random cyclicality in vp

[Dan2]

gilbert wrote (snip):

Oh well, I guess you're a software developer alright since you
disagree that "randomness cannot be programmed".

No, please don't twist my words. I AGREE that randomness cannot be programmed. A computer program cannot generate truly random numbers -- it can generate only pseudo-random numbers.

But that has nothing to do with the real points of this discussion:

1. Are the machines random? They are definitely random enough. The outcome of a play is never predetermined, nor is is ever modified after it is determined that the outcome would be a winner.

2. Are the games predictable? As I said, if they were, I would be making a fortune on them regardless of the payoff tables, as would several other computer experts. The machines are random enough that it is impossible to predict when it will be "hot" or "cold" (these are only observations after the fact) or to predict the draw cards.

3. Do the machines run in "cycles"? No, not predetermined winning/losing cycles. The games are streaky (i.e., they exhibit streaks of wins and losses), but such clumpiness is natural with randomness. Remember, the term "cycle" as used by game analysts is merely the reciprocal of the probability of a royal on any one play, and that probability never changes.

4. Because of the above facts, the only way to be a long-term winner at video poker is to find a situation (combination of payoff table, slot club rebate, promotion and accurate strategy) that offers over 100% expected return, and have a bankroll sufficient to withstand the natural win/loss streaks. The jazbo/Sorokin formula can be used to determine the bankroll necessary for your Risk of Ruin comfort level.

I don't believe that non-random is the same as being relatively
easy to figure out --- or even difficult to figure out but possible for
a master.

But since you apparently do have that belief, you should
probably consider playing at racinos in New York.

Again, you seem to be claiming I said just the opposite of what I really said. IF there were a predetermined algorithm, I believe I could figure it out. But that's a big "if." Since there is no such predetermination, there is nothing that I can analyze beyond what we can determine from the game's payoff table.

But you mention racinos. Are you suggesting betting on the horses? They might be beatable with enough inside information, but not by analyzing any algorithm, because there isn't any.

If this doesn't cut through the fog, there is nothing more worth saying, so this will be my last post on the subject.

Dan

···

--
Dan Paymar
Author of best selling book, "Video Poker - Optimum Play"
Developer of VP analysis/trainer software "Optimum Video Poker"
Visit my web site at www.OptimumPlay.com

"Chance favors the prepared mind." -- Louis Pasteur

[Gilbert_Escobar]

I AGREE that randomness cannot be programmed. A computer program

cannot generate truly random numbers

-- it can generate only pseudo-random numbers.

But that has nothing to do with the real points of this discussion
-------------------------------------------------------------------

gilbert wrote:
That's what you believe in and maybe you're right and maybe you're
wrong.
I wonder how you would explain how the randomness in the vp machines
I saw at Harrah's Rincon somehow significantly change BIG time after
the technicians switch the machines we were playing on (Normal mode)
to this Tournament Mode. We got kicked out from the machines to give
way to their VP Tournament. Quads after quads, Royal Flushes, Full
Houses, were raining like crazy. 3-4 hours later after the
Tournament, the machines were switch back to their Normal mode
(sorry, don't know what they call this mode) and the many quads we
were seeing during the Tournament magically started disappearing?
And the IGT programmers have nothing to do with it? Oh well.....

···

--- In vpFREE@yahoogroups.com, Dan Paymar <Dan@...> wrote:

[rgmustain]

gilbert wrote:
That's what you believe in and maybe you're right and maybe you're
wrong.
I wonder how you would explain how the randomness in the vp machines
I saw at Harrah's Rincon somehow significantly change BIG time after
the technicians switch the machines we were playing on (Normal mode)
to this Tournament Mode.

Tournament mode does not have the same legal requirements since
gamblers are not risking their own money. However, what you've shown is
what could happen without the legal requirement that the machines are
BOTH random and fair.

We got kicked out from the machines to give
way to their VP Tournament. Quads after quads, Royal Flushes, Full
Houses, were raining like crazy. 3-4 hours later after the
Tournament, the machines were switch back to their Normal mode
(sorry, don't know what they call this mode)

Normal is fine. You know, random and fair as required by law.

and the many quads we
were seeing during the Tournament magically started disappearing?
And the IGT programmers have nothing to do with it? Oh well.....

