BJ authority Ted Forester has indicated (below) in a recent CCCafe post that a certain-model CSMs are beatable. zg
--- In [email protected], ted_forrester
<no_reply@y...> wrote:
>
> I don't really want to go into details on this question. What I
will
> do in this post is present some representative SCOREs and related
> data for a widely used, current model CSM under the assumption
that
> only count-type (or compositional) information is considered by
the
> player. This means that the following results completely ignore
> available gains from sequencing, boundary effects, side bets, and
> other methods for acquiring hole-card or next-card information. I
> will not specify the machine for which these specific SCOREs have
> been derived (approximated, really), but I will say that it is a
new
> model, and that comparable results could have been presented for
the
> other state-of-the-art models currently employed in casinos. The
> results may leave some readers scratching their heads, especially
if
> they persist in thinking in terms of latency. The vulnerabilities
> come from specific knowledge of how the machine `shuffles', and
not
> necessarily from any sluggishness of cards to re-enter play upon
> reinsertion into the machine.
>
> Two sets of results are presented. In the first set, it is assumed
> that the player places wagers during all nonnegative true counts,
> but sits out or leaves at any negative true count. In the second
> set, the player takes a similar approach once s/he has joined a
> table, but only enters on true counts of +2 or greater. Except
when
> otherwise indicated, the player is assumed to play 2 boxes
whenever
> the optimal bet is at least 2 kelly units. By convention, it is
> assumed that the player observes 100 rounds per hour and has a
> $10000 bankroll.
>
> SET 1. BET IN ALL NONNEGATIVE ROUNDS
>
> . Bet . . . .Kelly . . . . . . . . . . . . . . . . Win/100
> Spread. . . . Unit . . . SCORE . . . IBA (%) . . . (units)
> ----------------------------------------------------------
> Flat 1 box. 41.10 . .. . 12.46 . . . 0.524 . . . . . 0.30
> Flat 2 box. 29.94 . .. . 18.14 . . . 0.524 . . . . . 0.61
> 2:1 . . . . 59.35 . .. . 42.54 . . . 0.911 . . . . . 0.72
> 4:1 . . . . 50.51 . .. . 72.19 . . . 1.252 . . . . . 1.43
> 6:1 . . . . 45.66 . .. . 84.97 . . . 1.531 . . . . . 1.86
> 8:1 . . . . 36.10 . .. . 93.27 . . . 1.566 . . . . . 2.58
> 12:1. . . . 31.15 . . . 101.99 . . . 1.758 . . . . . 3.27
> 16:1. . . . 25.91 . . . 105.28 . . . 1.882 . . . . . 4.06
> 20:1. . . . 21.98 . . . 107.98 . . . 1.952 . . . . . 4.91
> 40:1. . . . 12.58 . . . 113.50 . . . 2.046 . . . . . 9.02
>
> SET 2. ENTER >=+2 THEN BET IN ALL NONNEGATIVE ROUNDS
>
> . Bet . . . .Kelly . . . . . . . . . . . . . . . . Win/100
> Spread. . . . Unit . . . SCORE . . . IBA (%) . . . (units)
> ----------------------------------------------------------
> Flat 1 box. 81.90 . .. . 42.70 . . . 1.046 . . . . . 0.52
> Flat 2 box. 59.52 . .. . 62.10 . . . 1.046 . . . . . 1.04
> 2:1 . . . . 86.21 . .. . 90.95 . . . 1.390 . . . . . 1.06
> 4:1 . . . . 71.68 . . . 122.56 . . . 1.752 . . . . . 1.71
> 6:1 . . . . 56.24 . . . 134.17 . . . 1.927 . . . . . 2.39
> 8:1 . . . . 47.98 . . . 140.25 . . . 1.969 . . . . . 2.92
> 12:1. . . . 39.04 . . . 146.19 . . . 2.054 . . . . . 3.74
> 16:1. . . . 31.82 . . . 150.32 . . . 2.152 . . . . . 4.72
> 20:1. . . . 26.14 . . . 150.89 . . . 2.236 . . . . . 5.77
> 40:1. . . . 14.80 . . . 154.53 . . . 2.314 . . . .. 10.44
>
>
>
>
>
>
> --- In [email protected], leptokurtotic
> <no_reply@y...> wrote:
> >
> > What's the % edge Ted, please?.
