CSM - Can fed in cards be dealt right away?

[Sonny]

As cards are inserted into the top, the drum will rotate at an unknown amount for each inserted card, and the inserted card will become the last card in the compartment.

From what I’ve read, the machine will randomly place the card in either the first or last position in the compartment. One of the machine’s three RNGs is dedicated to this decision.

I said earlier that the drum rotates an unknown amount with each inserted card…When the staff opened the machine, the 2 cards that had successfully been loaded were placed 2 segments apart from each other, with the 3rd card ready to follow suit. In other words, for the first 3 cards, the machine inserted 1 card and then moved 2 compartments for the next one. This is by no means a random sequence, so I have my doubts that the machine would have continued with this very predictable progession.

After seeing two cards inserted into the machine you have no reason to assume anything. If the two cards were 8 slots apart would you change your strategy? In the absence of statistically significant evidence you should assume that the RNG is functioning properly.

What I have seen from my experiences playing many hands from these machines is that they are anything BUT random. Runs of cards and clumping are fairly common.

This has been documented by Shackleford on his website. I think you will find it very interesting. Not necessarily helpful, but interesting.

I have been taught by the skilled players (on the higher limit tables) to watch out for the rare situation when the machine is running hot. Specifically this means there is a nice mix of mostly high cards with a few low cards coming out.

What sort of indications did they tell you to look for? Just be careful that these high rollers aren’t Jerry Patterson disciples. Patterson has been pitching his TARGET system for CSMs for a long time now. They might have you chasing "hot" tables and "biased clumps" that don't actually exist.

Who knows, maybe card counting can provide a way to detect these cycles from the one2six before they actually come about.

Perhaps, but it sounds like sequencing would be a much more powerful technique to use when you consider the nature of the machine. Think about all those little “card sandwiches” being served out by the machine. If you could separate the layers then you might have a good idea of what “flavor” you’re going to get.

-Sonny-

 

[bigbjfan]

Has anyone ever bought a CSM and then reverse engineered it? I would think someone has but may not have shared their findings - don't blame them.

 

[Sonny]

Has anyone ever bought a CSM and then reverse engineered it? I would think someone has but may not have shared their findings - don't blame them.

You cannot buy the machines, they can only be leased from the manufacturer by casinos. However, there are a few consultants that have done extensive tests on certain models and have exposed weaknesses. That's why the first several generations of CSMs were all beaten so badly.

-Sonny-

 

[21forme]

You cannot buy the machines, they can only be leased from the manufacturer by casinos. However, there are a few consultants that have done extensive tests on certain models and have exposed weaknesses. That's why the first several generations of CSMs were all beaten so badly.

-Sonny-

Sounds like restraint of trade to me...

 

[Mr. T]

Here is a description of my current understanding of the one2six shuffler as relayed to me by a sharp dealer:

Firstly the machine can be set to use between one and six decks of cards, hence the name.
For Blackjack, casinos will almost invariably use 6 decks (312 cards).

Inside the shuffler is a rotating drum with 38 separate compartments for storing cards.

As cards are inserted into the top, the drum will rotate at an unknown amount for each inserted card, and
the inserted card will become the last card in the compartment.

The compartments will each store anything from zero up to a maximum of 10 cards.

Once 10 cards are stored the machine is ready to eject these cards into the buffered queue for play,
(in the same order they are stored in the compartment).

The buffered queue stores 1-2 sets of 10 cards at any given time.
Once the machine has detected the buffered queue is down to 10 cards it will then eject the entire contents of another full compartment right beneath the 10 cards that are already there.
In other words 10-20 cards in the output queue at any given time. Thereby ensuring that the front of the machine will never run out of cards.


I said earlier that the drum rotates an unknown amount with each inserted card.
The only evidence that I have on this was one time I got to see the inside of the machine as it was being loaded, this was at the opening of the BJ table and the machine jammed on the 3rd card. When the staff opened the machine, the 2 cards that had successfully been loaded were placed 2 segments apart from each other, with the 3rd card ready to follow suit. In other words, for the first 3 cards, the machine inserted 1 card and then moved 2 compartments for the next one. This is by no means a random sequence, so I have my doubts that the machine would have continued
with this very predictable progession. I will confidently say that the storage disk rotates in a clockwise direction when looking at it from the last box.

