vpFREE2 Forums

Wynn Update from Rob Singer

#21

>
> My computations are based on a kind of Bayesian-like approach

using

> exact (or nearly exact) PDFs and finite hand RoR curves.

I'm not sure I see how you would apply Bayesian conditional
probabilities since all the events are independent, at least if the
rng's are working properly. The only thing that is conditional, is
any strategy change or denomenation change that a player chooses to
make based on their wins or loses.

the RS strategy boils down to assuming that P(LA;SA),
> the
> probability of losing some LARGE amount (LA) before wining at

least

a
> small amount (SA) is small enough that P(LA;SA) << SA/LA.

Hence,

> one
> could assume (in the long run) that someone following this

strategy

> expects to win the SA many times (in separate successful

sessions)

> before losing the LA once, at least most of the time.
>
> Is this a reasonable way to approach a mathematical description

of

RS
> strategy? Comments?
>

For this to be a successful strategy the sum of those many SA wins
would have to be greater than the big LA loss.

The bottom line is, if you are playing negative expectation games you
will not be a long term winner, unless you are extremely lucky. No
betting modification system can counter the inherent odds of the
chosen game. You could try the system of doubling every losing bet
until you win, but that system only works with an infinite amount of
money and a casino with no betting limits.

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

#22

Why do casinos let progression spewing salesmen have
seminars on their property and why people buy them
foolishly.

One thing that casinos learned early on is as long as
they are offering a negative expectation game and they
put limits on their tables and games, ABSOLUTELY NO
PROGRESSION SYSTEM CAN WIN LONG TERM! It is pure and
simple math, so casino execs will gladly rent space to
someone selling a progression. They get a fee and a
house full of players who will follow a losing system.
Win/ win situation for both the casino and the
progression seller.

Unfortunately, many people just do not grasp the math
or feel that the bad streak of consecutive loses will
never hit them and they will win. The fact that these
systems will offer many small winning sessions before
devastation sets in makes it appealing because the
small wins are the only things constantly mentioned
and the impending doom is left up to someone who
understands the math to explain.

On a personal level, I can document two streaks in the
past four years on a blackjack table. First was 16
straight losing hands and the second was either 19 or
20 straight loses (not sure). Would any
progressionist believer please tell me just how
bankrupt I would have become if I had been playing a
progression chasing my loses? Of course the table
limits would have stopped the progression at some
point so getting the money back on one hand would have
been impossible. Since the vast majority of VP is
negative expectation, you can just from 25 cent
machines up to $100 machines if you have the bankroll
but if it continues badly at some point you have no
money or there is no machine with a high enough
betting level.

Victoria

···

--- thymos_one <thymos_one@yahoo.com> wrote:

>
> --- In vpFREE@yahoogroups.com, "cdfsrule"
<groups.yahoo@v...> wrote:
> >
> > My computations are based on a kind of
Bayesian-like approach
using
> > exact (or nearly exact) PDFs and finite hand RoR
curves.

I'm not sure I see how you would apply Bayesian
conditional
probabilities since all the events are independent,
at least if the
rng's are working properly. The only thing that is
conditional, is
any strategy change or denomenation change that a
player chooses to
make based on their wins or loses.

> the RS strategy boils down to assuming that
P(LA;SA),
> > the
> > probability of losing some LARGE amount (LA)
before wining at
least
> a
> > small amount (SA) is small enough that P(LA;SA)
<< SA/LA.
Hence,
> > one
> > could assume (in the long run) that someone
following this
strategy
> > expects to win the SA many times (in separate
successful
sessions)
> > before losing the LA once, at least most of the
time.
> >
> > Is this a reasonable way to approach a
mathematical description
of
> RS
> > strategy? Comments?
> >

For this to be a successful strategy the sum of
those many SA wins
would have to be greater than the big LA loss.

The bottom line is, if you are playing negative
expectation games you
will not be a long term winner, unless you are
extremely lucky. No
betting modification system can counter the inherent
odds of the
chosen game. You could try the system of doubling
every losing bet
until you win, but that system only works with an
infinite amount of
money and a casino with no betting limits.

test'; ">

#23

Why do casinos let progression spewing salesmen have
seminars on their property and why people buy them
foolishly.