> >
> > --- In [email protected], ted_forrester
> > <no_reply@y...> wrote:
> > >
> > > I agree with John that CSMs are beatable, including all
current
> > > models that I know of. I would not want to post specifics
about
> > such
> > > machines, because the techniques are in some cases
> > counterintuitive.
> > > This hopefully makes the techniques less transparent to
> > > surveillance, at least until they become more widely known.
> > >
> > > If people stopped to think about it, they would probably not
be
> > > surprised by the fact that CSMs remain beatable. It is not all
> > that
> > > easy to design a machine that performs an instantaneous and
> truly
> > > random 'shuffle' of multiple decks of cards, and yet does it
in
> a
> > > manner that is not overly prone to jams or destruction of the
> > cards.
> > >
> > > It may seem strange, but the machines that provide the
greatest
> > > challenge when it comes to finding flaws through shuffle
> analysis
> > > are usually not CSMs, but rather the noncontinuous machines.
> Some
> > of
> > > these machines are absolutely fascinating to examine. In
> contrast,
> > > CSM vulnerabilities are usually relatively quick to find, and
> > > usually less intriguing from an intellectual viewpoint (not
that
> > > this is a consideration from the perspective of advantage
> play!).
> > > Noncontinuous machines have much more time to perform their
> > > shuffles, which gives them an advantage when it comes to
foiling
> > > shuffle-based techniques. That they are exploitable at all is
> more
> > a
> > > testament to the difficulties involved in achieving a truly
> random
> > > shuffle than any deficiency of the machines themselves.
> > >
> > > If I had to choose between a 4/6 hand-shuffled shoe game, a
CSM
> > > game, and a 4/6 ASM game, all with the same rules, I would
> > probably
> > > opt first for the CSM, then the hand shuffle, and lastly the
> ASM.
> > > Special circumstances, such as a single-pass, one-riffle R&R
> would
> > > elevate the hand-shuffled game to first choice, but under most
> > > current, real-world conditions, the hand-shuffled game would
> come
> > in
> > > second to the CSM game. CSM games have the added benefit that
> > their
> > > vulnerabilities appear not to be widely understood, on either
> side
> > > of the tables. This reduces heat, and also reduces the
> > > contestability of the terrain (i.e. there is less competition
> for
> > > available space from other APs, especially wongers).
> > >
> > > Rather than posting any specific information about CSMs, I
will
> > > offer the following generic advice to anyone interested in
> trying
> > to
> > > identify effective techniques:
> > >
> > > 1. The first step is obviously to read the relevant patent,
and
> > any
> > > earlier patents that may be related to it. The patent,
> especially
> > > the diagrams, will usually give you a good idea of the
machine's
> > > basic procedure. However...
> > >
> > > 2. On certain key points of special relevance to tracking and
> > > sequencing, the patent will usually be hazy. On these points
it
> is
> > > necessary to propose some hypotheses about what the machine
> might
> > > do. Visual inspection of diagrams provided with the patent can
> be
> > > used to narrow the range of possible hypotheses. For example,
> what
> > > is the nature of the mechanisms that insert, sort, grab or
> select
> > > cards? The nature of the mechanisms may suggest certain things
> > about
> > > where cards are inserted, how they are sorted, from where
cards
> > are
> > > grabbed and which cards are selected. A presumption that the
> > > technically simplest machine procedure is likely to be adopted
> is
> > > also sometimes warranted.
> > >
> > > 3. Once you have devised one or more hypotheses, and
constructed
> > one
> > > or more working profiles of what the machine might do, examine
> the
> > > results and design some simple little tests of your hypotheses
> > that
> > > you can carry out at the tables. It is difficult to provide an
> > > example of such a test without giving up some information
about
> > the
> > > way in which machines can be vulnerable, but suffice to say
that
> a
> > > test would usually involve memorising one or more discards and
> > then
> > > observing later rounds to see if the appearance or
nonappearance
> > of
> > > the card(s) and its/their relation with other memorised cards
> (if
> > > any) is in accordance with your hypothesis.
