There is nothing scary about randomness in roulette.
I have done extensive research on systems that involve totally "random betting".
To my surprise random bets come a lot closer to their probability expectations than other systems that depend on previous results.
For example when I bet 3 random quads (each in its own dozen), on the spur of the moment (and changing them in every spin), I have a very difficult time to pick the wrong ones, for more than 6 spins. (That is if it gets there).
When I follow 3 missing quads it's not unusual to reach to 20 or more spins where they haven't appeared. And rarely even 30 spins.
But NEVER with random picks.
But that requires some experience in randomizing, so that you don't overuse some QUADS and underuse some others.
Some of my best systems involve totally random betting. In every spin.
I allow a few virtual losses, and bet the range from there to the point where the actual loss starts to become very rare. Always guided by empirical research results.
So you can use the same weapon the roulette uses (randomness), to fight back.
You will be very surprised with the results.
I am not clear on what you mean here, Palestis.
How can you make a "totally random bet " which is guided by "empirical research results " ?
Ok. What I mean is this.
By empirical test results in general I mean testing a system (any system), to find out how often I expect to hit the desired target. (after a predetermined trigger). And not only that, empirical research determines the most frequent range of a series of bets that will find its target.
The same goes with random betting.
I make a totally random bet (which changes after every spin), then observe how many spins it takes to find the target. And I do that for thousands of spins, just as we do with every system.
After a lengthy testing period, you come up with an average number of spins that it takes to hit the target. (No different than the method of testing any system). On top of that you determine the most
"frequent range" of bets that finds the target. So you bet that range only, instead of betting the entire range from the beginning to the end. (and hopefully the bank roll will survive the entire series of bets).
For example I make a random bet like R then see what happens. If lost, then I make another random bet like H then see what happens. Then E etc. Until my random selection wins.
In this particular case ( EC bets), I have found that you rarely exceed 5 bets before your random EC selection hits the target. With the most frequent range of hits in spin 2, 3 and 4.
A different random selection (like 3 quads or 3 DS's or 2 DS's) may have a different range of bets where a hit is most likely to occur.
I personally prefer 3 random QUADS, (each in its own dozen). and found that after long empirical research one of them will hit within 7 bets after the random betting starts. with the most frequent winning range that wins is in spin 3,4, 5 and 6 . So I only bets those 4 spins. The first 2 I lose virtually, (I wait till that happens), and then stop after the 6th bet.
If it exceeds the empirical results (which is on the rare side), I don't lose more than I should, because I will have stopped the betting by then.
So basically whatever empirical test results you get form testing a traditional system, you get the same from testing a system that uses random bet selections.
However it seems that random bet selections are less vulnerable to VARIANCE.