Theories, Myths, Facts And Ideas

[Joe]

Example of a dependent outcome is IF the probability of red hit next spin after 10consecutive reds is higher than 1/37.

lucky, the probability of red will always be higher than 1/37. I think you meant 18/37.

[luckyfella]

lucky, the probability of red will always be higher than 1/37. I think you meant 18/37.

Yes, you are correct.

[Joe]

Can someone please explain to me, in layman’s terms, what is a dependency as it relates to roulette?

Here’s hoping

I'll try, because I don't think luckyfella's reply was very helpful. IMO the best way to explain the difference between dependent outcomes and independent outcomes is using a deck of cards, and then relating it to roulette outcomes.

Suppose you draw cards from the deck. There are 52 cards; what's the probability that the card you pick is an Ace? Easy : there are 4 Aces so the chance is 4/52. Having picked a card there are now 2 options : you can 1) either put it back in the deck and re-shuffle, or 2) discard it.

Option 2 creates dependency, option 1 doesn't. If you knew that cards were not returned after being drawn but didn't know what the first card drawn actually was, and someone asked you to calculate the odds of getting another Ace, you would have to say 'It depends on what the first card drawn was'. If the first card drawn was an Ace, then the chance of drawing another Ace would be 3/51 because one Ace has been removed and there are only 51 cards remaining.

However, if the cards had been replaced and the deck shuffled after drawing the first Ace, calculating the chance of another Ace does not depend on the previous result; it's 4/52, just as it was for the first draw.

Roulette is like this (cards replaced after each draw) and the pockets around the wheel correspond to cards. Pockets are not blocked after successive spins, that's why the General keeps saying that the same number of pockets exist between spins, so it doesn't matter what previous spins actually were.  Red after 10 red in row is still 18/37, just as it is after 1 black.

lucky is right to say there must be a dependency to get an edge, but the question is - where does the dependency come from? It obviously doesn't come from the 'non replacement' principle I just explained, so where else? 

[Blood Angel]

I'll try, because I don't think luckyfella's reply was very helpful. IMO the best way to explain the difference between dependent outcomes and independent outcomes is using a deck of cards, and then relating it to roulette outcomes.

Suppose you draw cards from the deck. There are 52 cards; what's the probability that the card you pick is an Ace? Easy : there are 4 Aces so the chance is 4/52. Having picked a card there are now 2 options : you can 1) either put it back in the deck and re-shuffle, or 2) discard it.

Option 2 creates dependency, option 1 doesn't. If you knew that cards were not returned after being drawn but didn't know what the first card drawn actually was, and someone asked you to calculate the odds of getting another Ace, you would have to say 'It depends on what the first card drawn was'. If the first card drawn was an Ace, then the chance of drawing another Ace would be 3/51 because one Ace has been removed and there are only 51 cards remaining.

However, if the cards had been replaced and the deck shuffled after drawing the first Ace, calculating the chance of another Ace does not depend on the previous result; it's 4/52, just as it was for the first draw.

Roulette is like this (cards replaced after each draw) and the pockets around the wheel correspond to cards. Pockets are not blocked after successive spins, that's why the General keeps saying that the same number of pockets exist between spins, so it doesn't matter what previous spins actually were.  Red after 10 red in row is still 18/37, just as it is after 1 black.

lucky is right to say there must be a dependency to get an edge, but the question is - where does the dependency come from? It obviously doesn't come from the 'non replacement' principle I just explained, so where else? 

That’s brilliant Joe, thank you for your reply.

BA

[Tinsoldiers]

It obviously doesn't come from the 'non replacement' principle I just explained, so where else? 

What if we create unique streams of numbers as they come through roulette spins. What if we create a unique stream to trAck every time there is a repeat? Just thinking loud. 

[luckyfella]

What if we create unique streams of numbers as they come through roulette spins. What if we create a unique stream to trAck every time there is a repeat? Just thinking loud.

Why repeats ? What unique statistical property that repeats have ?

[Steve]

They "look" like a trend. Thats all.

Anyone saying otherwise hasnt properly tested.

Start by considering what countless people smarter than you have done, and consider their findings. Then check for yourself. Move on if they are correct.

Or keep going in circles.

[Joe]

What if we create unique streams of numbers as they come through roulette spins. What if we create a unique stream to trAck every time there is a repeat? Just thinking loud. 

tin, how does that create a dependency or identify one?

[Steve]

We if we start from where others finished, so we didnt need to repeat their wasted time?

What if we started with what is already well known?

Then maybe we could finish in a better place than others who spent their life reinventing the wheel.

[Joe]

Turbo's latest thread on GF is the best explanation of his betting strategy.

I've read it and it's more confusing than the first thread he made.  His strategy seems to be based on averages playing out in the short term which is a classic error - basically gambler's fallacy.

Now the naysayer might say "Heck - that last number might not show up for 300 spins !!!!!"
Well, then we wouldn't have an average of 6.3 per number in the winning set - and we
already know that's the case... because our data and bets are always on the set that wins.

[falkor2k15]

I only know 2 types of dependency: same spin and uniques vs. repeats.

1. When there's a unique dozen then there's also a unique number; likewise, when there's a repeat on the numbers then there's a repeat on the dozens = same spin dependency
2. A repeat depends on the uniques that came before. If we are betting for a repeat and we have 123 on the dozens then we can only get a cycle length 3 as CL1-2 is no longer available. On the other hand if we had just 12 then we could get a CL2 or a CL3.

Any other forms of dependency? I don't think so. Priyanka described summing up numbers: if you want the next spin and the last number to add up to 40 then it depends on how close to 20 is the first number. Somebody added up normal lines and positional lines to make 10. I don't find any application though.

[Joe]

I only know 2 types of dependency: same spin and uniques vs. repeats.

Neither of which are any good for knowing what the next outcome is going to be.

[falkor2k15]

Neither of which are any good for knowing what the next outcome is going to be.

Like I said if you have 123 next spin has to be CL3 and cannot be CL1-2, so at that point you can be sure of the next outcome. Similarly if just 12 then next outcome cannot be CL1, so there's some kind of predictability/dependency re: the next outcome, but I don't find any application.

[Blueprint]

What about unique repeats?

[Tinsoldiers]

tin, how does that create a dependency or identify one?

Steve, let this continue. We all know what you are saying and you don’t have to save us from losing our life savings on something we don’t know. Sometimes shaking up basics and believing blindly doesn’t help. We are currently having a discussion and first thing that we established that even math people cannot refuse is we cannot prove that there are no dependent streams in roulette.  So something that cannot be proven cannot be ignored. 

Joe and Lucky, I was just thinking loud to Joe’s question of explaining the non replacement principle. I was trying out things to see whether we can replicate the same card experiment. That’s all. I can’t see it creating any dependence though.