Number Six - this post is for you ...

[NS]

Gambler's fallacy.

[I have cookies]

Word  and prove me wrong that corelation of correction does not exist 

[NS]

Betting on these "corrections" or whatever doesn't work. I really don't know why you keep going back to this over and over again. You should know better.

Ah well.

[Bayes]

I tried this some time ago and, like mistarlupo, didn't get any better results than expectation. But, this time around, results are better, possibly because I'm waiting for the "correction" to begin before placing a bet. I'm not coming to any firm conclusions yet but I can confirm that I'm getting about the same results as IHC (2.5 - 3.0 SD).

Rather than gambler's fallacy, I would suggest reversion to the mean, although you shouldn't be able to make a flat bet profit in the long-run merely from this.

[Mr J]

Gamblers fallacy NS? So much for AP (cough), thats off the list.    Ken

[Number Six]

although you shouldn't be able to make a flat bet profit in the long-run merely from this.

Well, I agree with that. My findings from this type of bet are varied. Firstly, yes, this "correction" certainly does exist but by using only the SD as a benchmark the results seem random. Also, a bet can remain in negative fluctuation for hundreds and sometimes thousands of spins. This throws a monkey wrench in the works. It either means a lot of waiting around or a lot of bets lost and fundamentally makes it no different to any other system.

What I have found with the SD is that it moves in exactly the same way as the event itself. This seems obvious but it's not that clear cut. What I mean is, for example, in a measured sample you'll get 100 instances of 1.0 SD, 50 instances of 2.0 SD, 25 instances of 3.0 SD, 12 instances of 3.0+ SD etc., so there's really no difference between recording the SDs of an event and simply counting the event itself. So if you waited for 3.0 SD before an attack, half the time the SD will decrease, the other half it will go up. The upshot is that there is no advantage in using the SD in such an organised way.

I think you would need some secondary reference as an extra benchmark, some sort of criteria around which to base the attack. Again it seems obvious, but the way forward would be to lower the number of attacks needed to gain a profit, and that profit should be adequate to cover losses. For example, winning one unit per attack is not feasible since the hit rate will be 50%. You would have to ride the wave for as long as possible, as long as it's going in the right direction. Again, though, you're faced with certain impracticalities. You can't stand around and play thousands of continuous spins.  

[I have cookies]

NS you can not blame me for having fun and waste my time with 2.5 and 3.0 STD as result after every samepl of 10 000 is just plain and simpel fun - don´t you agree.
4 attempts with 700 Eruo wish becomes 1100 Euro and 5000 to 7000 Euro in 4 weeks play.

[NS]

Sure, it's fine to play with some ideas for fun, I'm not saying you shouldn't do that.

I'm basically just saying that this is the gambler's fallacy, and whatever your criteria is for betting on Red/Black, it's the same as betting randomly. No, I won't prove you wrong, you prove yourself wrong. I just hope it won't happen to you while you're at the casino betting €100 chips. I also just hope that you don't really think that 2.5-3.0 SD after 10k bets is statistically significant, it's weak and can/does happen by chance.

Give me the SD you got after 50k flat bets, and then we'll see if you have something.

[I have cookies]

One sampel of 10 000 hahaha come on NS I have run so many sampels of 10 000 I lost count  

And correlation of correction is not random as it has to happen after 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 no matter what you say my dear friend and you can not change that or sience   is part of nature  

You can even observe a std of 100 000 it does not matter as corelation of correction will apper no matter you like it or not ...

To capture corelation of correction is to argue and say fallacy and its random.
But to claim that you would get all reds for one year at the same tabe 8 hours a day with out black balance out to certan degree i would not agree.
It does not have to become even and it all has it up and downs and red can be ahead for a long time or for ever with out black ever cath back to even or get overrepresented.

[NS]

haha ok.

I really don't feel like explaining to you why betting on "corrections" can't give you the edge and is gambler's fallacy (and by the way, I'm not the one defying science and nature here). The fact that you don't know already why it doesn't work from a betting perspective says it all and goes against your statement that:

It is not many who can measruing there knowledge with or against me.

Also where did I say that 1 year of Reds is normal? The chances of that happening are next to nonexistent. Can you use these statistical facts to get an edge? No you can't.

As for the test, are these samples of 10k bets or spins? (I sure hope it's the former)

Also what's the SD of all of the results together? That's what matters, not the small separate SD's of separate samples.

[Number Six]

Also what's the SD of all of the results together? That's what matters, not the small separate SD's of separate samples.

I don't agree. Technically no system can ever be categorically proven to be effective since you can't test it enough. It has to remain consistent with reality. Beating 10 billion bets would be nice but is unecessary and pointless. Even 100k or 50k bets is pointless, because if it's the real deal it would become apparent long before you reached that milestone. Here I'm talking about flat betting, strictly. If you ran four or five samples and the bet selection reached 3.0 SDs in each one, then it's pretty safe to say that is the real deal, a proven long-term winner. This could happen in the space of only a few thousand bets, thus a major simulation of a flat bet is seldom required. Of course, the test has to be valid, you can't be changing the goal posts from sample to sample, which is what most people do.

[NS]

It would still be better to know the SD of the overall test (no separate samples). Let's say you have 14 SD after 50k EC flat bets. Well obviously in this case, you got a holy grail. But let's say you play up to 10k bets and stop whenever you reached 2.5-3.0 SD, and repeat the same with different samples. In this case you can't be as certain because 2.5-3.0 SD can and does happen (especially if he's using no-zero actuals which he probably is).

FIRST get the overall SD to know your chance of randomness. THEN you can divide your test into smaller samples to see if it's consistent (though if the overall SD is high enough, there is no need to do that because it can't get that high without being consistent).

So again, let's see the SD after at least 50k placed flat bets. Until then, this is all déjà vu BS.

[Mr J]

"Give me the SD you got after 50k flat bets, and then we'll see if you have something" >>> Guys like you crack me up NS. You'll rip methods and thats fine, I dont mind but if something is SHOWN to you to be 'good', then you're on board.  Choose one or the other NS, either rip methods  or dont rip them but you cant have it BOTH ways.  Ken

[I have cookies]

Well lets have fun   as the casino has no house edge against this way of play - any comment - I feel there is love in the air   well here is a red rose to you all that belive observation of corelation of correction is a fallacy   as we all know the future hand it over for us as we only have to observe and see how it unfold it self  

[Mr J]

The thing is, certain people call anything they DON'T PLAY, gamblers fallacy. AP (cough) is just as, if not more of a gamblers fallacy, good luck with that.  Ken

[Mr J]

-