Turner's spins (file attached under plot) -


[attachimg=1]

I will ask a question which will probably, once and for all times, show my ignorance!! 

How does one corrupt a spin file?  I've ran a few millions spins through various bots, and if the file was anything but numbers, the bot stopped.  By "corrupt" are you guys meaning numbers have been added to cause the thing to win or lose?

Here is corrupt to me:

34
2
Unicorn
5


Sam

QuoteBy "corrupt" are you guys meaning numbers have been added to cause the thing to win or lose?

That would be my view Sam, maybe "corrupt" in this instance means "manipulated".

As sputnik put it... The graph shows good days....bad days and average days
Having said that.... I wouldn't of minded being on table 2 in June and over christmas

Quote from: Number Six on February 20, 2014, 10:34:35 AM
Either way, you're correct I think the testing is become too divided and slapdash. I do think "official" tests should be done with spins from a verified source.

One of the basic rules to follow when you're doing any kind of statistical inference is to get data from as many and varied sources as possible, otherwise you run the risk of bias. Since XXVV is adamant that WF3 does not depend on any particular wheel, but only on "natural" cycles (although it's still not clear to me why he excludes RNG), it seems appropriate to apply this methodology in full. Multiple verified sources would be ideal. If you're going to get ALL the spins from Weisbaden, you should at least make sure they're taken from different tables and over multiple time spans.

Quote from: esoito on February 20, 2014, 05:21:11 AM
So, just to clarify, you're basically suggesting XXVV is a liar, are you?

In fact, Albalaha, why don't you come right out with it and actually call  him a liar?

XXVV isn't a liar, necessarily. He could just be mistaken. You've seen the plots: there are long stretches of good results, that's why you have to test over many spins to get the true picture. Having seen it, you can still decide to go ahead and play the system (or use it as a base for any tweaks and modifications). But would you rather know or not know the long-term results? I know which I prefer.

XXVV,

You haven't addressed any of my points about the spins from roulette research. The question remains: why are those spins unacceptable and "rogue"? And why are the Weisbaden spins deemed "Approved"?

Can we please stay on topic. In what sense is my testing flawed, in your opinion? Please be specific and suggest how you think it should be done, and why.


You talk about how the bet can be "tuned" and modified. This is not the place for that; my concern is to test WF3 specifically and present the results for that system and that system only.

Quote from: TwoCatSam on February 20, 2014, 02:48:34 PM
I will ask a question which will probably, once and for all times, show my ignorance!! 

How does one corrupt a spin file?  I've ran a few millions spins through various bots, and if the file was anything but numbers, the bot stopped.  By "corrupt" are you guys meaning numbers have been added to cause the thing to win or lose?

Here is corrupt to me:

34
2
Unicorn
5

Sam

Sam, spins can be corrupted in various ways. For example, the reader boards in casinos can misread, omit, or duplicate spins. Also if spins are hand written mistakes and duplications can occur. I remember seeing a file some time ago in which a long sequence of spins was repeated exactly (the odds against such a sequence would be trillions to one). I'm not suggesting that spins are necessarily deliberately added or removed.

Bayes.
Taking my spins and method of extraction...do you feel they are more reliable or fall foul of the same issues. (obviously the hand written issue is elliminated)

XXVV.
what is your view on the result from Bayes test using my numbers?

Quote from: Turner on February 21, 2014, 09:34:05 AM
Bayes.
Taking my spins and method of extraction...do you feel they are more reliable or fall foul of the same issues. (obviously the hand written issue is elliminated)

Turner, it seems to me that your method of extraction is reliable. In any case the spins can be checked against the existing database at Weisbaden. I think the Weisbaden spins themselves can be trusted.

XXVV,

QuoteFrom the outset I have requested live spins. You know that

Yes, and what makes you think that they aren't? I told you I had a reply from the owner confirming that the spins were taken from a real wheel under licensed gaming conditions.

By the way, I had a heads up from Xander telling me that the owner of the site is probably the chief wheel engineer for TCS Huxley George Melas (and I can confirm it because that's the name he used in his email reply to me).

http://www.worldgameprotection.com/speakers/george-melas.html

I don't think there should remain any doubts as to the reliability of those spins.

You seem to want everything on your terms. You don't trust the spins, and yet we're supposed to take on trust your Macao data, which isn't even in the public domain. A clear case of double standards, don't you think?

And I note you've ignored my point above that you were the one who emphasized that at least 20,000 spins should be tested, and yet appear now to have taken a complete u-turn on that issue. 

Nevertheless, I'll continue to test using the Weisbaden spins which members are helpfully providing.

we can try Weisbaden spins witch i prepared in other theard (40k spins is enough to see how it going) if it show some promise i can continue.

Table 3, Weisbaden 01/02/03 2013. Thanks to Biagle for providing the spins.


[attachimg=1]

[attachimg=1]