Of course they did. Who else would have does the programming. So, are
you claiming that because you've seen machines set to pay back better
than random/fair when in tournament mode that somehow the machines are
set up to pay back less than random/fair at other times? Are you
claiming that IGT ignores the law and programs VP machines to operate
illegally? If not, then what is your point? The point that random and
equal distribution (fair) are two SEPARATE requirements has already
been mentioned several times.

···

--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@...> wrote:

[Gilbert_Escobar]

Normal is fine. You know, random and fair as required by law.

> and the many quads we
> were seeing during the Tournament magically started

disappearing?

> And the IGT programmers have nothing to do with it? Oh well.....

Of course they did. Who else would have does the programming. So,

are

you claiming that because you've seen machines set to pay back

better

than random/fair when in tournament mode that somehow the machines

are

set up to pay back less than random/fair at other times? Are you
claiming that IGT ignores the law and programs VP machines to

operate

illegally? If not, then what is your point? The point that random

and

equal distribution (fair) are two SEPARATE requirements has already
been mentioned several times.
-----------------------------------------

gilbert wrote:
I don't understand why you keep switching the discussion
of "randomness can or can be programmed" vs. what you always want to
talk about which is "vp machines being legal or illegal". It's not
even close.
I was just trying to express my point of view to Dan P. that if
programmer wants control the randomness of the numbers being
generated by his/her program via the Random Seed value and other
variables, s/he can do it with ease. I'm not saying this is what
they do in the vp machines.
About a year ago, we had a guy working with us we called "409". He
won't take notes, never pays attention as he claims he already know
it, then you ask him the next day and... as clear as it
gets "409" ... everything you showed him the day before is gone.
Is there another one out there?

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:

[2-WILD1]

gilbert wrote:

About a year ago, we had a guy working with us we called "409". He
won't take notes, never pays attention as he claims he already know
it, then you ask him the next day and... as clear as it
gets "409" ... everything you showed him the day before is gone.
Is there another one out there?

You're talking about yourself, right?

Seriously, it's obviously a struggle for you to keep up with the
discussion.

[rgmustain]

gilbert wrote:
I don't understand why you keep switching the discussion
of "randomness can or can be programmed" vs. what you always want

to

talk about which is "vp machines being legal or illegal". It's not
even close.

The question of whether randomness can be programmed is dependent on
your definition of randomness (yep, there are more than one, truly
random, weakly random, strongly random, etc.). RNGs are the basis of,
not only gaming machines, but ALL secure transactions between
computers. So, again I ask, what is your point? Are you trying to say
that non-randomness can also be programmed, guess what? you'll get no
arguments there either.

I was just trying to express my point of view to Dan P. that if
programmer wants control the randomness of the numbers being
generated by his/her program via the Random Seed value and other
variables, s/he can do it with ease.

Quit already with the oxymorons. If it's being controlled, then IT
ISN'T RANDOM. Just because someone uses code called an RNG doesn't
mean they will get randomness. The two are NOT the same. To
get "sufficient and non-breakable randomness" out of computer code
takes more effort.

I'm not saying this is what
they do in the vp machines.

Then I wonder what you are trying to say? You have yet to make any
point other than "programmers can produce code".

About a year ago, we had a guy working with us we called "409". He
won't take notes, never pays attention as he claims he already know
it, then you ask him the next day and... as clear as it
gets "409" ... everything you showed him the day before is gone.
Is there another one out there?

Appears so.

···

--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@...> wrote:

[Tom_Robertson1]

gilbert wrote:
I don't understand why you keep switching the discussion
of "randomness can or can be programmed" vs. what you always want

to

talk about which is "vp machines being legal or illegal". It's not
even close.

The question of whether randomness can be programmed is dependent on
your definition of randomness (yep, there are more than one, truly
random, weakly random, strongly random, etc.). RNGs are the basis of,
not only gaming machines, but ALL secure transactions between
computers. So, again I ask, what is your point? Are you trying to say
that non-randomness can also be programmed, guess what? you'll get no
arguments there either.

