The Naked Scientists

The Naked Scientists Forum

Author Topic: Online Texas holdem poker is flawed by ''time travelling'' probability.  (Read 1718 times)

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Part-1   Time Travelling probability

In a vast Universe of the unknown, probability is a measurement, a measurement of the expected chance of an event occurring at a specific moment in time.  In example, if we were to role a six sided dice, we would have a one of six chances of the dice to stop rolling and stopping on  any specific number at that specific moment in time.
The same applies to a shuffled deck of cards, any of the fifty two individual variants of a deck of cards has  one of fifty two chances of being the top card from each individual shuffle of the deck.
In mathematics we would represent the dice example by 1/6 and in the deck of cards example we would write 1/52.

In games that are probability based, the observed outcomes over continuous time or a continuous time period produce a random sequence of results,  a history of results.  A good example would be the result screen on a roulette table, red,red,black,red,black,red.   Another example would be a poker players hand histories, these are the results and the history of a past event, of a specific moment in continuous time from the past.

Continuous time being of importance to my theory.

In explaining some fundamentals, I will now move on to the actual question and what the theory entails.

In a live game of Texas holdem poker, a single deck of cards is used per table for distribution of the cards to the players, the same deck of each table is then reshuffled and then re-distributed to the same table!
Online Texas holdem poker,a new pre-shuffled deck from a queuing system is brought to each table every hand.

Can this have effect on probabilities and outcome over time?

My theory can lead to only one conclusion, I will show by incorporating vectors and time into my theory, that conclusively online probabilities are dependent to players intercepting variants by ''time travelling'' rather than receiving variants throughout continuous time.



Before I continue and put in a lot of effort writing this all up, based on so far, does anyone want to be my peer and peerview this?


 



 




 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Does this fit my problem?

''Considers a random walker
which starts on the origin at time t = 0. It stays fixed to its position until time t1, it then
makes a jump to ∆r1, the particle waits on ∆ri until time t2 > t1 when its jumps to a new
location ∆r1+∆r2, the process is then renewed. The dots on the time axis {t1,t2, } define
the times of jumping events. The times τ1 = t1 − 0, τ2 = t2 − t1 etc are called waiting times.
In the CTRW the waiting times {τ1, τ2...} and the displacements {∆r1, ∆r2...} are mutually
independent identically distributed random variables. ''
 

Offline chiralSPO

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 1879
  • Thanked: 145 times
    • View Profile
The mechanism of the random walk is very different from selecting various pre-shuffled decks of cards. The decks are not transformed from one to the next to the next.

However the end result is the same. The order in which the deck is drawn is mutually independent of the order of the cards within the deck.

(answering this question does not mean that I volunteer to peer review this--I would advise not putting to much effort into this line of questioning until you have spent some time learning more about probability calculations. They're tricky at first, but there are only so many types of scenarios that commonly arise that once you have learned a few of the tricks, it's not so bad)
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
The mechanism of the random walk is very different from selecting various pre-shuffled decks of cards. The decks are not transformed from one to the next to the next.

However the end result is the same. The order in which the deck is drawn is mutually independent of the order of the cards within the deck.

(answering this question does not mean that I volunteer to peer review this--I would advise not putting to much effort into this line of questioning until you have spent some time learning more about probability calculations. They're tricky at first, but there are only so many types of scenarios that commonly arise that once you have learned a few of the tricks, it's not so bad)

Thank you for your reply it seems lonely on here at the moment, I believe I have the maths proof now and a good explanation,


posted elsewhere on a maths forum

If we had two variants 1 and 2, and were to conceal their identities, then randomly shuffled the 2 variants, we would have a 1/2 chance of the value 1, being the left aligned of the two variants.

If we were to add a second set of the two variants and randomized them, this would also have a 1/2 chance of 1 being the left aligned.

Now if I was to say, that 1 is a winner and 2 is a loss and your value you would receive is always the left position, we know that if we choose either set, our chance of 1, being aligned to the left by either set is 1/2.

At this point I considered the above and used (n) to represent the above unknown variants in diagram form.

set (a)nn
set (b)nn


The first thing I noticed was each set expanded across a X-axis. Each set remaining a 1/2 chance of 1 being the variant left aligned of each sets X-axis.

Secondly we notice that using multiple sets creates a Y-axis. Rows randomly shuffled that once stopped, creates columns of variants.


