Yes, that’s very true. The real scientific way of doing it, which the statistical results actually assumes is employed for their traditional interpretations to be genuinly valid. Doing it the other way around is scornfully dismissed as “data mining” by REAL staticians. But mathematical perfectionism isn’t always all that practical either.

I didn`t fabricate the sheet myself, so i gotta say i haven`t looked closer in the cell formula for the Chi. I can send you the cell formula for the Chi Square if its any help.Thank you, but there is no need. I just use the chi2test()-function which is built-in in Excel. The difference is probably just about using the money versus raw pocket numbers.

PS: Yes you can get the numbers in most larger european casinos, but its far better to collect them by yourself because you can then seperate them in clock and anticlock spins.Ah, yes. But maybe it is possible for one person to keep track of spin direction (and maybe ball change) on several wheels at the same time in a casino, if winning numbers are taken from official lists.

Anyway, I’m primarily interested in analysis of data in order to find drop zones and scatter distribution for use together with predictions made while the ball is spinning. That’s hard enough…

It is disheartening to realize how large samples of data are needed in order to achieve good statistical significance. And add to that how small fraction of the bankroll one must bet in order to avoid Risk of Ruin, and the initial dream om getting rich quick on predicting roulette suddenly seems to be harder than one would have thought from the beginning. Even with an edge over the casino, it still helps alot to be lucky too!