But he DID get an argument. Dan wrote: "randomness cannot be
programmed." I don't see the controversy over saying that randomness
is programmed. That's the whole point. Wikipedia has an interesting
article about randomness. In it, it says that during World War 2, the
German code was easy to break because those who sent the messages,
even though they were explicitly asked to make their codes random,
would make their codes fall into patterns, such as their girlfriend's
initials, etc. Randomness isn't natural, so it has to be programmed.
Isn't that what RNGs are designed to do? It doesn't happen by
accident.

I was just trying to express my point of view to Dan P. that if
programmer wants control the randomness of the numbers being
generated by his/her program via the Random Seed value and other
variables, s/he can do it with ease.

Quit already with the oxymorons. If it's being controlled, then IT
ISN'T RANDOM.

Are you sure you two aren't just talking past each other? Randomness
must be programmed, but the actual numbers that an RNG generates
cannot be. Are you sure that isn't what you're both saying?

···

--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@...> wrote:

[Gilbert_Escobar]

--- In vpFREE@yahoogroups.com, Tom Robertson <thomasrrobertson@...>
wrote:

Are you sure you two aren't just talking past each other? Randomness
must be programmed, but the actual numbers that an RNG generates
cannot be. Are you sure that isn't what you're both saying?
------------------------------------------------------------------

gilbert wrote:
Thank you Tom!!!
I'm so glad you put this thing in a very simple way.
I sure missed talking about those "actual numbers" that cannot be
programmed. :<

[rgmustain]

--- In vpFREE@yahoogroups.com, Tom Robertson <thomasrrobertson@...>
wrote:

>> gilbert wrote:
>> I don't understand why you keep switching the discussion
>> of "randomness can or can be programmed" vs. what you always

want

>to
>> talk about which is "vp machines being legal or illegal". It's

not

>> even close.
>
>The question of whether randomness can be programmed is dependent

on

>your definition of randomness (yep, there are more than one, truly
>random, weakly random, strongly random, etc.). RNGs are the basis

of,

>not only gaming machines, but ALL secure transactions between
>computers. So, again I ask, what is your point? Are you trying to

say

>that non-randomness can also be programmed, guess what? you'll get

no

>arguments there either.

But he DID get an argument. Dan wrote: "randomness cannot be
programmed." I don't see the controversy over saying that

randomness

is programmed. That's the whole point. Wikipedia has an

interesting

article about randomness. In it, it says that during World War 2,

the

German code was easy to break because those who sent the messages,
even though they were explicitly asked to make their codes random,
would make their codes fall into patterns, such as their

girlfriend's

initials, etc. Randomness isn't natural, so it has to be

programmed.

Isn't that what RNGs are designed to do? It doesn't happen by
accident.

RNG's are pseudo-random code as you know. If you know the algorithm
and exactly where you are in the algorithm then you can predict
future results. I don't think anyone would consider that random.
Progammming will NEVER be able to produce truly random results.
Programming can produce a sequence of numbers with an even
distribution over time. When combined with reseeding and user
interaction the result is random enough. However, the main problem is
Gilbert keeps inferring that IGT may be controlling the output of
their RNG to produce "control"ed results. He also infers this would
still be random.

>> I was just trying to express my point of view to Dan P. that if
>> programmer wants control the randomness of the numbers being
>> generated by his/her program via the Random Seed value and other
>> variables, s/he can do it with ease.
>
>Quit already with the oxymorons. If it's being controlled, then IT
>ISN'T RANDOM.

Are you sure you two aren't just talking past each other?

Randomness

must be programmed, but the actual numbers that an RNG generates
cannot be. Are you sure that isn't what you're both saying?

I'm sure.

···

>--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@> wrote:

[Tom_Robertson1]

mroejacks wrote:

RNG's are pseudo-random code as you know. If you know the algorithm
and exactly where you are in the algorithm then you can predict
future results. I don't think anyone would consider that random.

