Bayes has been one of the most intelligent and rational guy arounds forum and I have heard lesser irrational things from him than anybody else. What he is suggesting is to put 3 units as 1 on the coldest doublestreet, 1 on hottest doublestreet and 1 on average. There are lots of trouble deciding them. At times, there will be two hottest or coldest or average doublestreets together. So, it is easier said than done. There can be cleaner parameters like farthest hitting line, the latest hitting line and one in the median leaving the last line i.e. 3rd in nearness. That can be played with decisive pick.
All these won't seriously help in any manner, in my humble opinion. I have worked on these for pretty long time and found that no way of picking bet is superior by itself than all others. It is impossible. Hence, although it could be interesting to pick three doublestreets instead of RED, it won't matter in short or long run and it can not reduce variance by even an iota. Trying to find a better bet in a purely random game is a joke. We better make strategies to survive the worst cases(whenever they come in our bet) and how to win if there are average wins after a horrible drag.
Looking at the complex mathematics, that Bayes has produced on this forum, together with being directed
to even more complex mathematics, it appears he is a very competent mathematician. If I was a maths
student, I would of course be thrilled. I don't see how it helps a roulette player.
Any suggestion using hot & cold double streets (presumably lines) or any hot, cold numbers is a non
starter. I mentioned on a previous occasion, the casino's are only too happy to provide such info on
their indicator boards and they would not do that, if they thought it would give a player an advantage.
It is a pity that mathematicians don't spend some of their time on random numbers, rather than for
example their latest amazing discoveries, that there are two types of infinities !!