Guaranteeing an infinite loss on BV Roulette?

[falkor2k15]

Falkor, just to make sure I understand properly:
when you mention ''repeat'', are you refering to the spin following the closing cycle?  Ex: NCL2: 23, 15, 23, rep, or the spin where it closes, or the previous spin?

When a number has 2 appearances in total then it's a repeat, i.e. above 23 is responsible for the repeat on spin 3 = NCL2.

When the repeat happens then the defining element, i.e. 23, is carried over to the next number cycle (principle A) or you can do a fresh virtual spin (principle B) - doesn't really matter:
23...

Some people seem to think the principle A forms some kind of magical relationship - perhaps akin to fractals - but I haven't noticed any advantage using principle A or B. There is also Principle C:
23, 15, 23 = repeat. Now carry over the 15 and 23:
15, 23...
15, 23, 6...
15, 23, 6, 6 = repeat on Order 3/Position 1.

A lot of terminology to learn! 

[Bigbroben]

I understand now what you mean by order and cycle length, ok.

The question was: you said: on a NCL2, the ECs repeat 92% of times.
Do you mean they repeat after the cycle is closed, on next spin;
Or do they repeat when the cycle closes, i.e. a black nr closes in 2 spins, previous spin was black?

What are your spin reference to call it a repeat? The spin where the cycle closes and the next spin, or the previous?

[falkor2k15]

More stats:

Let's look at NCL3o3:
Dozens repeat 97% of the time.
CL: 50%, 32%, 19%
Order: 68%,27%,5%
Columns also repeat 97% of the time in the same game as the dozens.
CL: 54%, 30%, 16%
Order: 68%,27%,5%

NCL3o1:
Dozens repeat 70% of the time
CL: 38%, 40%, 22%
Order: 79%, 17%, 3%
Columns repeat 71% of the time in different games to the dozen repeats.
CL: 38%, 39%, 23%
Order: 79%, 17%, 3%

[falkor2k15]

I understand now what you mean by order and cycle length, ok.

The question was: you said: on a NCL2, the ECs repeat 92% of times.
Do you mean they repeat after the cycle is closed, on next spin;
Or do they repeat when the cycle closes, i.e. a black nr closes in 2 spins, previous spin was black?

What are your spin reference to call it a repeat? The spin where the cycle closes and the next spin, or the previous?

The exact spin where the number closes the cycle, i.e. the first to 2 appearances, defines the repeat on that one spin alone; and it's this same spin number where we refer to the neighboring EC streams to see if they also repeated at the same time as the numbers (next test: streets).

So the idea is this: if the number cycle is open at, say, spin 4:
16, 23, 15, 33...
If we want to bet for a number repeat on the next spin (would be NCL4) then we can express this by betting an EC or Dozens instead of the 4 numbers, based on the stats covered in this topic (still more to publish if it turns out to be useful).

[falkor2k15]

Street Repeats Cycle Length 3 (SCL3):

Dozens repeat 84% of the time.
CL: 39,33,28%
O: 68,28,4%

Columns repeat 53% of the time.
CL: 34,45,22%
O: 63,30,7%

Splendid, splendid... that's the result we wanted! Very good progress after all these years of not knowing... I'm happy. 

[falkor2k15]

Looks like edge to me...

EDIT: failed! Back to the drawing board... 

[Bigbroben]

Looks like edge to me...

EDIT: failed! Back to the drawing board... 

Do you  mean: you found no way of exploiting these stats or you found an error in your  file?

[Bigbroben]

Falkor,
I am running a macro right now with Excel.

First thing sorted is what you would call the number cycle.  Numbers come out and a win is recorded when a nr repeats.  Then restart with the won nr and proceed again.
Then, on a win, I record the dozen in which the repeating nr is, and the dozen on the previous spin.
As opposed to be about 33% same dozen and 66% different dozen, I get 43% of a repeating dozen when a nr repeats.

I get 44,7% order1 (for me it is the newest, not the oldest), 34,5% order2 and 20,8% order3.

I suspect the difference between expected 33% and 43% lies in the amount of nrs per dozen: a more frequently hit dozen gathers more nrs and is more often at Order1.

I've also checked the sequences of SameDoz-DiffDoz when a nr repeats.  Turns out that when a nr repeats with a different dozen than previous spin, most of the time it happens only once, 62%.
Below one can see easily the difference in the slopes: when a nr repeats on the same dozen as the previous spin, it tends to do it consecutively, as opposed to when a nr repeats on a different dozen than previous spin.


That is, let's say you'd play the nrs that appeared so far, but only the ones that were in the previous dozen.  If playing all ''number cycles'', as you would put it, it would be a 43-44% win,  56-57% loss.  With the most populated dozen hitting more frequently, it is not enough to overcome HE.
If 62% of ''losses'' only appear once then switch back to ''win'', would it make sense to ''shed'' a loss then hope for a hit in a repeating dozen?  Maybe go for a run of consecutive hits on same dozen?

Max consecutive DiffDoz is 11, max consecutive SameDoz is 21, average maximums in 1000-spin tests, after 1700 tests, is 4 for the DiffDoz, 7 for SameDoz.  So indeed SameDozen tends to run.

That's what I got so far with my beloved Excel Random.

[Bigbroben]

Average of 2.37 nrs in the dozen when it's a different dozen;
Average of 2.81 nrs in the dozen when it's the same dozen.

[Bigbroben]

I've also checked the sequences of SameDoz-DiffDoz when a nr repeats.  Turns out that when a nr repeats with a different dozen than previous spin, most of the time it happens only once, 62%.
Below one can see easily the difference in the slopes: when a nr repeats on the same dozen as the previous spin, it tends to do it consecutively, as opposed to when a nr repeats on a different dozen than previous spin.

Falkor, I got it wrong.  Consecutive Same and Consecutive Different should be switched.  Just a tiny typing mistake...