Theories, Myths, Facts And Ideas

[luckyfella]

Yes you can create 'conditional' bets, but that doesn't make the outcomes dependent.

That's what what I meant.

The purpose of the videos is to cover both these aspects.

[Herby]

Herby, this is a bit confused.

Thanks for correcting.
Since the user Bayes isn't here any more all the math here seems a bit confused.

Maybe you allow to say it is helpful to differentiate between discrete and continous distributions.

[luckyfella]

If the trials were dependent the binomial would be a lousy model and its predictions would be poor, but they are very good.

This has to be the most important basis then.


If roulette outcomes were to be,
R B R B R B R B 0 R B R 0 B R B R B.....

then nobody have a chance of winning, losing to house edge of extra pocket and unfair payout.

Since roulette outcomes is something like this,

R R B B B B R R B R B 0 R R R B B R R...

then some players win and others lose, the casino always win.

If the player bets R and it turns out B, he sees it as negative variance or "unlucky".

While another player bets B, he sees it as positive variance or "luck".

That's why I wrote earlier that the winning systems bet has to be good at predicting variance aka "luck".

[luckyfella]

That's why I wrote earlier that the winning systems bet has to be good at predicting variance aka "luck".

We all know that it's not possible to predict "luck".

The point I want to convey across with this predict "luck" posts is to highlight what YOU expect the systems bet has to do, that is to predict accurately.

To win you have to increase the accuracy of your predictions. -----steve unquote

So long as you have this predict requirement it's bound to fail.

No matter how awesome your method of prediction, it still lose against this spin sequence which is theoretically possible

R B R B R B R B 0 R B 0 R B R B R B 0 R ....

[luckyfella]

 (I can't edit the above post)
Or,

0 0 0 0 0

Or,

1 2 3 4 5

Or,

R R B B B R B R R 0 B B B B B B B B R B R B R B 0 R B 0 0 0 0 R R R R B R B.....

A mix of any of the above. Where everything and anything is theoretically possible that makes prediction impossible.

[MoneyT101]

This is what i was saying as contradicting. You say they dont and then say sometimes it can happen However, i got what you meant. A couple of questions for you, why 18, why not 17, why not 16 as the stop, why do you say it is 18, what is magical about it? At the end of the day, it is pure probability laws.

The statistic show that it happens at half... so it’s 36 numbers and half is 18.

The repeat comes past 19-36 less then 5% of the time.  I believe it’s around 1% but I don’t have the exact number.

So maybe sometimes isn’t the right word.  Because it’s much less then sometimes. It’s not a contradiction.  I’m just pointing it out that it can happen but 95%/99%vs 1%/5%.  Contradiction would be 90/10 maybe or 85/15.  But the percent is so low that I don’t see why anyone needs to even focus on it.

The Fact is repeat happens from last 18 recent numbers

Once it hits half at 19 anything beyond that isn’t recent.

Birthday paradox takes all numbers into account.  So if it’s past 19-36.  It will be considered.  But recent numbers are only 18.

Same as if we use splits... recent numbers is up to 9 splits.

Most recent numbers and birthday paradox have almost the same meaning but they are two different things.

[MoneyT101]

I will tell you what is difficult for everyone to get. You are saying LLN doesnt impact you, as you win every session you play within a certain number of spins. The argument from everyone is such a system doesnt exist which can win every session you play. Your example doesnt make sense as it is not the right one and as you say it needs to evolve. So when you are shaking the basic assumption of LLN to say that you are able to control your sessions to win every session, either we need better examples or we need better mathematic representations like i was showing in columns/rows. You think you might be able to try that? Else, you will not achieve your objective of sharing what you know.

I tried to show the bankroll against the bets taking place.

Column 1
L
H
L
L

Column 2
3
5
2
2

Basically if you have 1 bet going down column 1 in the long run it will be negative

If you have another bet  in column 2 in the long run it will be negative

In both cases LLN will kick in and give you a negative outcome.  We agree up to this point!

Column 1 has a 3 spin limit for a repeat to come. Column 2 has a 7 spin limit.

So I said if you can use the 3 spin limit from column 1 while betting on column 2. 

Then you have changed the LLN affect on column 2.  If you can always get a win in 3 spins maximum.

So it’s like picking a double street at an EC odds.  But more or less the spins it takes to end.