> > >
> > > 4. At this point you are ready to test your hypotheses in the
> > > casino. If your hypothesis is quite specific, but
unfortunately
> > > turns out to be incorrect (which is quite likely, especially
in
> > the
> > > preliminary stages of your investigations), it should not take
> too
> > > long to realise it. If early evidence supports your
hypothesis,
> > you
> > > should be cautious and test it until you have reason to be
very
> > > confident of its validity.
> > >
> > >
> > > --- In [email protected], sys381
> > > <no_reply@y...> wrote:
> > > >
> > > > --- In [email protected],
> popcornlover01
> > > > <no_reply@y...> wrote:
> > > > >
> > > > > > After checking out the official webpage
> > > > (http://www.borisbj21.com/) I
> > > > > > would have to agree with John. This guy is full of bad
> > > > information
> > > > > > and his advice is worthless. There is no modified basic
> > > strategy
> > > > or
> > > > > > betting progression that will give you the advantage
> against
> > > the
> > > > CSMs.
> > > > > >
> > > > > > I don't know how accurate his software is, but Casino
> Verite
> > > from
> > > > > > QFit will do all the same things (and MUCH, MUCH more)
for
> > $60
> > > > > > cheaper. Also, Casino Verite is the industry standard
> > > simulator
> > > > and
> > > > > > shuffle analysis software. Card counters do not like to
> > > gamble,
> > > > they
> > > > > > like to know what to expect. I don't know why any AP
would
> > > take
> > > > their
> > > > > > chances with the Boris software.
> > > > >
> > > > > If many casinos start using the continous shuffle
machines,
> is
> > > > there any hope for the
> > > > > advantage player?
> > > >
> > > >
> > > > Yes. There are ways to beat the things.
> > > > I published a few techniques in the book you can see above
> > there.
> > > > There are others. After the book was published a lot of
places
> > > > started dumping the devices and they have never gained the
> > > industry
> > > > domination the shufflemachine companies hoped for.
--- In [email protected], ted_forrester
<no_reply@y...> wrote:
>
> I don't really want to go into details on this question. What I
will
> do in this post is present some representative SCOREs and related
> data for a widely used, current model CSM under the assumption
that
> only count-type (or compositional) information is considered by
the
> player. This means that the following results completely ignore
> available gains from sequencing, boundary effects, side bets, and
> other methods for acquiring hole-card or next-card information. I
> will not specify the machine for which these specific SCOREs have
> been derived (approximated, really), but I will say that it is a
new
> model, and that comparable results could have been presented for
the
> other state-of-the-art models currently employed in casinos. The
> results may leave some readers scratching their heads, especially
if
> they persist in thinking in terms of latency. The vulnerabilities
> come from specific knowledge of how the machine `shuffles', and
not
> necessarily from any sluggishness of cards to re-enter play upon
> reinsertion into the machine.
>
> Two sets of results are presented. In the first set, it is assumed
> that the player places wagers during all nonnegative true counts,
> but sits out or leaves at any negative true count. In the second
> set, the player takes a similar approach once s/he has joined a
> table, but only enters on true counts of +2 or greater. Except
when
> otherwise indicated, the player is assumed to play 2 boxes
whenever
> the optimal bet is at least 2 kelly units. By convention, it is
> assumed that the player observes 100 rounds per hour and has a
> $10000 bankroll.