Assuming that the dealer popped the cards back into the machine at the end of every hand, (something which rarely happens), the soonest you would possibly get any of those cards back out the other end would be 21 cards later, and it would likely only be only one card. As a general rule, since the inserted cards are going behind the existing cards in the compartments, most of those cards won't come back out for quite a while.


What I have seen from my experiences playing many hands from these machines is that they are anything BUT random. Runs of cards and clumping are fairly common. I am still working out a way to reliably exploit them, but I am happy to share with you what I have already discovered so far.........


It is common for the dealer to draw a 20/21 off a 5 or 6 face card when dealing from these machines. This is a disaster for us, when we always stand on low totals from basic strategy. For this to happen, there needs to be clumping of low cards, which seems to happen way too often. Even when the high cards clump together, the dealer will draw 20 and only the lucky players will get by with a stand off. I have often seen seasoned players break with the basic strategy and treat the dealers 5 or 6 like it is an Ace and hit aggressively on totals up to 16.

So how might we exploit the one2six..........? I have been taught by the skilled players (on the higher limit tables) to watch out for the rare situation when the machine is running hot. Specifically this means there is a nice mix of mostly high cards with a few low cards coming out. These cards will bust the dealer most hands. The table will be full of people, but we will only open 3-4 boxes and will back-bet each other to the maximum allowable amount. There are also a lot of winning perfect pairs bets at these times, usually one or more every hand. I suspect that the cause is the machine burys the 5's and 6's somewhere in it's bowels and the dealer simply can't build a decent hand until they re-emerge. I have seen these fantastic cycles go from anything between 5-30 hands. By the time the cycle ends and everything is balanced again, the dealer will have given away thousands of dollars. These cycles happen on the cheap tables too, but nobody really wins much, maybe a run of a few wins at $5 a pop, for a half a dozen hands. Never lasts long when all seven boxes are open, and the $5 players never know to increase their bets and milk it.

Learn how to spot these cycles, when they happen, and only have a small number of boxes open when they do (to get more hands out of the rare situation), and bet big until it stops. Of course, this effect can happen dealing from a regular BJ shoe as well, but I think that the difference is that the effect and length of time can be far more pronounced from an automatic shuffler.

Who knows, maybe card counting can provide a way to detect these cycles from the one2six before they actually come about.

Good Work.
Are you saying the CSM is random or non random?
You are right. I did notice many times the dealer getting 20 and 21 when his face card is 5 or 6. I have also notice many times when the dealer has 10 or Ace and then bust.
Do you have any comments on my 15 times losing streak presently when I am playing perfect BS agaainst the CSM.

 

[Mr. T]

So how might we exploit the one2six..........? I have been taught by the skilled players (on the higher limit tables) to watch out for the rare situation when the machine is running hot. Specifically this means there is a nice mix of mostly high cards with a few low cards coming out. These cards will bust the dealer most hands. The table will be full of people, but we will only open 3-4 boxes and will back-bet each other to the maximum allowable amount. There are also a lot of winning perfect pairs bets at these times, usually one or more every hand. I suspect that the cause is the machine burys the 5's and 6's somewhere in it's bowels and the dealer simply can't build a decent hand until they re-emerge. I have seen these fantastic cycles go from anything between 5-30 hands. By the time the cycle ends and everything is balanced again, the dealer will have given away thousands of dollars. These cycles happen on the cheap tables too, but nobody really wins much, maybe a run of a few wins at $5 a pop, for a half a dozen hands. Never lasts long when all seven boxes are open, and the $5 players never know to increase their bets and milk it.

Learn how to spot these cycles, when they happen, and only have a small number of boxes open when they do (to get more hands out of the rare situation), and bet big until it stops. Of course, this effect can happen dealing from a regular BJ shoe as well, but I think that the difference is that the effect and length of time can be far more pronounced from an automatic shuffler QUOTE FROM MARQUOISE

It sounds like just the old Superstition to me.

 

[Sonny]

So how might we exploit the one2six..........?

Sandwiches.

-Sonny-

 

[dacium]

How hard is a CSM anyway?

Consider a 7 person game where +1 TC is even and +2 TC is advantage to the player.

At a 7 person table the average amount of cards delt in one hand is about 24 (more if lots of ploppies playing). All of these are in the out slide already before the previous hands cards get shuffled in.