I believe you're somewhat mixed up. Those spewing salesmen (& women)
who give seminars on casino properties are in fact those who are
advocates of advantage play. Mr. Singer has to my knowledge tried to
hold classes teaching his short term play methodology, but the
casinos weren't ready for that. There also seems to be a lot of
misconceptions about his play strategy. It is "progression in
denomination and volatility" (as taken from his site) but it
definitely is not Martingale. For people to have a coherent
discussion here on it, I suggest they read and understand his
philosophy and approach first.

···

--- In vpFREE@yahoogroups.com, Victoria Rosado <ros4144@y...> wrote:

One thing that casinos learned early on is as long as
they are offering a negative expectation game and they
put limits on their tables and games, ABSOLUTELY NO
PROGRESSION SYSTEM CAN WIN LONG TERM! It is pure and
simple math, so casino execs will gladly rent space to
someone selling a progression. They get a fee and a
house full of players who will follow a losing system.
Win/ win situation for both the casino and the
progression seller.

Unfortunately, many people just do not grasp the math
or feel that the bad streak of consecutive loses will
never hit them and they will win. The fact that these
systems will offer many small winning sessions before
devastation sets in makes it appealing because the
small wins are the only things constantly mentioned
and the impending doom is left up to someone who
understands the math to explain.

On a personal level, I can document two streaks in the
past four years on a blackjack table. First was 16
straight losing hands and the second was either 19 or
20 straight loses (not sure). Would any
progressionist believer please tell me just how
bankrupt I would have become if I had been playing a
progression chasing my loses? Of course the table
limits would have stopped the progression at some
point so getting the money back on one hand would have
been impossible. Since the vast majority of VP is
negative expectation, you can just from 25 cent
machines up to $100 machines if you have the bankroll
but if it continues badly at some point you have no
money or there is no machine with a high enough
betting level.

Victoria

#24

Go to a professional blackjack site: bj21.com or
advantageplayer.com or cardcounter.com and ask those
there about Mr Singer.
Victoria

···

--- aces_hii <aces_hii@yahoo.com> wrote:

--- In vpFREE@yahoogroups.com, Victoria Rosado
<ros4144@y...> wrote:
> Why do casinos let progression spewing salesmen
have
> seminars on their property and why people buy them
> foolishly.

I believe you're somewhat mixed up. Those spewing
salesmen (& women)
who give seminars on casino properties are in fact
those who are
advocates of advantage play. Mr. Singer has to my
knowledge tried to
hold classes teaching his short term play
methodology, but the
casinos weren't ready for that. There also seems to
be a lot of
misconceptions about his play strategy. It is
"progression in
denomination and volatility" (as taken from his
site) but it
definitely is not Martingale. For people to have a
coherent
discussion here on it, I suggest they read and
understand his
philosophy and approach first.

> One thing that casinos learned early on is as long
as
> they are offering a negative expectation game and
they
> put limits on their tables and games, ABSOLUTELY
NO
> PROGRESSION SYSTEM CAN WIN LONG TERM! It is pure
and
> simple math, so casino execs will gladly rent
space to
> someone selling a progression. They get a fee and
a
> house full of players who will follow a losing
system.
> Win/ win situation for both the casino and the
> progression seller.
>
> Unfortunately, many people just do not grasp the
math
> or feel that the bad streak of consecutive loses
will
> never hit them and they will win. The fact that
these
> systems will offer many small winning sessions
before
> devastation sets in makes it appealing because the
> small wins are the only things constantly
mentioned
> and the impending doom is left up to someone who
> understands the math to explain.
>
> On a personal level, I can document two streaks in
the
> past four years on a blackjack table. First was
16
> straight losing hands and the second was either 19
or
> 20 straight loses (not sure). Would any
> progressionist believer please tell me just how
> bankrupt I would have become if I had been playing
a
> progression chasing my loses? Of course the table
> limits would have stopped the progression at some
> point so getting the money back on one hand would
have
> been impossible. Since the vast majority of VP is
> negative expectation, you can just from 25 cent
> machines up to $100 machines if you have the
bankroll
> but if it continues badly at some point you have
no
> money or there is no machine with a high enough
> betting level.
>
> Victoria

test'; ">

#25

note that 'aces_hii' and Rob Singer's posts (on freevpfree) are coming
from the same computer.

click on the message, click 'view source', the 'X-Yahoo-Post-IP' field
identifies the computer.

the old admin chose the right time to retire I think.