>
> SET 1. BET IN ALL NONNEGATIVE ROUNDS
>
> . Bet . . . .Kelly . . . . . . . . . . . . . . . . Win/100
> Spread. . . . Unit . . . SCORE . . . IBA (%) . . . (units)
> ----------------------------------------------------------
> Flat 1 box. 41.10 . .. . 12.46 . . . 0.524 . . . . . 0.30
> Flat 2 box. 29.94 . .. . 18.14 . . . 0.524 . . . . . 0.61
> 2:1 . . . . 59.35 . .. . 42.54 . . . 0.911 . . . . . 0.72
> 4:1 . . . . 50.51 . .. . 72.19 . . . 1.252 . . . . . 1.43
> 6:1 . . . . 45.66 . .. . 84.97 . . . 1.531 . . . . . 1.86
> 8:1 . . . . 36.10 . .. . 93.27 . . . 1.566 . . . . . 2.58
> 12:1. . . . 31.15 . . . 101.99 . . . 1.758 . . . . . 3.27
> 16:1. . . . 25.91 . . . 105.28 . . . 1.882 . . . . . 4.06
> 20:1. . . . 21.98 . . . 107.98 . . . 1.952 . . . . . 4.91
> 40:1. . . . 12.58 . . . 113.50 . . . 2.046 . . . . . 9.02
>
> SET 2. ENTER >=+2 THEN BET IN ALL NONNEGATIVE ROUNDS
>
> . Bet . . . .Kelly . . . . . . . . . . . . . . . . Win/100
> Spread. . . . Unit . . . SCORE . . . IBA (%) . . . (units)
> ----------------------------------------------------------
> Flat 1 box. 81.90 . .. . 42.70 . . . 1.046 . . . . . 0.52
> Flat 2 box. 59.52 . .. . 62.10 . . . 1.046 . . . . . 1.04
> 2:1 . . . . 86.21 . .. . 90.95 . . . 1.390 . . . . . 1.06
> 4:1 . . . . 71.68 . . . 122.56 . . . 1.752 . . . . . 1.71
> 6:1 . . . . 56.24 . . . 134.17 . . . 1.927 . . . . . 2.39
> 8:1 . . . . 47.98 . . . 140.25 . . . 1.969 . . . . . 2.92
> 12:1. . . . 39.04 . . . 146.19 . . . 2.054 . . . . . 3.74
> 16:1. . . . 31.82 . . . 150.32 . . . 2.152 . . . . . 4.72
> 20:1. . . . 26.14 . . . 150.89 . . . 2.236 . . . . . 5.77
> 40:1. . . . 14.80 . . . 154.53 . . . 2.314 . . . .. 10.44
>
>
>
>
>
>
> --- In [email protected], leptokurtotic
> <no_reply@y...> wrote:
> >
> > What's the % edge Ted, please?.
> >
> > --- In [email protected], ted_forrester
> > <no_reply@y...> wrote:
> > >
> > > I agree with John that CSMs are beatable, including all
current
> > > models that I know of. I would not want to post specifics
about
> > such
> > > machines, because the techniques are in some cases
> > counterintuitive.
> > > This hopefully makes the techniques less transparent to
> > > surveillance, at least until they become more widely known.
> > >
> > > If people stopped to think about it, they would probably not
be
> > > surprised by the fact that CSMs remain beatable. It is not all
> > that
> > > easy to design a machine that performs an instantaneous and
> truly
> > > random 'shuffle' of multiple decks of cards, and yet does it
in
> a
> > > manner that is not overly prone to jams or destruction of the
> > cards.
> > >
> > > It may seem strange, but the machines that provide the
greatest
> > > challenge when it comes to finding flaws through shuffle
> analysis
> > > are usually not CSMs, but rather the noncontinuous machines.
> Some
> > of
> > > these machines are absolutely fascinating to examine. In
> contrast,
> > > CSM vulnerabilities are usually relatively quick to find, and
> > > usually less intriguing from an intellectual viewpoint (not
that
> > > this is a consideration from the perspective of advantage
> play!).
> > > Noncontinuous machines have much more time to perform their
> > > shuffles, which gives them an advantage when it comes to
foiling
> > > shuffle-based techniques. That they are exploitable at all is
> more
> > a
> > > testament to the difficulties involved in achieving a truly
> random
> > > shuffle than any deficiency of the machines themselves.
> > >
> > > If I had to choose between a 4/6 hand-shuffled shoe game, a
CSM
> > > game, and a 4/6 ASM game, all with the same rules, I would
> > probably
> > > opt first for the CSM, then the hand shuffle, and lastly the
> ASM.