Yes, from the perspective of one who knows how to determine the
results of an RNG, those results are not random, but the definition of
randomness must include perception in order to be relevant, since
"true" (or "pure" or "intrinsic") randomness can't exist.

Progammming will NEVER be able to produce truly random results.
Programming can produce a sequence of numbers with an even
distribution over time. When combined with reseeding and user
interaction the result is random enough.

"Random enough" could be rephrased as "completely random, from the
perspective of an observer who is completely ignorant of what results
will be generated."

However, the main problem is
Gilbert keeps inferring that IGT may be controlling the output of
their RNG to produce "control"ed results. He also infers this would
still be random.

I don't see what's wrong with that. There's no empirical way to prove
that IGT has made their machines fair and if they aren't, they can be
just as random as if they were. Gilbert is probably being
misunderstood by using the phrase "changing randomness" when, at least
in the case of the tournament machines at Rincon, he means changing
fairness, I assume.

[Gilbert_Escobar]

--- In vpFREE@yahoogroups.com, Tom Robertson <thomasrrobertson@...>
wrote:

mroejacks wrote:

>RNG's are pseudo-random code as you know. If you know the

algorithm

>and exactly where you are in the algorithm then you can predict
>future results. I don't think anyone would consider that random.

Yes, from the perspective of one who knows how to determine the
results of an RNG, those results are not random, but the definition

of

randomness must include perception in order to be relevant, since
"true" (or "pure" or "intrinsic") randomness can't exist.

>Progammming will NEVER be able to produce truly random results.
>Programming can produce a sequence of numbers with an even
>distribution over time. When combined with reseeding and user
>interaction the result is random enough.

"Random enough" could be rephrased as "completely random, from the
perspective of an observer who is completely ignorant of what

results

will be generated."

>However, the main problem is
>Gilbert keeps inferring that IGT may be controlling the output of
>their RNG to produce "control"ed results. He also infers this

would

>still be random.

I don't see what's wrong with that. There's no empirical way to

prove

that IGT has made their machines fair and if they aren't, they can

be

just as random as if they were. Gilbert is probably being
misunderstood by using the phrase "changing randomness" when, at

least

in the case of the tournament machines at Rincon, he means changing
fairness, I assume.
-------------------------------------------------------

gilbert wrote:
Of course, it's all my fault to be misunderstood -- for being not
smarter than a 5th grader. What's clear to my brain is not really
what's clear to another's brain.
So here's hopefully a sample that may clear things up:
(pretending that I am an IGT programmer, where IGT stands for Indian
Gaming Technology)

(from a software test tool we used)
testcase tcGetMeRandomNumbers () appstate none
  INTEGER iMin, iMax
  iMin = 1
  iMax = 10
  INTEGER iSeed
  iSeed = 2
  RandSeed (iSeed)
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10

Results:
1st run:
1
10
8

2nd run:
1
10
8

Note: same results each time. I didn't tell the computer to give me
the numbers 1, 10, and 8. The RandInt () function got it for me.

Change the iSeed to 3 from 2 and run the same test.

Results:
1st run:
1
3
4

2nd run:
1
3
4

If I change the values for iSeed, iMin and iMax, the "what I refer to
as random" numbers generated will be different again.

NOTE: This is just to demonstrate that I, the programmer, did
control what random numbers to generate each run. No numbers outside
of the iMin and iMax that I specify in my program --- will come out.
This is all what I meant when I refer to "randomness controlled by
the programmer".

ALL of these has nothing to do with what the random number generation
that is done in the vp machines - completely different.

[rgmustain]

gilbert wrote:
Of course, it's all my fault to be misunderstood -- for being not
smarter than a 5th grader. What's clear to my brain is not really
what's clear to another's brain.

It's clear to me. Read on.

So here's hopefully a sample that may clear things up:
(pretending that I am an IGT programmer, where IGT stands for

Indian

Gaming Technology)

(from a software test tool we used)
testcase tcGetMeRandomNumbers () appstate none
  INTEGER iMin, iMax
  iMin = 1
  iMax = 10
  INTEGER iSeed
  iSeed = 2
  RandSeed (iSeed)
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10
  Print (RandInt (iMin, iMax)) // get me a number between 1 and 10

Results:
1st run:
1
10
8

2nd run:
1
10
8

Note: same results each time. I didn't tell the computer to give

me

the numbers 1, 10, and 8. The RandInt () function got it for me.