···

On 9/10/05, aces_hii <aces_hii@yahoo.com> wrote:

--- In vpFREE@yahoogroups.com, Victoria Rosado <ros4144@y...> wrote:
> Why do casinos let progression spewing salesmen have
> seminars on their property and why people buy them
> foolishly.

I believe you're somewhat mixed up. Those spewing salesmen (& women)
who give seminars on casino properties are in fact those who are
advocates of advantage play. Mr. Singer has to my knowledge tried to
hold classes teaching his short term play methodology, but the
casinos weren't ready for that. There also seems to be a lot of
misconceptions about his play strategy. It is "progression in
denomination and volatility" (as taken from his site) but it
definitely is not Martingale. For people to have a coherent
discussion here on it, I suggest they read and understand his
philosophy and approach first.

> One thing that casinos learned early on is as long as
> they are offering a negative expectation game and they
> put limits on their tables and games, ABSOLUTELY NO
> PROGRESSION SYSTEM CAN WIN LONG TERM! It is pure and
> simple math, so casino execs will gladly rent space to
> someone selling a progression. They get a fee and a
> house full of players who will follow a losing system.
> Win/ win situation for both the casino and the
> progression seller.
>
> Unfortunately, many people just do not grasp the math
> or feel that the bad streak of consecutive loses will
> never hit them and they will win. The fact that these
> systems will offer many small winning sessions before
> devastation sets in makes it appealing because the
> small wins are the only things constantly mentioned
> and the impending doom is left up to someone who
> understands the math to explain.
>
> On a personal level, I can document two streaks in the
> past four years on a blackjack table. First was 16
> straight losing hands and the second was either 19 or
> 20 straight loses (not sure). Would any
> progressionist believer please tell me just how
> bankrupt I would have become if I had been playing a
> progression chasing my loses? Of course the table
> limits would have stopped the progression at some
> point so getting the money back on one hand would have
> been impossible. Since the vast majority of VP is
> negative expectation, you can just from 25 cent
> machines up to $100 machines if you have the bankroll
> but if it continues badly at some point you have no
> money or there is no machine with a high enough
> betting level.
>
> Victoria

------------------------

#26

Let me outline the computation for you. Then, if you don't think it is
a Bayesian-like approach, please supply another name. [BTW, I don't
think Bayesian statistics requires independent events; if the events
are independent however, the statistics are easier to compute, since
you don't need to know the joint distributions, but rather only the
independent distributions of the 2 events]

1) Compute the PDF for the fist hand of VP: PDF(1) = Starting State X
PDF (game), where X stands for convolution, and the starting state has
a probiliaty of 1 at 0 (no win or loss)-- that is a delta function.

2) Adjust the PDF(1) to take into account the loss limit and the win
limit. Call this PDF(1)'. To do this, set the PDF(1) <= Loss limit to
0 and PDF (1)>= Win Limit to zero. Also store the Total Probability <=
Loss Limit {I will call this the RoR for the 1st hand} and store the
total prob >= win limit (this is the Prob. of Success, I call it PoS
for 1 hand)

3) Now compute the PDF for the second hand, PDF(2)= PDF(1)' X PDF(game).
4) Next adjust PDF(2), computing PDF(2)' as above in step(2). This
yeilds the RoR & PoS for 2 hands.
5) Continue this procedure until: you reach some large number of hands,
the total Prob of PDF(n)' becomes 0, or you get sick of it.

So, in the end, one gets RoR(n), PoS(n), PDF(n)' for n=1 to a big
number. Now note that for each RoR(j), PoS(j) or PDF(j)' , with j>=2,
depends on PDF(j-1)' , etc. Hence, each PDF(j)' is dependant on all
prior PDFs. [BTW,If you know another way to compute the PDFs, or a
closed form expression for the convolution, please, please, let me
know!]
So, the quantities we are interested in here, are, I think, conditional
probabilities of a sort: That is, when we ask, what is the probability
of not losing all our money by hand 10000, we are really asking, what
is the prob. of not losing all our money by hand 10000 , given a
certain distribution for hand 9999 (which depends conditionally on hand
9998, and so on).

