One of my favourite bets is the hidden or secret  dozen.
This is occurs when 25 numbers have been hit and just 12 remain, it usually occurs after around 40 spins and is often considered a pretty reliable bet.

However, after recent discussions on Even Chance bets I noticed a similar phenomenon* on an even chances and have modified RouletteKeyGold accordingly.

At some stage in virtually every Roulette session there is a time when just 18 numbers remain unhit.

We've all probably seen the normal even chances sleep for over 20 spins but this set of 18 random numbers never seem to sleep for less !
So when only 18 numbers remain I believe they won't remain for long and so are probably worth betting on.
I'd even consider a Martingale progression! As in the second video below

I'd be interested to hear from anyone just how long 18 numbers have slept in your own play  and whether it was live or RNG
My record so far is 32 spins.
An explanation and two demos of cash play are below.
Any feedback will be much appreciated (I think !)







(phenomenon* I won this threw my spell chequer witch says its awl wright.)

Hi!

As you well know, the fact that 12 silent numbers can be sleeping for a lot more than normal dozens or columns do is just a reflex of the innumerable possible combinations forming clusters of 12 numbers.
The table layout has only six possible combinations of such possibilities.

I recordered a sleeping dozen created by 12 silent numbers that remained silent for 92 spins (I tried your same methodology several years ago taking a Marigny De Grilleau suggestion).

Don't know about 18 numbers, I think you can expect to get them silent for 50 or more spins.

More interesting is the study of the exact position of dozens or, better, lines (6 numbers) coming along the way in order to build a dynamic new even chance.
You'll have to wait the situations when the positions in the last three spins are different, then betting the remaining three positions. This is your trigger.
Since you won't bet when positions are repeating, you are excluding a large part of the random world and at the same time you'll be more confident that sooner or later at least one of the three remaining positions you bet will be filled.
In a word, you shift the problem from "what" (line) to "when" (position) and "how" (actual distribution).

Cheers
   
as.




   





   

I usually see 12 sleepers occur between 37 & 50 spins.
The equivalent for 18 numbers would be 24 to 34 spins

I often bet on these as they frequently get hit within a few spins
Admittedly I have seen 12 sleep for 50 spins but this is quite rare.

In addition to this one off bet on 18 sleepers RKG watches the six usual (HL,RB,OE)
combinations on the table layout and 37 combinations on the wheel and 20+ combinations
of 3 table sixains. So I'm watching over 70 even chances that may occur occasionally
and repeatedly but this one set of 18 sleepers only occurs once in every game.

I believe the best way to explain is by example so I've just played Roulette Pro at William Hill.
I played 42 spins in about 5 minutes stopping after the 'Hidden' dozen.
During the demo I won £14.00!

Here are the numbers played 17,3,33,8,1,12,7,0,16,27,32,15,5,15,36,29,21,28,5,13,34,8,3,23,5,26,14,15,28,26,20,9,34,29,16,33,9,28,24,0,27,4

In this session the maximum normal even chances missed are Black-6, Even-3, High 5.
However, RKG identifies 1 even chance that slept 21 times and several that slept 8 or more times.

There's a general theorical problem when betting a "sleepers" oriented strategy: you'll never know whether the given results are real fluctuations or produced by a biased wheel.
In the doubt I'd prefer to either not to bet at all or to wager what happened instead of what it didn't happen so far.
Moreover, a sleeping numbers cluster could easily contain one or more number to sleep several hundreds of spins. And just for normal statistical reasons. So I'm betting maybe 16 or 17 numbers vs a larger pool than a 18/37 ratio dictates.
Of course a smaller numbers ratio could get an empirical slight higher probability to appear and it seems you have found such situations to be more probable than not. 

 
as. 

Whether the results are real fluctuations or produced by a biased wheel it doesn't matter as the 18 numbers selected differ in every game.

In every session there is a time when there are 18 sleepers and 12 sleepers and these always get hit and with sufficient capital can be the source of profitable bets.

I don't only advocate a sleepers" oriented strategy but at the present time it seems as good as any

It doesn't matter if one or two of the sleepers go on to sleep hundreds of times the trick is to be aware of all of them as I'm not betting on 1 or two but twelve or eighteen numbers.

To bet on 'what has happened' there are also always numbers that repeat. Out of 37 or so spins there are always numbers that repeat (usually 25 or or so) so it can be worth betting on a repeater and sleepers simultaeneously.

For example in the video below after 13 spins hoping for a repeat of the penultimate spin  I play the *Rational system  which after 32 spins shows a profit of £10.00 (its low point is a loss of £26.00)
After 26 spins I identify 18 sleepers these win on the 28th spin

I continue to play the rational system hoping for a repeat but simultaneously play sleeping *finals and at 45 spins a sleeping dozen

So I'm not betting only on what hasn't happened so far but also on something that has happened.


*http://winningrouletteplayer.com/systems.html