Change the iSeed to 3 from 2 and run the same test.

Results:
1st run:
1
3
4

2nd run:
1
3
4

If I change the values for iSeed, iMin and iMax, the "what I refer

to

as random" numbers generated will be different again.

Gilbert. If you CONTROL the results they ARE NOT random. How many
times do I have to say this. Just because you're calling a function
named RandInt does NOT mean the results are random.

Now, if you put the call to RandInt in a loop running constantly and
then interrupt that loop at some arbitrary time in the future, guess
what? You won't be able to determine what number will appear and the
result WILL BE random for all practical purposes.

NOTE: This is just to demonstrate that I, the programmer, did
control what random numbers to generate each run.

Like I said before, they ARE NOT random. You're still confusing
random with the name of a function that MAY be used to create random
numbers but does not do so in and of itself.

Another way to make it random is to seed the function RandInt with
something that is unpredictable. VP machines could have many values
that can be used to help seed the RNG. For example,

- time between last two hands
- average time between hands for some variable interval of time
- lifetime coin-in
- daily coin-in
- number of inserted bills of various denoms
- last time machine door opened
- etc., etc., etc.

Some algorithm utilizing variables such as these to seed the RNG will
make it pretty much impossible to predict the output.

No numbers outside
of the iMin and iMax that I specify in my program --- will come out.
This is all what I meant when I refer to "randomness controlled by
the programmer".

And, once again, it is NOT randomness.

ALL of these has nothing to do with what the random number

generation

that is done in the vp machines - completely different.

Absolutely true. So why did you say

"In short, the randomness programmed in these vp machines are still
in favor of the casinos." ? and

"If the randomness of vp machines are set to favor the players, many
casinos will be out of business. It is probably set to such a huge
number that no player can see any pattern to memorize or do something
to win all the time - which is what I mean by it in favor of the
casinos." ???

In both of these quotes you are claiming that the results of VP play
are not really random. You can't "favor" anything if it's random and
fair. If your words are not what you meant to say then I apologize
for debating this point.

Dick

···

--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@...> wrote:

[rgmustain]

--- In vpFREE@yahoogroups.com, Tom Robertson <thomasrrobertson@...>
wrote:

mroejacks wrote:

>RNG's are pseudo-random code as you know. If you know the

algorithm

>and exactly where you are in the algorithm then you can predict
>future results. I don't think anyone would consider that random.

Yes, from the perspective of one who knows how to determine the
results of an RNG, those results are not random, but the definition

of

randomness must include perception in order to be relevant, since
"true" (or "pure" or "intrinsic") randomness can't exist.

>Progammming will NEVER be able to produce truly random results.
>Programming can produce a sequence of numbers with an even
>distribution over time. When combined with reseeding and user
>interaction the result is random enough.

"Random enough" could be rephrased as "completely random, from the
perspective of an observer who is completely ignorant of what

results

will be generated."

I have no problem with this assertion although I think some security
pundits might disagree. The gaming commission tests the output of VP
machines with specific randomness tests. If they pass those tests,
they are "random enough".

>However, the main problem is
>Gilbert keeps inferring that IGT may be controlling the output of
>their RNG to produce "control"ed results. He also infers this

would

>still be random.

I don't see what's wrong with that. There's no empirical way to

prove

that IGT has made their machines fair

Hopefully, the gaming commission's testing is "empirical enough"

and if they aren't, they can be
just as random as if they were.

Absolutely. Random and fair are two distinct elements.

Gilbert is probably being
misunderstood by using the phrase "changing randomness" when, at

least

in the case of the tournament machines at Rincon, he means changing
fairness, I assume.

That's only part of Gilbert's problem. As I've said before, he has
stated that IGT controls ("favors") the results (which mean they
cannot be random and fair as required by law).

