Randomer Thoughts

[psimoes]

It might be that the spins where actual bets were placed are less than 10000. Spins waiting for triggers after a loss must not be added to the stats.

[falkor2k15]

Well, since each dozen has always 12/37 chances of hitting, no matter what the bet selection is, the maximum number of repeats found in 10000 spins for a dozen will be almost equal to the maximum number of losses for any double dozens bet selection, plus the zeros. Any dozen is certain to repeat itself much more than five times in 10000 spins...

I had another look (see attachment above) and there were 6 losses in a row - not including zero. And yes - it's 10,000 spins but not all spins had placed bets due to waiting out lost cycles.

So what are you suggesting, psimoes: that there is something exploitable from the distribution pattern of betting the last 2 dominant whose stats resemble any 2 dozen bet of 66%? So we should "triple up" 6+ times to recoup losses?

[psimoes]

Now six losses in a row, that´s too much. Tripling up 6+ times would be impossible I believe. Table limits. A mild progression such as +1/-1 might be worth trying. Problem is something like LLLLLLWLLLL KILLS IT.

It´s still a small number of losses in a row compared to say betting dz1&2 on a rolling basis. If the LLLs stay low after a much larger sample there´s something worth of your time. If not, forget it.

[falkor2k15]

Now six losses in a row, that´s too much. Tripling up 6+ times would be impossible I believe. Table limits. A mild progression such as +1/-1 might be worth trying. Problem is something like LLLLLLWLLLL KILLS IT.

It´s still a small number of losses in a row compared to say betting dz1&2 on a rolling basis. If the LLLs stay low after a much larger sample there´s something worth of your time. If not, forget it.

Exactly. I'm sure Manrique would have found his way through this dispersion somehow - but we are not here to test dispersion killing on 2 dozen bets that result in 66% - we are here to try and find edge with Priyanka's Principle A Cyclic framework! And the last the 2 dominant dozens seems like a red herring...

[psimoes]

Betting on the last two dominant dozens was just a playful exercise. Nothing in the "Cycles Theory" hints at it. We could bet for the last dominant single dozen as well, with the expected 33% wins vs 66% losses. Then add the zeros to worry about.
One major flaw with betting last two dominants. or second last two, or last and third last or whatever, is that it expects losses to occur, since nothing stays the same forever.
This static bet selection doesn´t respond dynamically to Change.

We could watch out for the "cycles of dominant dozens" or something like that instead.

Example:

Consider the following outcome 232112331323211231

Start large cycle

232 small cycle completed - dz2 dominant
                                                                             
11 small cycle completed - dz1 dominant                                                                     
                                                                                                                                       
233 small cycle completed - dz3 dominant                                                                       
                                                                                                                                       
1323 small cycle completed - dz3 dominant again - end of large cycle.

After dominants 1233, dz3 predominates                                                                               

211 small cycle completed - dz1 dominant

231 small cycle completed - dz1 dominant again - end of large cycle.

After dominants 11, dz1 predominates

Now we´ve had two larger cycles where dozens 3 and 1 predominated. Sticking with the same "bet last two dominant dozens" criteria, we´re now ready to bet dz3 and dz1 on a rolling basis until, of course a new predominant dozen appears. It just gets complicated. Knowing the fractal nature of Chaos, I suspect the result will be 66% wins against 33% losses...

[falkor2k15]

Betting on the last two dominant dozens was just a playful exercise. Nothing in the "Cycles Theory" hints at it. We could bet for the last dominant single dozen as well, with the expected 33% wins vs 66% losses. Then add the zeros to worry about.
One major flaw with betting last two dominants. or second last two, or last and third last or whatever, is that it expects losses to occur, since nothing stays the same forever.
This static bet selection doesn´t respond dynamically to Change.

We could watch out for the "cycles of dominant dozens" or something like that instead.

Example:

Consider the following outcome 232112331323211231

Start large cycle

232 small cycle completed - dz2 dominant
                                                                             
11 small cycle completed - dz1 dominant                                                                     
                                                                                                                                       
233 small cycle completed - dz3 dominant                                                                       
                                                                                                                                       
1323 small cycle completed - dz3 dominant again - end of large cycle.