The question, what is the chance of receiving a 1 from the aligned left column if we were to pick any one of the sets after the shuffle?
(baring in mind, the right column is removed from the equation)

n
n

My maths to this point-

P(1)/X=(1/2)/t1


P(1)/Y=(?/2)/t2

With a range of entropy minimum 0/2 to a maximum of 2/2.

further maths-

nn
nn

P(1)/^Y=(?/2)/(d/t2)

I believe that if science can not replace ? with a true value, then this alone proves I am correct
 

Offline Colin2B

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 1921
  • Thanked: 125 times
    • View Profile
I believe that if science can not replace ? with a true value, then this alone proves I am correct
Neither science nor maths can do this.
It does not prove you are correct as unfortunately your maths does not make sense.
The reasons have been explored in the other thread and I don't think you are going to find anyone willing to spend the time on this thread.
I echo what ChiralSPO said, do learn basic probability, also the maths of fractions and decimals.
Please don't waste more of your time and energy on this, there are better things to do with your life.
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile

Neither science nor maths can do this.
It does not prove you are correct as unfortunately your maths does not make sense.
The reasons have been explored in the other thread and I don't think you are going to find anyone willing to spend the time on this thread.
I echo what ChiralSPO said, do learn basic probability, also the maths of fractions and decimals.
Please don't waste more of your time and energy on this, there are better things to do with your life.

so although i am correct, and online poker is flawed, you suggest i just drop it and let 10000000s of  players to continue playing a game that they think is poker and lose their money?
 

Offline Colin2B

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 1921
  • Thanked: 125 times
    • View Profile
I'm saying you are not correct with your maths, it doesnt prove anything.
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
I'm saying you are not correct with your maths, it doesnt prove anything.

it does prove it, if nobody can fill in the ? it proves i am correct.


nn
nn

x axis = 1/2    always, there can only ever be 1 one.

y=?


probability of 1 from y is ?


this theory is real and true,by adding choice, the axis shifts from x,52 individual variants to Y, unknown.

X=52

Y=?


I even know how to fix it.

each deck starts out 1/52 a x-axis


1/52
1/52


we add choice we switch to y-axis ?/2


?/52
?/52
p

if the dealer removed the alignment by not dealing a specific order, the deck returns to 1/52 and the x-axis
« Last Edit: 14/08/2015 11:18:31 by Thebox »
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
new maths


P(x)/X=1=(x)/(d/t1)

P(x)/Y=σ2=(x)/(d/t2)

X=....................................................................x..........................
X=..................x............................................................................
X=.............................................x.................................................
X=....................................................................x..........................
///////////////Y//////////////////Y//////////Y/////////////Y///////////////////
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Would it please be possible for one of the mods to please change the title of the theory to ,  By adding choice, Vector X is switched to Vector Y and this creates a probability function change.
 

Offline Kenyonm

  • Jr. Member
  • **
  • Posts: 18
  • Thanked: 2 times
    • View Profile
Is probability just a manifestation of the human mind trying to make some sense of ordered chaos. By ordered meaning apparent chaos being produced by historical inputs. To demonstrate this consider that a deck of cards cannot be shuffled by hand or by a machine into a completely random order. Such that the chance of turning over an ace at the top of the pile is not 4/52 as the non random shuffling has placed the ace at the top of the pile resulting in the probability being 1.
The national lottery is not random as the balls are put in in order. The time of the button press is more or less the same to start the picking process. This produces a assumed random result but in fact it isn't. The recent move to 59 balls was a result of people seeing patterns in the result and writing programs to predict the outcome. A low set with balls all under 15 come out frequently. The select 6 balls from 15 dramically increases the chance of winning. People have realised this. History therefore changes probability as it does the outcome.
 

Offline alancalverd

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 4725
  • Thanked: 155 times
  • life is too short to drink instant coffee
    • View Profile
probability is a measurement
Wrong. Probability is  an a priori estimate. Measurement is an a posteriori statement of fact.

Quote
Can this have effect on probabilities and outcome over time?
No.

Quote
History therefore changes probability as it does the outcome.
no. History (i.e. measurement) allows us to refine our estimate of probability. We can start by assuming the dice are not loaded, but if we get a significant excess of 6s we might suspect that the dice are not entirely symmetrical.

The national lottery ball story is a simple example of a problem that turns up in many mass-production industries. How do you ensure a consistently even distribution of, say, pigments in car paint? Easy enough if you only have one pigment, but if you add something like metalflake you need to be sure that the mixing process doesn't allow clumping. My favourite example is Ambriosa Rice Pudding. This used to (maybe still does) come in ordinary and deluxe varieties. The difference is that deluxe contains exactly four sultanas per can: any more and it turns a yukky brown color on heating, and any less produces complaints that it doesn't taste as good as it used to. Simply stirring an appropriate number of sultanas into a ton of rice pudding doesn't  produce an adequately homogeneous mixture. The machinery actually inserted four in each can, and even that isn't sufficiently consistent: each can was then x-rayed to count the sultanas! 