For 6 deck CSM that leaves 288 cards, or about 5.5 decks. So to get +2 you need 2*5.5 = 11. So how often is a single hand end with a +11 count? You would be surprised, its rare, but seems to happen.

 

[Sonny]

At a 7 person table the average amount of cards delt in one hand is about 24 (more if lots of ploppies playing). All of these are in the out slide already before the previous hands cards get shuffled in.

Are you sure about that? What if the first round uses up all (or most) of the cards in the “snout” of the CSM? Does the machine shuffle the used cards first then reload the snout? If so then the machine will be dealing from a full (or almost full) shoe on each round and there will be no latency for some of those cards from the first round.

I don't have an answer to those questions, but they are very important things to consider.

-Sonny-

 

[Ringer]

Are you sure about that? What if the first round uses up all (or most) of the cards in the “snout” of the CSM? Does the machine shuffle the used cards first then reload the snout? If so then the machine will be dealing from a full (or almost full) shoe on each round and there will be no latency for some of those cards from the first round.

I don't have an answer to those questions, but they are very important things to consider.

-Sonny-

It does not require a shuffle before offering the cards in the shoe of the machine... at least in the machines I've studied.

 

[marquoise]

Xtra info

OK I have some new information since my earlier post which may be of interest.

Assuming that the insertion of the cards into the one2six is random, it turns out that way the machine outputs the piles of cards is not.
As a general rule, it seems that the machine will eject one pile of cards, then move one slot around and eject the next and so on. I won't say that it will do this ad infinitum, but if you ever take the chance to look inside an opened up machine that has already ejected a few cards, you will notice that 95% of the time the empty compartments are all together.

I also had the good fortune to be playing a game recently where one of the cards in the shoe had a black smudge on it. This was on a 4 deck pontoon game.
After each time I saw the marked card, I would count the number of hands until it reappeared again. Each hand had 2 boxes open (plus the dealers) for an average of 8-15 cards per hand. The soonest it reappeared was 3 hands later. The latest it reappeared was 32 hands later. But on average it took between 20 and 30 hands to reappear. This basically confirmed what I had already stated to this group but it is real data for those who might need something tangible.

Secondly, I have copped a little bit of heat regarding things I had included in my initial post. That is fair enough. The initial post actually had 2 sections, the first section was regarding actual facts, answering several questions first proposed in the beginning of this thread (which had not been adequately answered by anyone up to that stage). And the second section was speculation on my part on how the machines might be beaten. In hindsight I should have left the speculation out of it. Fear not, I won't make that mistake twice.

I am very curious about the information that the one2six can insert a card to the front or back of a compartment, this is at odds with what my friend the dealer told me, so I will endeavour to verify it.

Marq

 

[alienated]

If you discover something, keep it to yourself

This thread is interesting in some respects, but I don't understand why you guys would want to discuss your ideas about this topic on a public forum.

I don't know if you'll be able to find a way to beat the machines or not, but if you could manage it, it would be a much better situation for you if your discovery was not widely known.

My guess would be that if you found something, it would be somewhat weaker than the edge you could get from a regular shoe game. This does not mean such a discovery (assuming for a moment it is possible) would be worthless - but it would only have significant value for as long as it remained relatively unknown to others. Otherwise it would become no better (in fact worse) than traditional shoe games with widely understood (and hence protected) vulnerabilities.

In the past, I felt that posting about possible flaws in CSMs - or even alluding to supposed "flaws" without any basis in fact - might be a way of somehow deterring casinos from using them. I no longer consider this to be an effective tactic. If a vulnerability in the current models happened to exist and became widely recognized (not just by the casinos but, more critically, by many players), I think this would pose at least two dangers. First, and most obviously, it would merely provide an impetus for a new, improved model to be introduced that nullified whatever edge had been uncovered and broadcast over the internet by players. Second, I think it threatens the longevity of playable blackjack itself. If casinos felt they couldn't "protect" themselves from skilled blackjack players even with the use of CSMs (irrespective of whether this view arose as a result of deliberately false information purveyed by players), they might try even harder to slip in inferior blackjack games and replace them as much as possible with other, less favorable games. How long will it be before 6:5 multideck blackjack becomes the norm? Dealer wins on ties? Player blackjack loses all to dealer 20 or 21? There is very little to stop any of these rules changes being instituted other than informed gamblers voting with their feet (I won't hold my breath).