> > > Special circumstances, such as a single-pass, one-riffle R&R
> would
> > > elevate the hand-shuffled game to first choice, but under most
> > > current, real-world conditions, the hand-shuffled game would
> come
> > in
> > > second to the CSM game. CSM games have the added benefit that
> > their
> > > vulnerabilities appear not to be widely understood, on either
> side
> > > of the tables. This reduces heat, and also reduces the
> > > contestability of the terrain (i.e. there is less competition
> for
> > > available space from other APs, especially wongers).
> > >
> > > Rather than posting any specific information about CSMs, I
will
> > > offer the following generic advice to anyone interested in
> trying
> > to
> > > identify effective techniques:
> > >
> > > 1. The first step is obviously to read the relevant patent,
and
> > any
> > > earlier patents that may be related to it. The patent,
> especially
> > > the diagrams, will usually give you a good idea of the
machine's
> > > basic procedure. However...
> > >
> > > 2. On certain key points of special relevance to tracking and
> > > sequencing, the patent will usually be hazy. On these points
it
> is
> > > necessary to propose some hypotheses about what the machine
> might
> > > do. Visual inspection of diagrams provided with the patent can
> be
> > > used to narrow the range of possible hypotheses. For example,
> what
> > > is the nature of the mechanisms that insert, sort, grab or
> select
> > > cards? The nature of the mechanisms may suggest certain things
> > about
> > > where cards are inserted, how they are sorted, from where
cards
> > are
> > > grabbed and which cards are selected. A presumption that the
> > > technically simplest machine procedure is likely to be adopted
> is
> > > also sometimes warranted.
> > >
> > > 3. Once you have devised one or more hypotheses, and
constructed
> > one
> > > or more working profiles of what the machine might do, examine
> the
> > > results and design some simple little tests of your hypotheses
> > that
> > > you can carry out at the tables. It is difficult to provide an
> > > example of such a test without giving up some information
about
> > the
> > > way in which machines can be vulnerable, but suffice to say
that
> a
> > > test would usually involve memorising one or more discards and
> > then
> > > observing later rounds to see if the appearance or
nonappearance
> > of
> > > the card(s) and its/their relation with other memorised cards
> (if
> > > any) is in accordance with your hypothesis.
> > >
> > > 4. At this point you are ready to test your hypotheses in the
> > > casino. If your hypothesis is quite specific, but
unfortunately
> > > turns out to be incorrect (which is quite likely, especially
in
> > the
> > > preliminary stages of your investigations), it should not take
> too
> > > long to realise it. If early evidence supports your
hypothesis,
> > you
> > > should be cautious and test it until you have reason to be
very
> > > confident of its validity.
> > >
> > >
> > > --- In [email protected], sys381
> > > <no_reply@y...> wrote:
> > > >
> > > > --- In [email protected],
> popcornlover01
> > > > <no_reply@y...> wrote:
> > > > >
> > > > > > After checking out the official webpage
> > > > (http://www.borisbj21.com/) I
> > > > > > would have to agree with John. This guy is full of bad
> > > > information
> > > > > > and his advice is worthless. There is no modified basic
> > > strategy
> > > > or
> > > > > > betting progression that will give you the advantage
> against
> > > the
> > > > CSMs.
> > > > > >
> > > > > > I don't know how accurate his software is, but Casino
> Verite
> > > from
> > > > > > QFit will do all the same things (and MUCH, MUCH more)
for
> > $60
> > > > > > cheaper. Also, Casino Verite is the industry standard
> > > simulator
> > > > and
> > > > > > shuffle analysis software. Card counters do not like to
> > > gamble,
> > > > they
> > > > > > like to know what to expect. I don't know why any AP
would
> > > take
> > > > their
> > > > > > chances with the Boris software.
> > > > >
> > > > > If many casinos start using the continous shuffle
machines,
> is
> > > > there any hope for the
> > > > > advantage player?
> > > >
> > > >
> > > > Yes. There are ways to beat the things.
> > > > I published a few techniques in the book you can see above
> > there.
> > > > There are others. After the book was published a lot of
places
> > > > started dumping the devices and they have never gained the
> > > industry
> > > > domination the shufflemachine companies hoped for.
] used to disappear for a couple or three months at a time, but he's been gone so long this time. Master of the run-on sentence, and perpetrator of outlandish bullshit, he still was quite a theorist.