand shuffling. These include a more random shuffle then hand shuffling, and a faster paced game, since there is no waiting for the dealer to shuffle. It is for this reason that the Deck Mate is employed on over 80% of all poker tables in the United States.

For poker use, the Deck Mate shuffles one deck of 52 cards. Two decks of cards are generally used (usually different color decks) for game play. While one is in play, the other is being shuffled. The cards are counted, but not identified as to rank and suit, as they are shuffled in. The shuffle is aborted if there are more or less than 52 cards. At the beginning of each cycle, a random number generator assigns a new deck position to each card of the un-shuffled deck. The cards are loaded into a feeder. As the cards are feeding in (from the bottom of the deck), the program executes the outcome determined by the RNG. That is, using a moving platform and set of grippers, each card is put into its newly determined position. A moveable platform positions the cards in the shuffling compartment to the correct location. A set of grippers will grip the cards at that location and the platform will move down creating a gap for the next card to be inserted. This process is repeated for each card, until all cards are shuffled in. So, unrelated to rank and suit, the cards are randomized via this method. As an example, the first card in may be the 25th card of the new deck, the second card in may be the 43rd card of the new deck, and so on.

The Deck Mate does not contain an apparatus to recognize the rank and suit of a playing card; therefore, it is impossible for the shuffler to “create” any combination of cards as desired by a programmer. If, in future versions of the Deck Mate, Shuffle Master’s patented card recognition technology is used, this will be for deck verification purposes only, and will not be used to order the deck. This technology is already in use in several of Shuffle Master’s new third generation products and the technology has been carefully and thoroughly examined by gaming regulators around the world. In fact, the manner in which data is stored in the shuffler prevents any such use, since the card position data and card value data cannot be combined until after the completion of the hand. In fact, all Shuffle Master shufflers operate with hardware and software that has been evaluated and approved by numerous gaming regulators/test labs including those in Nevada, New Jersey and Mississippi, as well as Gaming Labs International.

Josh Marz,

Vice President of Product Management

Shuffle Master Americas

I think the windmill Forum is the other way.

Remember, one of the major arguments against you has been is that nobody sees the biased boards you report. You are trying to offer an explanation for a phenomenon that nobody else sees.

I agree with Angus. Seems like solving a problem that isn't really noticed by most players. It may not be perfect Random Number Generation but close enough.

So then you can predict the board?

Because as far as I am concerned if its not predictable its random......

Regarding posts on the Shuffle Master DeckMate. Some have suggested that a cut after the shuffle would safegaurd against non-random shuffles or setup decks.

In at least 3 major poker rooms in LV and CA they instruct the dealers to cut the deck as close to center as possible. I have requested for more random cuts not in the center but each time the dealer said they weren't allowed and that they were required to cut the deck as close to center as possible.

At all three locations the manager confirmed this but denied that the decks were being setup. I argued that cuts in a specific predetermined location serve no purpose and should be random. They either just repeated that the decks were not setup or in one instance said that players might complain if the cuts not were always in the middle.

I've tried to center cut a deck myself to see how often I hit the exact center. With no practice, in 10 tries I hit it 7 times. Each time I would riff and spread the deck several times to ensure that air between cards would not bias the next cut.

You don't need to trust me on this. If your card room does a cut after shuffle, request a non center cut yourself and see what happens.

where I play the dealer lets a player cut the cards (like blackjack), this cuts down on the auto shuffler stacking the deck. this is important because the guy that built the shuffle program for shuffle master plays in the game.