But from an advantage player's perspective, the biggest reason for keeping any discovery to yourself is that the opportunity will only last in its strongest form for as along as it remains relatively unknown. If you really found something and posted it here, and it was actually valid, how long do you think it would be before big teams came down from a great height and killed the game? Before you could blink an eye, intimidating fellows would be spilling drinks on your betting box or in your lap to expel you from the tables. And all because you shared your secret for free. There would be no one to blame but yourself.

All this may be moot in any case, since (as indicated earlier) I don't know if any of you will be able to find a vulnerability with one or other of the current models in use. I'll admit I'm skeptical. In fact, the lack of any mention in company reports of greater-than-usual drink spilling at the CSM tables is probably one of the more reliable indicators that these machines are free of any huge vulnerabilities. Either that or the big teams haven't looked closely yet (I'd doubt this) and a tiny number of solo players have kept any secrets they might have to themselves. I'd suggest others do the same.

 

[zengrifter]

This thread is interesting in some respects, but I don't understand why you guys would want to discuss your ideas about this topic on a public forum.

I don't know if you'll be able to find a way to beat the machines or not, but if you could manage it, it would be a much better situation for you if your discovery was not widely known.

Its our motto: All public, all the time!

Say, Ted, why don't you post that ground-breaking analysis of yours on the subject.

Don't worry about it being public, no one, including me will understand it! zg

 

[Mr. T]

Sonny is right......( this one time)

I have played many hours of BJ against the CSM and I have re-read this thread and what Sonny says about Marquoise is correct.
He actually said the CSM is "anything but random" having seen the inside of the machine a few times. Need better material than that to make that claim, which is against all conventional knowledge.
To play against the 5 & 6 like it is an Ace is unthinkable.
To open only 4 or 5 boxes and not the full table is superstition.
To bet very high in the so-called streak or when the dealer is "hot" is playing by hunch. Come on you can do better than this.

 

[bigbjfan]

I took a quick trip to the casino with the CSM's to practice my counting speed the other night and came to the conclusion that the way I was counting these things wasn't correct. I'm going to apply what I've learned since from this great thread, regroup, and then give it another try to see what I find.

 

[Sonny]

Sonny is right......( this one time)

Well, maybe not! After looking at the patent number I can see that he is talking about a completely different CSM than I was. My research has been with the models that use a round “carousel” of cards, not the “elevator” type that he is talking about.

In any case, I stand by my comments that much more research has to be done before any conclusions can be made.

-Sonny-

 

[Sonny]

I don't know if you'll be able to find a way to beat the machines or not, but if you could manage it, it would be a much better situation for you if your discovery was not widely known.

Howdy alienated! I haven't seen that name for a while. Good to see you here.

Don't worry. I'm sure anyone who makes a breakthrough will keep it to themselves.

-Sonny-

 

[taipafan]

Once the machine has detected the buffered queue is down to 10 cards it will then eject the entire contents of another full compartment right beneath the 10 cards that are already there.

Marq, your info is very useful. Thanks. Here is my question:
Do you mean only for those compartments with 10 cards will be ejected, but not for those less than 10 cards?
There are chances for all compartments hold less than 10 cards.

 

[dacium]

If the machines do dump the whole trays off the carrasell, then clearly they are far from random.

Its practically impossible to find out how these work exactly without having one to work with. Assume you are lucky enough to find a table with a single marked card so you can follow it out and you are counting its re-occurances. If its 6 decks then you expect the average repeat to be 3 decks (156 cards). Now work out the variance and standard deviation and to get an accurate result you would need to track the card through about 300 times over to get an average result that is highly likly to be within +/- 10 cards.

This is about 10,000 hands if you are playing just with the dealer.

 

[alienated]

Hi & a minor point

Hi Sonny. I'm still around. I enjoy reading everyone's posts. Just don't have much to say these days ...

If the machines do dump the whole trays off the carrasell, then clearly they are far from random.

Actually, this doesn't follow. As long as cards are inserted randomly, it doesn't matter how they are ejected. (If they are randomized at one point, they don't need any more randomizing.)