You increase the accuracy because instead of losing double streets all the way up to 7 spins your winning at 3.  So the LLN affect even tho you can place 1,000,000 bets.  It won’t affect you.

That’s basically what I said and tried to show.  Now I am not saying this exist, especially not in this format.

We are strictly talking LLN affecting this idea and how it neutralizes the affect.

This was the Example I used to make a point.

[luckyfella]

If you can always get a win in 3 spins maximum.

In 100spins, you win 33x for lines.

This has to be the best roulette predictor ever.

You have beaten steve's rc at 150% edge.

My prediction is mostly poor, have to admit no talent. So I gave up on prediction.

[Joe]

If the trials were dependent the binomial would be a lousy model and its predictions would be poor, but they are very good.

This has to be the most important basis then.

When I say 'predictions' I just mean that the binomial distribution is a good model for the numbers of failures/successes in a sequence of trials. It doesn't mean you can use it to increase predictive accuracy of the next spin.

[Joe]

The only dependence, or more specifically "link", between winning number, is what causes numbers to win.tm This is the physical variables like rotor speed. What else would it be? The number of reds?

Statistics is just past. It doesnt influence the future, but still holds clues. Distinguish between irrelevant and relevant statistics.

There is physical independence (or dependence) and statistical independence (or dependence). Obviously there is no physical dependence between spins, just as between successive flips of a coin. Where there is physical independence there must be statistical independence.

[luckyfella]

When I say 'predictions' I just mean that the binomial distribution is a good model for the numbers of failures/successes in a sequence of trials. It doesn't mean you can use it to increase predictive accuracy of the next spin.

There are 2 posters on this thread who possess god level predicting capablility, steve rc with his 150% edge and OP.

Ask both of them to help you improve your prediction, make sense ?

Or you can ask Kav for his no need for accuracy method.
(I'm here hopefully pick up some secret sauce  )

[redhot]

I tried to show the bankroll against the bets taking place.

Column 1
L
H
L
L

Column 2
3
5
2
2

Basically if you have 1 bet going down column 1 in the long run it will be negative

If you have another bet  in column 2 in the long run it will be negative

In both cases LLN will kick in and give you a negative outcome.  We agree up to this point!

Column 1 has a 3 spin limit for a repeat to come. Column 2 has a 7 spin limit.

So I said if you can use the 3 spin limit from column 1 while betting on column 2. 

Then you have changed the LLN affect on column 2.  If you can always get a win in 3 spins maximum.

So it’s like picking a double street at an EC odds.  But more or less the spins it takes to end.

You increase the accuracy because instead of losing double streets all the way up to 7 spins your winning at 3.  So the LLN affect even tho you can place 1,000,000 bets.  It won’t affect you.

That’s basically what I said and tried to show.  Now I am not saying this exist, especially not in this format.

We are strictly talking LLN affecting this idea and how it neutralizes the affect.

This was the Example I used to make a point.

Thanks MoneyT, your point makes sense now.

So one way to avoid the LLN is to increase the accuracy of your bet so you win more frequently than the expectation. If you're betting double streets and winning at a frequency greater than 1/6, the LLN can't catch you.

The big question is, can this be done? 

Using your example:

Column 1
L
H
L
L

Column 2
3
5
2
2

At spin 4, if betting for the repeat in column 2, you would need to bet on double streets 3, 5 and 2.

However, if for whatever reason we think column 1 is going to repeat on the EC, that means the result will be 'L'

We could then only bet the 'Low' double streets, so 2 and 3. Our accuracy has increased as we've narrowed down the bet selection from 3 double streets down to 2.

However you look at this, it comes down to being able to 'rule out' certain options. If there's 6 possible options and we can find a reason/dependency to rule out 3 of those options, we'll be betting at better odds than expectation.

[MoneyT101]

The big question is, can this be done? 

If you can’t do this particular thing are others similar ideas that can....?!

[Blueprint]

If you can’t do this particular thing are others similar ideas that can....?!

You may want to be able to answer your own questions before you riddle others with them.

[MoneyT101]

I did some reading of redd old material yesterday.

I woke up this morning thinking about some of the stuff and I had some sort of revelation 😱

A new level of understanding on cycles and events that I never saw before.  This sh*t is mind blowing.

I wrote 3 different versions of this post but I keep deleting it cause it’s to cryptic.  At the moment I don’t know how to share this new idea.