···

--- In vpFREE@yahoogroups.com, "thymos_one" <thymos_one@y...> wrote:

I'm not sure I see how you would apply Bayesian conditional
probabilities since all the events are independent, at least if the
rng's are working properly. The only thing that is conditional, is
any strategy change or denomenation change that a player chooses to
make based on their wins or loses.

#27

Let me outline the computation for you. Then, if you don't think

it is

a Bayesian-like approach, please supply another name. [BTW, I don't
think Bayesian statistics requires independent events; if the

events

are independent however, the statistics are easier to compute,

since

you don't need to know the joint distributions, but rather only the
independent distributions of the 2 events]

My understanding of Bayesian analysis is that it requires events to
be dependent, if they are independent there is no conditional
probability. If you have two events A and B which are dependent then
given that event A has occurred you can modify your estimation that
event B has occurred by applying Bayes theorem.

I don't see how your recursive formula below implies statistical
dependence. If I'm playing a 99% ev game with a goal of winning $100
and a stop loss at $1,000 if after 1000 hands I am down (500) what
does that tell me about the probability or ev of the 1001 hand,
nothing more than what I knew when I started playing, the ev is still
99%, the pdf hasn't changed.

1) Compute the PDF for the fist hand of VP: PDF(1) = Starting

State X

PDF (game), where X stands for convolution, and the starting state

has

a probiliaty of 1 at 0 (no win or loss)-- that is a delta function.

2) Adjust the PDF(1) to take into account the loss limit and the

win

limit. Call this PDF(1)'. To do this, set the PDF(1) <= Loss limit

to

0 and PDF (1)>= Win Limit to zero. Also store the Total Probability

<=

Loss Limit {I will call this the RoR for the 1st hand} and store

the

total prob >= win limit (this is the Prob. of Success, I call it

PoS

for 1 hand)

3) Now compute the PDF for the second hand, PDF(2)= PDF(1)' X PDF

(game).

4) Next adjust PDF(2), computing PDF(2)' as above in step(2). This
yeilds the RoR & PoS for 2 hands.
5) Continue this procedure until: you reach some large number of

hands,

the total Prob of PDF(n)' becomes 0, or you get sick of it.

So, in the end, one gets RoR(n), PoS(n), PDF(n)' for n=1 to a big
number. Now note that for each RoR(j), PoS(j) or PDF(j)' , with
=2,
depends on PDF(j-1)' , etc. Hence, each PDF(j)' is dependant on

all

prior PDFs. [BTW,If you know another way to compute the PDFs, or a
closed form expression for the convolution, please, please, let me
know!]
So, the quantities we are interested in here, are, I think,

conditional

probabilities of a sort: That is, when we ask, what is the

probability

of not losing all our money by hand 10000, we are really asking,

what

is the prob. of not losing all our money by hand 10000 , given a
certain distribution for hand 9999 (which depends conditionally on

hand

9998, and so on).

>
> I'm not sure I see how you would apply Bayesian conditional
> probabilities since all the events are independent, at least if

the

> rng's are working properly. The only thing that is conditional,

is

> any strategy change or denomenation change that a player chooses

to

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

--- In vpFREE@yahoogroups.com, "thymos_one" <thymos_one@y...> wrote:
> make based on their wins or loses.
>

#28

I am really not an expert on Bayesian Analysis (BA), and I never
claimed a strict BA approach, but rather only a Bayesian-like
method. I wasn't hedging, but rather being respectful to the fact
that BA is not really a single well defined method, per se, but
rather a collection of related statistical methods. Indeed "There are
many varieties of Bayesian analysis." (see
http://www.bayesian.org/bayesexp/bayesexp.htm )

I don't see how your recursive formula below implies statistical
dependence. If I'm playing a 99% ev game with a goal of winning

$100 > and a stop loss at $1,000 if after 1000 hands I am down (500)
what >does that tell me about the probability or ev of the 1001 hand,

nothing more than what I knew when I started playing, the ev is

still >99%, the pdf hasn't changed.