Dick

[brucem]

In theory only, not in practice. There are other factors preventing
a player from taking advantage of knowing the algorithm, as I'm sure
you know. The primary one is the speed of the RNG. RNG's process
through a giant "loop" in random numbers, billions of numbers long.
But they whiz through them so fast it's impossible to know where you
are at any given time, no matter how precise your stopwatch is.
Plus there is the "mapping" issue ... how a random number is
translated into a specific card. I'm pretty sure the mapping rules
change frequently, possibly after every game, or every few games.
Another problem is the "seed" number. We don't know what it is, or
when it resets. Finally, are we sure random numbers are used
successively? I seriously doubt they are, for reasons I won't go
into. So, what rules define the "skips" in selecting random
numbers? I doubt IGT programmers can answer that question.

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:

RNG's are pseudo-random code as you know. If you know the algorithm
and exactly where you are in the algorithm then you can predict
future results.

[Gilbert_Escobar]

gilbert wrote:
Maybe this can help you understand it this time...
Even the best computer hackers can't probably figure out how to deal
with these random numbers generated for vp machines . How else can
ordinary vp players figure them out?
If they can then I'll say "this randomness favors the players" if its
easy enough for them to find ways to beat the system.
Since this randomness is something that will be difficult to hack or
something, then I'll say again that "this randomness favors the
casinos".

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:

That's only part of Gilbert's problem. As I've said before, he has
stated that IGT controls ("favors") the results (which mean they
cannot be random and fair as required by law).

Dick
--------------------------------------

[rgmustain]

> RNG's are pseudo-random code as you know. If you know the

algorithm

> and exactly where you are in the algorithm then you can predict
> future results.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>
In theory only, not in practice. There are other factors

preventing

a player from taking advantage of knowing the algorithm, as I'm

sure

you know. The primary one is the speed of the RNG. RNG's process
through a giant "loop" in random numbers, billions of numbers

long.

But they whiz through them so fast it's impossible to know where

you

are at any given time, no matter how precise your stopwatch is.
Plus there is the "mapping" issue ... how a random number is
translated into a specific card. I'm pretty sure the mapping rules
change frequently, possibly after every game, or every few games.
Another problem is the "seed" number. We don't know what it is, or
when it resets. Finally, are we sure random numbers are used
successively? I seriously doubt they are, for reasons I won't go
into. So, what rules define the "skips" in selecting random
numbers? I doubt IGT programmers can answer that question.

I was only referring to RNGs in general, not VP machines. Everything
you added here is absolutely right on. If I were an IGT programmer I
might constantly be changing the mapping algorithm, reseeding the RNG
with other available values and have multiple algorithms to choose
from. Nothing like a little overkill just in case ...

If I was really paranoid I might even reclock the processor every so
often.

Dick

···

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

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@> wrote:

[rgmustain]

>
> That's only part of Gilbert's problem. As I've said before, he

has

> stated that IGT controls ("favors") the results (which mean they
> cannot be random and fair as required by law).
>
> Dick
>--------------------------------------

gilbert wrote:
Maybe this can help you understand it this time...
Even the best computer hackers can't probably figure out how to

deal

with these random numbers generated for vp machines .

And how would anyone know? Do you think they'd advertise? The group
that hacked early IGT machines many years ago didn't advertise
although they did get caught.

How else can
ordinary vp players figure them out?
If they can then I'll say "this randomness favors the players" if

its

easy enough for them to find ways to beat the system.

If players can find a way to "beat the system" then by definition it
was never random.

Since this randomness is something that will be difficult to hack

or

something, then I'll say again that "this randomness favors the
casinos".

No it doesn't. You can try to "help (me) understand" all you want but
it doesn't change the fact that if it's random it doesn't favor
anyone. I think you're getting in trouble with the word "random", if
you continue to use it you are always going to be contradicting
yourself.

I get the impression what you are trying to say is ... THE CARD
SELECTION ALGORITHM used in VP machines can not be determined and
therefore it could be weighted (that is, NOT RANDOM) to favor
casinos. Once again, that means the algorithm IS NOT random and fair
as required by law AND the comprehensive testing by the gaming
commission (which also may include code inspections) has failed to
detect it. YES, they do test it to make sure it doesn't FAVOR anyone.