After dominants 1233, dz3 predominates                                                                               

211 small cycle completed - dz1 dominant

231 small cycle completed - dz1 dominant again - end of large cycle.

After dominants 11, dz1 predominates

Now we´ve had two larger cycles where dozens 3 and 1 predominated. Sticking with the same "bet last two dominant dozens" criteria, we´re now ready to bet dz3 and dz1 on a rolling basis until, of course a new predominant dozen appears. It just gets complicated. Knowing the fractal nature of Chaos, I suspect the result will be 66% wins against 33% losses...

That's the "Secondary" or "Outer" Cycles framework that I tested, albeit preliminary, over at the other Random Thoughts topic, and I've maintained it in the newest version of my simulator - the "arithmetic parallel edition"!  When I tested it on Quads, the dominant dozen went from 55% to 65% - but that's for one large cycle of smaller cycles (to use your terminology). Again, that larger ratio could be artificial without knowing the cycle length since the cycle length is a specific outcome based on Non-Random, but the defining dozen is more a Random based constant that is describing an attribute for any one of those cycle length events - and we don't know which one it will be. It's a bit like VdW... we don't know which type of APs will make up the final 50/50 on ECs - 1,2,3 or 2,4,6 (etc.) - but we know in the end that it will be 50/50. So it's not useful to try to pin down only 1,2,3s and ignore the 3,5,7s since the number of occurrences for each will be different every time - but together they work towards the constant ratio. The defining element is the same problem as that - we can't pinpoint which cycle length the defining dozen is going to be applied to so we always end up with 66% instead of edge. With cycle lengths we are on much firmer ground - and I'm thinking that this foundation needs to be laid before we try to create a biased game out of the defining element. The only way to improve VdW on R/B was to use dispersion killing - but this process is secondary to gaining edge. As for playing the last dominant dozens within the Large cycle framework I think a different approach may be needed: it would most likely entail something along the lines of what Priyanka was describing with the "Dozen Triplets" before she introduced cycles proper. I can test your basic play on the small cycles taking into account previous results of larger cycles, but this unsophisticated method I fear may just result in the 66% again.

[falkor2k15]

I've now got this up and running, but it's not using advanced tracking yet:

Scarface "I wonder if we take it a step further, could we find some sort of edge.  Maybe, always bet the last 2 dominant, or repeating dozens.  But only play the most recent hit 3 streets from each one (total of 6 streets).   Seems like a good way to catch hot sections being hit."
link:://:.rouletteforum.cc/index.php?topic=15938.330 (page 23)



I doubt it will result in any edge...

[falkor2k15]

50/50... no edge...



Should I bother with this one?

Scarface "What if we bet the last 3 hit lines in the dozen, instead of the whole dozen.  If a cycle ends with dozen 2 as dominant, bet the last 3 hit lines in dozen 2 on the first bet.  If dozen 1 hits next, bet the last 3 hit lines in dozen 1 and 2. Seems like there is always 1 line in a dozen that stays cold.  Thought this might increase the odds of a hit better."
link:://:.rouletteforum.cc/index.php?topic=15938.495 (page 34)

[falkor2k15]

Remember: the penultimate dominant quad/dozen did take effect when we were tracking for multiple quad repeats - but for cycles of 1 repeat only it's useless. Also, jumping from dozens to streets parachutes a cheap progression.

[falkor2k15]

For sure it seems we have to get edge on the cycle length for all this other defining/dominant stuff to start working otherwise it's going to remain inactive...

[falkor2k15]

I'm running a few tests... without a shadow of a doubt this is the holy grail. I keep getting the same results across all data sets. Before I proceed to devising a final strategy and checking profit over 1 million spins I am trying to figure out if my interpretation is flawed in some way. Am I missing something obvious? Why is this game suddenly so predictable? It can't be that easy to beat... surely. Mission complete?

[praline]

arithmetic parallel ?

[falkor2k15]

arithmetic parallel ?

[Tomla021]

maybe the Falkor got the answer-congrats if you did

[falkor2k15]

Tom, I dropped you an email with an invite... password is 123ADDED2123246 see you on the other side...