What Does Reading Random Mean

[cheese]

It comes out with a loss of - 1.35% of the wagered amount, same shit happens if I bet red after 4 or 5 blacks.

Its called unconditional probability, it rules roulette. Thats why you have to invent a game from the random outcomes that can use conditional probability. Its doesn't have to be anything close to accurate, just so it wins more than 50% of the time.

[Kelly]

The calculation for unconditional probability is:


To find the unconditional probability of an event, sum the outcomes of the event and divide by the total number of possible outcomes.

What could that possibly gain  in roulette ?

[cheese]

What could that possibly gain  in roulette ?

Thats what I said, its rules roulette, you can't win with it. You have to invent a game from the actuals where you can use conditional probability and play that instead.

[Kelly]

Conditional probabilty relys on that spins are not independend, which they are. If they are not,  it must be possible to construct some sort of proof that they aren`t. You can apply conditional probability to BJ and Poker where some cards cannot be in 2 seperate places at once. In roulette there is nothing that stops an occurence although the accumulated probability might get smaller or larger. In BJ, there is a physical condition when a lot of small cards has dissapeared from the deck, that gives the player a small advantage in the pay out on hitting a Black Jack. In roulette there are no physical conditions that has changed between 2 spins.

[Mike]

What kelly said. 

Of course it's possible to do probability calculations using Bayes' theorem because it can be used even if you're ignorant of whether the distribution is independent or not. You can also invent a "game within a game" if you want, but none of that will help you to pick more winners than expectation because on each spin there are no outcomes which are eliminated (as in the case of BJ or Poker).

Bayes' theorem will only confirm what should be obvious: that outcomes are independent. You can't just "decide to use" conditional probability (which doesn't even make sense) and somehow magically make the outcomes dependent! this is a serious misunderstanding of probability and math in general.

Conditional probability is only sensible when the set of outcomes is restricted in some way, so if cards are removed from a deck then obviously those cards cannot re-appear and any probability calculations are with respect to that reduced set of cards.

[Mike]

It's all very well having the electric hat and a T-shirt with "Bayes or Bust" emblazoned on it, but unless you understand the concepts you end up looking like an idiot. 

[cheese]

Bayes' theorem will only confirm what should be obvious: that outcomes are independent.

Nobody's arguing they aren't independent. What I'm saying is, you have a invent a game to play inside the game of roulette where you can use conditional probability to win more than 50% of the time.

[gizmotron]

So, now its your turn. How do you apply conditional probability to create an edge over the house ?

For one thing I don't start with the wheel has no memory. That's a red herring if I ever saw one. The wheel doesn't need a memory in order to create a unique sequence of spins. I start with the fact that the most recent spins have formed a unique sequence. From that sequence I can accumulate data that is real. I can check it for baseline references to deviations from the baseline. I can apply rule based decisions on it to form both trend and pattern formations. These formations can take on the shapes of a rhythm, much like music does. I use this to produce win streaks. I know how to use these moments to take down the casinos. Sometimes the rhythm of the patterns are so perfect that they form into a perfect win streak. That's a moment when opportunity reaches an almost 100% quality to it. The fact that there are those that won't take advantage of these moments doesn't surprise me. They would rather listen to the wood in the wheel. It talks to them. The wood has no memory.

[Mike]

The wheel doesn't need a memory in order to create a unique sequence of spins.

What?

Of course it does. So you're saying that a sequence can't repeat when on each spin each number has the same chance of appearing?

The only way a sequence would be guaranteed to be unique is if some numbers were removed from the wheel - ie: outcomes were dependent.

[cheese]

I can apply rule based decisions on it to form both trend and pattern formations.

Yes, you can. You can apply them all day long. You can apply rules to random events till you're blue in the face and the cows come home. You can apply them 24/7. You can also attempt to staple jello to the wall. See which one works first.

[Mike]

Nobody's arguing they aren't independent. What I'm saying is, you have a invent a game to play inside the game of roulette where you can use conditional probability to win more than 50% of the time.

Ok, you can invent your game but eventually you have to deal with reality. Having a game where you pretend that outcomes are dependent doesn't change the reality that they aren't, and sooner or later you have to place a bet. I'm not sure whether you're saying that outcomes are independent or they're not independent. You can't have it both ways.

[cheese]

Having a game where you pretend that outcomes are dependent doesn't change the reality.

Of course not. But it skews it enough to gain an edge. It doesn't change anything except the way you look at it. I makes guesses based on my game, I have no idea whats coming next. If I'm wrong, I'm never ever surprised, there are no 'good' bets, there is no 'better' time to bet.

[gizmotron]

Bayes' theorem will only confirm what should be obvious: that outcomes are independent. You can't just "decide to use" conditional probability (which doesn't even make sense) and somehow magically make the outcomes dependent! this is a serious misunderstanding of probability and math in general.

A recent unique sequence of spins is independent of the expectation inference of the long termed sample. You can use pattern recognition technique on any unique sequence.

"In machine learning, pattern recognition is the assignment of some sort of output value (or label) to a given input value (or instance), according to some specific algorithm. An example of pattern recognition is classification, which attempts to assign each input value to one of a given set of classes "

I'm using an algorithm to assign output values to unique sequences of current data that is real.

You are assigning that all data from past spins obtains a qualification value of nothing. You demand that all others seek this same data sample of nothing from witch you extrapolate nothing. You are guilty of your own inference. You are assigning it a value too.

[cheese]

I'm using an algorithm to assign output values to unique sequences of current data that is real.

Assigning values  to unique random outcomes is an exercise in futility. Random follows no rules. No matter how hard you wish, and how hard you curve fit, and how much you dream, they will never follow any rules.

[cheese]

Pattern recognition is not an exercise in futility.

Assigning rules to pattern recognition is an exercise in futility. I do this day in and day out, I know no rules whatever can work with random outcomes. You can't give ONE example where it works.

[cheese]

-