Roulette Tool Edge Cases

[ati]

No, because these are unrelated sequences and there can be a huge variances in random. Sure, things are likely to even out in the long run (regression towards the mean, or RTM), but the short term variance remains. I'm sure there are hundreds of RTM threads on the forums, where people tried various things to go against the variance.

[Joe]

For argument sake lets say you tested 100, I know that is small, 100 million would be better, but anyway. The results show that only one time, red came up 15 times. So during play, red comes up 14 times and you will play the next spin.

Math says that red has an18/38 chance, but it also says it has a 1 in a hundred chance to happen.

The two probabilities are not inconsistent and don't contradict each other because they are referring to different distributions, and that's what people fail to understand. As players we can only ever bet on the next spin, never on a sequence of spins, and for any bet on the table the distribution of successive probabilities of the next spin is always uniform. link:s://en.wikipedia.org/wiki/Discrete_uniform_distribution

That is the simplest distribution and all other distributions like the binomial or geometric depend on and use the basic probability of the next spin! eg you can't figure out that you have a 99% chance of getting at least one hit in X spins unless you know the probability of winning the next spin. The basic probabilities dictate the results for all more complicated calculations.

What you and many others are trying to do is work back from these more complex probabilities (like eg the law of the third) and come to the conclusion that they affect the probability of the next spin, ie affect the probability that was used to infer the complex probability! Do you realize how insane that is? It's completely backwards. 

And more to the point, it explains why your systems based on this mistake can never work. If you understand what I've written above, you'll understand why.