Back to poker. Hand-shuffling isn't entirely random. Cards are "clumped" by the players trying to assemble the best possible hand by discard and purchase. A simple shuffle at the end of a round redistributes clumps of cards, so pairs and even threes may be more frequent than random in the next deal. If the dealer deals "threes and twos" the hands gradually strengthen as the evening wears on. A riffle shuffle splits pairs, but still retains more order than a random distribution. Online poker, as described here by Mr B, should be entirely random at each deal since there is no "history" in the new pack, and each deal is therefore truly independent.
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
probability is a measurement
Wrong. Probability is  an a priori estimate. Measurement is an a posteriori statement of fact.

Quote
Can this have effect on probabilities and outcome over time?
No.

Quote
History therefore changes probability as it does the outcome.
no. History (i.e. measurement) allows us to refine our estimate of probability. We can start by assuming the dice are not loaded, but if we get a significant excess of 6s we might suspect that the dice are not entirely symmetrical.

The national lottery ball story is a simple example of a problem that turns up in many mass-production industries. How do you ensure a consistently even distribution of, say, pigments in car paint? Easy enough if you only have one pigment, but if you add something like metalflake you need to be sure that the mixing process doesn't allow clumping. My favourite example is Ambriosa Rice Pudding. This used to (maybe still does) come in ordinary and deluxe varieties. The difference is that deluxe contains exactly four sultanas per can: any more and it turns a yukky brown color on heating, and any less produces complaints that it doesn't taste as good as it used to. Simply stirring an appropriate number of sultanas into a ton of rice pudding doesn't  produce an adequately homogeneous mixture. The machinery actually inserted four in each can, and even that isn't sufficiently consistent: each can was then x-rayed to count the sultanas! 

Back to poker. Hand-shuffling isn't entirely random. Cards are "clumped" by the players trying to assemble the best possible hand by discard and purchase. A simple shuffle at the end of a round redistributes clumps of cards, so pairs and even threes may be more frequent than random in the next deal. If the dealer deals "threes and twos" the hands gradually strengthen as the evening wears on. A riffle shuffle splits pairs, but still retains more order than a random distribution. Online poker, as described here by Mr B, should be entirely random at each deal since there is no "history" in the new pack, and each deal is therefore truly independent.
probability is a measurement
Wrong. Probability is  an a priori estimate. Measurement is an a posteriori statement of fact.

Quote
Can this have effect on probabilities and outcome over time?
No.

Quote
History therefore changes probability as it does the outcome.
no. History (i.e. measurement) allows us to refine our estimate of probability. We can start by assuming the dice are not loaded, but if we get a significant excess of 6s we might suspect that the dice are not entirely symmetrical.

The national lottery ball story is a simple example of a problem that turns up in many mass-production industries. How do you ensure a consistently even distribution of, say, pigments in car paint? Easy enough if you only have one pigment, but if you add something like metalflake you need to be sure that the mixing process doesn't allow clumping. My favourite example is Ambriosa Rice Pudding. This used to (maybe still does) come in ordinary and deluxe varieties. The difference is that deluxe contains exactly four sultanas per can: any more and it turns a yukky brown color on heating, and any less produces complaints that it doesn't taste as good as it used to. Simply stirring an appropriate number of sultanas into a ton of rice pudding doesn't  produce an adequately homogeneous mixture. The machinery actually inserted four in each can, and even that isn't sufficiently consistent: each can was then x-rayed to count the sultanas! 

Back to poker. Hand-shuffling isn't entirely random. Cards are "clumped" by the players trying to assemble the best possible hand by discard and purchase. A simple shuffle at the end of a round redistributes clumps of cards, so pairs and even threes may be more frequent than random in the next deal. If the dealer deals "threes and twos" the hands gradually strengthen as the evening wears on. A riffle shuffle splits pairs, but still retains more order than a random distribution. Online poker, as described here by Mr B, should be entirely random at each deal since there is no "history" in the new pack, and each deal is therefore truly independent.

I think I have give up on this one Alan although I am sure I am correct , but me versus the world I can not win ever.

A 4/52 period of time is not the same has a 100*(4/52) based on using 100 decks.

But like I said I give up, everyone keeps quoting back x and ignoring y.




 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Is probability just a manifestation of the human mind trying to make some sense of ordered chaos. By ordered meaning apparent chaos being produced by historical inputs. To demonstrate this consider that a deck of cards cannot be shuffled by hand or by a machine into a completely random order. Such that the chance of turning over an ace at the top of the pile is not 4/52 as the non random shuffling has placed the ace at the top of the pile resulting in the probability being 1.
The national lottery is not random as the balls are put in in order. The time of the button press is more or less the same to start the picking process. This produces a assumed random result but in fact it isn't. The recent move to 59 balls was a result of people seeing patterns in the result and writing programs to predict the outcome. A low set with balls all under 15 come out frequently. The select 6 balls from 15 dramically increases the chance of winning. People have realised this. History therefore changes probability as it does the outcome.




Every one will argue that is gamblers fallacy, I call it ''reverse odds. ''   

 

Offline alancalverd

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 4725
  • Thanked: 155 times
  • life is too short to drink instant coffee
    • View Profile
Quote
everyone keeps quoting back x and ignoring y.