Dick

···

--- In vpFREE@yahoogroups.com, "gilbert_616" <gilbert_616@...> wrote:

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@> wrote:

[Peter]

It is the following that makes things "random" no matter what the pseudo random number
generator is (as long as the output of the PRNG is not a constant).

You will never be able to "PREDICT" what will be the output of your function call to RandInt.

.....bl

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:

Now, if you put the call to RandInt in a loop running constantly and
then interrupt that loop at some arbitrary time in the future, guess
what? You won't be able to determine what number will appear and the
result WILL BE random for all practical purposes.

[rgmustain]

--- In vpFREE@yahoogroups.com, "bornloser1537" <bornloser1537@...>
wrote:

>
> Now, if you put the call to RandInt in a loop running constantly

and

> then interrupt that loop at some arbitrary time in the future,

guess

> what? You won't be able to determine what number will appear and

the

> result WILL BE random for all practical purposes.
>

It is the following that makes things "random" no matter what the

pseudo random number

generator is (as long as the output of the PRNG is not a constant).

You will never be able to "PREDICT" what will be the output of your

function call to RandInt.

.....bl

True for any single number and almost any algorithm would work in
this case. The actual randomness would be in the "arbitrary time
period". Only if multiple numbers were taken at that arbitrary time
period would it be necessary for the algorithm itself to produce
random results.

This raises a question about RNGs in general. It's been discussed
before that there are many possible methods IGT could use, but what
would be the MOST random? Some have indicated that the RNG could be
used to constantly shuffle a virtual deck of cards, others have
indicated that the random numbers themselves could be used (which was
the method used initially according to "Hacking the Casinos for a
Million Bucks").

Both approaches require mapping the random number from it's range (32
bits?) into a value from 1-52 ("which, according to AP
Lawrence, "Anytime you are hashing, folding or mapping, you run the
risk of weighting your distribution somewhere"). This operation
also "reduces the randomness" of the result as does any mapping from
a binary RNG to a value between 1-52. My question is, does anyone
know of a method IGT might use to utilize the full randomness of 32
bits without weighting the distribution?

One possibility might be to separate each 32 bit value into 5
distinct 6 bit values and use each of them to select a card or
shuffle the deck. This only loses about 4-5 bits of randomness but
still has some weighting issues.

Dick

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@> wrote:

[cdfsrule2]

This raises a question about RNGs in general. It's been discussed
before that there are many possible methods IGT could use, but what
would be the MOST random?

The answer of coarse depends upon you deifne "MOST random. But generally speeaking, a
well designed hardware RNG based on a "random" physical process, such as thermal noise
in a resistor or noise in a diode, would be "MORE random" than a software only approach.
You might want to take a look at: http://en.wikipedia.org/wiki/
Hardware_random_number_generator

I think pentium chips (not in VP machines!) use a thermal approach. You can buy "boxes"
that provide quite random numbers such as http://qrbg.irb.hr/

BTW, that the wiki article isn't complete or perfect. But its a good place to start your
investigation

My question is, does anyone
know of a method IGT might use to utilize the full randomness of 32
bits without weighting the distribution?

"Whitening" the distribution (actually whitening the random bit stream) for VP is easy and
likewise it is easy to make a mistake and add both bias or non-whiteness. The bits, 6
(2^6 = 64) or 32, or whatever have no real impact. 32-bits is not MORE random than 6
bits, etc. The resolution has nothing to do with "randomness" at all, so long as there are at
least 1 bit.
Also These kinds of random numbers are always band limited, and so not truly white ("flat
DC to Daylight" as I like to say). The key is to make sure numeric artifacts don't f-things
up, espcially binning or rounding. Finally, the requirement for "whiteness" is not so
restrictive as the requirement for bias. That is a slightly positve biased mean (higher EV)
could be a big problem for the casino, while a non-biased mean but slightly non-white
distribution would not be a big deal.