Well, I'm not quite sure what you objecting to. We all know that the
PDF for any particular hand hasn't changed, and in my computation, I
use the same PDF(Game) throughout. That said, the PDF'(hand n)
depends heavily of the PDF's of all "prior" hands. As you know, In BA
is said to involve what is called the "prior" distribution (which is
a usually subjective thing, btw). What is the prior distribution for
VP in my computation? Well, imagine that you send out a lot of
people to play the same VP game with the same strategy and tactics
(stop loss, stop win, etc). I could ask the question, "what is the
likelihood of a particular VP player to be still playing after 1000
hands?". Well that question is conditional on the actual state of
the VP player before hand 1000 (that is after hand 999). For example,
after hand 999, the player may be teetering on the edge of the stop
loss or stop win, or solidly (safely) away from those critical
points. So the answer to the question DEPENDS on the seemingly
arbitrary state after hand 999. But we can handle this question in
statistical terms. In VP, this state is perhaps best described as a
PDF, and this prior PDF of sorts isn't actually subjective for VP, so
long as we can assume that the player started in a known state and
played by the rules we set forth. So what we are actually computing
is not the probability of a particular player (which would require an
arbitrary and subjective "prior" assumption) but rather the PDF for
all players (any player), which seems to me to be nicely described by
a non-arbitrary (non subjective) prior PDF. In other words, I'd
rather not get into a discussion about a priori -vs- a posterior
statistics (since I would muck everything up), though I think in your
question to me (where you assumed an outcome, namely being down 500)
you presented an a posteriori treatment, which, as you point out,
doesn't utilize Bayesian statistics, and isn't relevant to my
computation (since I don't know that you or anyone else was actually
down 500)

#29

I think we are converging on agreement or at least being on the same
page, so to speak. From a personal point of view I don't think the
effort you propose is worth the outcome since we know apriori that
Singer's strategies are not worth using. You can't change a negative
expectation game with betting strategies, and that would be a real
benefit.

I am really not an expert on Bayesian Analysis (BA), and I never
claimed a strict BA approach, but rather only a Bayesian-like
method. I wasn't hedging, but rather being respectful to the fact
that BA is not really a single well defined method, per se, but
rather a collection of related statistical methods. Indeed "There

are

many varieties of Bayesian analysis." (see
http://www.bayesian.org/bayesexp/bayesexp.htm )

> I don't see how your recursive formula below implies statistical
> dependence. If I'm playing a 99% ev game with a goal of winning
$100 > and a stop loss at $1,000 if after 1000 hands I am down

(500)

what >does that tell me about the probability or ev of the 1001

hand,

>nothing more than what I knew when I started playing, the ev is
still >99%, the pdf hasn't changed.

Well, I'm not quite sure what you objecting to. We all know that

the

PDF for any particular hand hasn't changed, and in my computation,

I

use the same PDF(Game) throughout. That said, the PDF'(hand n)
depends heavily of the PDF's of all "prior" hands. As you know, In

BA

is said to involve what is called the "prior" distribution (which

is

a usually subjective thing, btw). What is the prior distribution

for

VP in my computation? Well, imagine that you send out a lot of
people to play the same VP game with the same strategy and tactics
(stop loss, stop win, etc). I could ask the question, "what is the
likelihood of a particular VP player to be still playing after 1000
hands?". Well that question is conditional on the actual state of
the VP player before hand 1000 (that is after hand 999). For

example,

after hand 999, the player may be teetering on the edge of the stop
loss or stop win, or solidly (safely) away from those critical
points. So the answer to the question DEPENDS on the seemingly
arbitrary state after hand 999. But we can handle this question in
statistical terms. In VP, this state is perhaps best described as a
PDF, and this prior PDF of sorts isn't actually subjective for VP,

so

long as we can assume that the player started in a known state and
played by the rules we set forth. So what we are actually computing
is not the probability of a particular player (which would require

an

arbitrary and subjective "prior" assumption) but rather the PDF for
all players (any player), which seems to me to be nicely described

by

a non-arbitrary (non subjective) prior PDF. In other words, I'd
rather not get into a discussion about a priori -vs- a posterior
statistics (since I would muck everything up), though I think in

your

question to me (where you assumed an outcome, namely being down

500)

you presented an a posteriori treatment, which, as you point out,
doesn't utilize Bayesian statistics, and isn't relevant to my
computation (since I don't know that you or anyone else was

actually

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

down 500)