That is because each online hand is entirely independent of all previous hands, unlike real cards. Your "y" is by defintion irrelevant in online poker, but very relevant in the real game. Playing with real cards, you are much more likely to be dealt three aces at the end of the evening than at the beginning. But so is everyone else! 
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Quote
everyone keeps quoting back x and ignoring y.

That is because each online hand is entirely independent of all previous hands, unlike real cards. Your "y" is by defintion irrelevant in online poker, but very relevant in the real game. Playing with real cards, you are much more likely to be dealt three aces at the end of the evening than at the beginning. But so is everyone else!

Huh, that sounds contradictory. How are  you making the two different? 


 ''Your "y" is by definition irrelevant in on-line poker, but very relevant in the real game''
 

Offline alancalverd

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 4725
  • Thanked: 155 times
  • life is too short to drink instant coffee
    • View Profile
Just read what I wrote. Every deal in online poker is entirely random and independent of all other deals. Every successive 3/2 deal in real poker is subtly dependent on the previous round. They are different because they are different!
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile
Just read what I wrote. Every deal in online poker is entirely random and independent of all other deals. Every successive 3/2 deal in real poker is subtly dependent on the previous round. They are different because they are different!

So to confirm, you are saying, that using a single deck, your chance of an ace is 4/52, if you receive an ace, the chance the next shuffle of receiving an ace, is (4/52)^2.


On-line, my chance remains 4/52, a greater chance of receiving another ace?
 

Offline chiralSPO

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 1879
  • Thanked: 145 times
    • View Profile
Just read what I wrote. Every deal in online poker is entirely random and independent of all other deals. Every successive 3/2 deal in real poker is subtly dependent on the previous round. They are different because they are different!

So to confirm, you are saying, that using a single deck, your chance of an ace is 4/52, if you receive an ace, the chance the next shuffle of receiving an ace, is (4/52)^2.


On-line, my chance remains 4/52, a greater chance of receiving another ace?

No, the chance is always 4/52. There is a 4/52 chance of drawing an ace from one shuffled deck. If you draw one card from each of two shuffled decks, there is a (4/52)2 chance of drawing two aces. But if you draw from one deck and get an ace, the chance of drawing an ace from the next deck is still 4/52. (together this makes (4/52)2...)

Seriously, just read an introductory book on statistics. I learned this stuff in grade school, so there should be many different introductory level textbooks available that don't require a lot of background knowledge.

Something like this might be a good place to start: https://www.mathsisfun.com/data/probability-tree-diagrams.html
 

Offline Thebox

  • Neilep Level Member
  • ******
  • Posts: 3164
  • Thanked: 46 times
    • View Profile


No, the chance is always 4/52. There is a 4/52 chance of drawing an ace from one shuffled deck. If you draw one card from each of two shuffled decks, there is a (4/52)2 chance of drawing two aces. But if you draw from one deck and get an ace, the chance of drawing an ace from the next deck is still 4/52. (together this makes (4/52)2...)

Seriously, just read an introductory book on statistics. I learned this stuff in grade school, so there should be many different introductory level textbooks available that don't require a lot of background knowledge.

Something like this might be a good place to start: https://www.mathsisfun.com/data/probability-tree-diagrams.html

Like I said before, I know your way and where you get 4/52 from all the time, I know why you say 4/52, but this does not consider Y at all. 

Like I said I give up because there is no way you will ever be able to see the paradox, so I will just leave it at that.

 

Offline alancalverd

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 4725
  • Thanked: 155 times
  • life is too short to drink instant coffee
    • View Profile
So to confirm, you are saying, that using a single deck, your chance of an ace is 4/52, if you receive an ace, the chance the next shuffle of receiving an ace, is (4/52)^2.
Not what I said at all. The probability of receiving one ace in a random deal of 5 cards is 5 x 4/52. The probability of receiving a pair of aces in one random deal is about (5 x 4/52)2. But during that play, anyone with a pair has a further 4/52 chance of getting a third ace with a single discard, so at the end of the play there is a likelihood that some pairs or even high singles have been improved into 3 or even 4 card groups.

Now if the deck is collected and hand-shuffled, these groups will be split, but not completely randomised, so there is a better than (4/52)2 chance of being dealt a pair in a 3/2 deal as the game progresses.

The maths is complicated by players making other choices. Whilst 2, 3, or 4 of a kind is an obvious target, straights and flushes are also attractive if you happen to be dealt with 4 or even 3 members, but it is unlikely that anyone would discard a picture or an ace in the hope of gaining a straight or a flush, so the effect on the distribution of high pairs is small.
« Last Edit: 22/12/2015 00:07:46 by alancalverd »
 

The Naked Scientists Forum


 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums