Random Thoughts

[RayManZ]

Ok guys. We know we need a parallel game right? We can use the positions for that:


For this example is use the quads:

Code Select Expand


number quad cycle define positions position cycle define
17 2 2 1 3 4 2
19 3 3 2 1 4 3
2 1 1 3 2 4 3 2 3
30 4 4 1 3 2 4
9 1 4 1 1 4 3 2 2
32 4 4 1 3 2 2 3 2
14 2 2 4 1 3 4
1 1 3 1 1 2 4 3 3
30 4 4 1 2 3 3 3 3
21 3 3 4 1 2 4
5 1 3 1 1 3 4 2 3 2 3
28 4 4 1 3 2 3 1 3
28 4 2 4 4 1 3 2 1
35 4 1 4 4 1 3 2 1 2 1
34 4 1 4 4 1 3 2 1 1 1
32 4 1 4 4 1 3 2 1 1 1
30 4 1 4 4 1 3 2 1 1 1
26 3 3 4 1 2 3
6 1 1 3 4 2 3 2 3
18 2 2 1 3 4 4
24 3 4 3 3 2 1 4 3 2 3
34 4 4 3 2 1 4


Now who can help me get this to something that is winning. I'm still missing some part of the puzzle.

[falkor2k15]

Below is the perfect example how two losing games played together can be turned into a winning game.

I think that's a bad example because it's not directly applicable to Roulette, right?

Cycles and statistical relations can help avoid the variance for our bet  selection, and can increase the odds.

Cycles have constant ratios, but they still seem to suffer from variance. My datasets always come out different even when based on the event of a cycle closure instead of spins. Increasing the odds... you must be referring to stitching bets to increase payout odds for an unofficial event? But what about increasing predictability? It does that too, right?

It was proven that the odds for each individual bets are the same, but we should not play individual spins. We must think in group of numbers.

Groups or Events? Is there a concept based around groups? By groups you mean events that are not equally likely? But then they do happen to be equal in terms of cost vs. payout. A 2 dozen bet will win more than a single dozen bet (unequal), but the payouts mean they break even (or lose to the house edge) = equal.

I think that's the key. If we mean the same thing by parallel.
Priyanka has emphasized many times that 2 or more games need to played, because each individual game on its own is a loser.

Yeah, and I think there exists many different ways of playing parallel games that could confuse the meaning - some more difficult that others. The reason to avoid VdW and if possible, parallel streams, is to keep any method of obtaining edge as simple as possible to play in a real casino with less chance of making mistakes. I understand parallel games could be as simple as observational spins for 1 game, such as tracking multiple repeats (= less dozens), and then the 2nd parallel stream/game is the one we bet. Or we could be tracking 2 streams and playing both of them in an alternating fashion - or perhaps stitching a bet across both streams simultaneously or what not... But does all this mean that we cannot find edge using just one stream alone..? If so then why? Let me give an example of how we might obtain edge with a single stream game...

We play inner cycles wrapped by outer cycles. 121 11 1322 22 = outer cycle completed. The outer cycle is dependent on the inner cycle, and it could be described as a parallel game? But really we are just using a single stream of dozens, inner cycles, outer cycles, albeit with the "repeat types" flagged at different points along the stream based on relationships? So this could be described as dependency without a parallel stream. And we would always be betting in terms of repeats - not individual spins. We wait for an outer cycle to be defined by, say, inner CL2. That then becomes our entry point and we now have a bunch of constants to create a biased game: Outer Defining Dozen, Outer Defining Cycle Length, and Inner Defining Dozen (Pri said we just need a couple of constants if that). Outer is dependent on Inner in terms of defining cycle length, and the chances of the next outer cycle being defined the same are as follows:

2 > 1	269	16%
2 > 2	1299	76%
2 > 3	136	8%
	1704
1 > 1	524	61%
1 > 2	263	31%
1 > 3	73	8%
	860
3 > 1	68	19%
3 > 2	141	40%
3 > 3	145	41%
	354

Since we waited for outer to be defined by CL2, we've attained an entry point with the highest possible ratio of 76% for the next outer to be defined by another CL2 (same), which exceeds the probability ratio of a double dozen bet of 66% or an inner CL2 @ 44%. So we proceed to stitching the next outer cycle based on being defined by CL2 with other inner cycles being defined by the same dozen to reduce losses. The third constant - the all-powerful Outer Defining Dozen - could also play a part in the stitching.

So please confirm...
1) Is the parallel part indeed missing?
2) Is there any other concepts missing or are all the required ones there?
3) Could it gain edge providing the aforementioned concepts are played correctly?

[falkor2k15]

Ok guys. We know we need a parallel game right? We can use the positions for that:


For this example is use the quads:

Code Select Expand


number quad cycle define positions position cycle define
17 2 2 1 3 4 2
19 3 3 2 1 4 3
2 1 1 3 2 4 3 2 3
30 4 4 1 3 2 4
9 1 4 1 1 4 3 2 2
32 4 4 1 3 2 2 3 2
14 2 2 4 1 3 4
1 1 3 1 1 2 4 3 3
30 4 4 1 2 3 3 3 3
21 3 3 4 1 2 4
5 1 3 1 1 3 4 2 3 2 3
28 4 4 1 3 2 3 1 3
28 4 2 4 4 1 3 2 1
35 4 1 4 4 1 3 2 1 2 1
34 4 1 4 4 1 3 2 1 1 1
32 4 1 4 4 1 3 2 1 1 1
30 4 1 4 4 1 3 2 1 1 1
26 3 3 4 1 2 3
6 1 1 3 4 2 3 2 3
18 2 2 1 3 4 4
24 3 4 3 3 2 1 4 3 2 3
34 4 4 3 2 1 4


Now who can help me get this to something that is winning. I'm still missing some part of the puzzle.

Please could we hold off for a bit till ati - and ideally Priyanka - have confirmed that a parallel game is absolutely necessary for obtaining edge (see my example above)? Or is it another case of VdW: it helps as an additional method to complete the Non-Random framework, but the other concepts can work without them?

[RayManZ]

Please could we hold off for a bit till ati - and ideally Priyanka - have confirmed that a parallel game is absolutely necessary for obtaining edge (see my example above)? Or is it another case of VdW: it helps as an additional method to complete the Non-Random framework, but the other concepts can work without them?

Only pri can confirm anything. I do know that you're looking in the wrong direction. Innner cycles? outer cycles? Pri said we should look at positions. That is a part of the puzzle. You have posted alot of information but i did not find anything usefull.

Pri has confirmed we should look at positions and use cycles. I'm trying to get this thread back on the right track. Something you did not do at all. The only thing all your data has confirmed is that there is a house egde.

Also i can't but that complex at all. How the hell can you play it if you use all kind a concepts together? It not posible. Think logical and simple.

[falkor2k15]

Thanks for the constructive criticism.

Positions - meaning parachuting dozens, lines, numbers? But many of Pri's videos only feature dozens or only feature quads - and they won when designed to win and lost when designed to lose.

How the hell can you play it if you use all kind a concepts together? It not posible. Think logical and simple.

I'm sure Pri said we should use most of the concepts - presumably that's why she went through so many? I doubt edge could be achieved with just one or two concepts, so what are the minimal conceptual requirements and can we at least miss out VdW and parallel, but still gain edge?

[falkor2k15]

Oh I understand what RayManPS meant by "position", which previously meant type of official bet selection such as quad, dozen, line... he means how far back was the last dozen, cycle length, etc. This is another type of parallel game that we should ideally avoid as overly complex unless it's absolutely necessary to include them - TBC.

[RayManZ]

I shall share one final string in this whole piece and my way is not as complex as Russian dolls and outer circles. For this final string you need to look at the post by rrbb titled

Basically it is the ordinality of numbers. Ordinality for people who don't know is nothing but the order in which numbers are positioned in a set.  Put it simply in the explanation of dozens consider the order in which the dozens appear as follows;

3,2,3,3,1,3,2,1,3,3,3,1,1

If you have to right it in positions it will read like

3,3,1,1,3,2,3,3,3,1,1,2,1

Let us assume that repeat of the last two dozen is more likely than the dozen that is sleeping and combine it with cycles to say that you are more likely to see the defining dozen appearing again. That makes it clear that you should not play for cycles of 3 position or one should ignore 3 position when it occurs.

Look at the results.
0,0,0,2,-1,-1,-2,0,0,0,2,-1,1

This is one more step in the direction of how I play.

This is what i mean with positions.
More details about it: link:://:.rouletteforum.cc/index.php?topic=17115.0

So we have cycles and a parallel game. The only question that remains is; How can we combine these things so we can get rid of the losses while keep getting winners.

Maybe we can use some statistics for this. If everything points out we would win. We bet.

[3Nine]

From page 1!  Sound familiar?!

This is just the first stepping stone. Before getting further into the world of random and non-random and how we can combine these two worlds, another question. As I touched upon dozens, “A dozen on the carpet, a dozen on the wheel, a selection of 12 numbers that changes constantly. Are they different? Do these bet selections result in changes to your predictions or the distribution?”

[falkor2k15]

From page 1!  Sound familiar?!

That's one of those questions/hints that doesn't lead to any eureka moments. I can attempt to answer it though. Let's say our cycle was defined by dozen 2, so we expect more chance that the next cycle will be defined by dozen 2 on the carpet. But that dozen 2 could close the cycle at an unspecified length, so when we try to guess which length to bet - or bet @ all lengths (and lose many attempts at trying) we end up at break even/house edge. So if we were to substitute the carpet defining dozen for a position based dozen or a parallel dozen of some kind, I think we are still heading for break even. So I think it might affect changes to distribution.

[falkor2k15]

edit: post still wip...

[falkor2k15]

Incidentally, "Defined by same" has been the most interesting problem since this topic began... I got some stats for Defined by Same in terms of regular dozen cycles...

CL1s	3794	33%
CL2d	2608	23%
CL2s	2582	22%
CL3d	1675	15%
CL3s	880	8%
	11539

CL1s + CL2s +CL3s = 64% same

Same is a 1 dozen bet, but Different is a 2 dozen bet, so we somehow need to re-calculate. To get defined by same on CL1 is a 1 dozen bet of 33% (same as normal).

Let's say we miss out CL1:

CL1s
CL2d	2608	34%
CL2s	2582	33%
CL3d	1675	22%
CL3s	880	11%
	7745

Still 33% for same.

Miss out CL1 and CL2:

CL1s
CL2d
CL2s
CL3d	1675	66%
CL3s	880	34%
	2555

Still 33% same.

So as individual spins it's always 1 dozen/33% for same, but together as 1 event it's 64%... how to take advantage of that?

Perhaps we need to stitch/hedge bets so that we always break even when it's defined by different instead of same?

[falkor2k15]

Everyone disappeared again? I was going to talk about the "Defined by Same" problem in a bit more detail, as I believe it's the key to Pri's Non-Random strategy providing it even works and in spite of all the smoke and mirrors and criticism/discouragement/silent treatment I've received throughout the course of this topic.

EC Cycles - Defined by Same
HLH
H...

Ratio: 75% same
Non-Random Limit: 2 spins
Risk: 1 to 2 units
Reward: 0 to 1 units

Dozen Cycles - Defined by Same
1231
1...

Ratio: 64% same
Non-Random Limit: 3 spins
Risk: 1 to 3 units
Reward: 0 to 2 units

We have at least 50% predictability in both cases that the next repeat event is going to be defined the same as last cycle, but we've got no obvious way of expressing this through betting the entire "all-in-one" event to finish up as "same" without risking 1-3 units. I doubt there is the ability to stitch it, and unfortunately it seems not within our power to effectively bet for it in a cost-effective manner. So I think we need to try to improve on the situation somehow through workarounds even if we cannot find a direct solution to the problem, and I've got a few ideas... Firstly, "defined by same" is a misleading way of looking at the problem. The real problem can be expressed as simply as this:

HL...

12...

The previous cycle actually has nothing to do with the next repeat. The next repeat is dependent on whatever pre-defining element (H, 1) and Halves (L, 2) - let's refer to them both as "uniques" - have appeared before the repeat occurs or reaches it's Non-Random limit. So what started out as 75% for H to repeat or 64% for 1 to repeat is now split equally between H and L + 1 and 2, respectively. Parallel games most likely use the same concept; I guess they simply provide extra support for either the HL or the 12 to repeat - perhaps via the birthday paradox where "parallel" halves may impact them further. The other parallel games - though I've not tried them - keep track of prior repeats that may impact the current cycle in some other way; however, looking back at the previous cycle resulted in no impact and independence in all the tests I did. The only dependency I did find were coming from the aforementioned uniques occurring in a fresh cycle.

So there's a few problems with playing for repeats that we could possibly improve upon:
1) The entire repeat event starts out costly and is out of reach in terms of trying to effectively bet the entire thing even though we can predict > 50% what the result will be, and this continues as the halves/uniques drop in till the limit is reached and we reach stalemate. By then it's too late to bet the repeat for the very next spin - and the odds are always having to be recalculated for a reduced sequence. Looking at the Defining Order instead of Defining Element could potentially improve the situation - or better still using outer cycles then we have more flexible stitching available for betting unique Cycle Lengths:
CL1 CL2 CL2
CL2...
Each CL is dependent on the unique dozens that came before the closing repeat that sealed the event; likewise, each outer cycle is dependent on the unique CLs that came before that repeat. However, even with Outer Cycles and unique Cycle Lengths I still don't think we can effectively bet the entire outer cycle as one event starting with only 1 bet - but I think it would be an improvement still. Through outer cycles we do at least have secondary dependencies wrapped inside primary dependencies and multiple constants are overlapping in terms of their affected impact zone.

2) The repeat doesn't come quick enough. For dozens the limit is 3 spins, so we could speed up the repeat by converting to SD cycles.
12322123 = DDDSDDD
or
1231 = S
11 = S
1233 = D
(just compare first and last)
So we can express Dozen Cycles as SD cycles then the limit is reduced from 3 to 2, so the repeat has to come quicker. SD on Order looked the most promising. We also have Reverse Order/Positions as a single stream - RO on Order/Defining Order looked the most promising there. The limit could be reduced further by betting for simultaneous repeats on all constants, such as defining + order + CL, so that we are guaranteed a repeat on the very next cycle in some cases- whichever constant repeats first.

3) Each unique has an equal chance of repeating once it's appeared. This is a problem - particular when you reach the end of the limit - because once all uniques have shown we cannot say which has more chance of repeating. The solution to this is through creating Outer Cycles - not based on Cycle Length - but based on Order+CL as one to give 6 options: CL1o1, CL2o1, CL2o2, CL3o1, CL3o2, CL3o3; from left to right they are each more likely to repeat after one appearance. So if we reached Outer CL6, rather having stalemate we could bet the first option, the first three options or even the first 5 options to close the outer cycle and seal the next repeat with only a small chance of CL3o3 occurring. This is the opposite of trying to reduce the 3 outcomes to 2 SD outcomes - we now have 6 options. In standard Outer Cycles, CL2 is also a more likely event than CL1 or CL3, so that's why outer cycles defined by CL2 have a 76% chance of being defined same next cycle because CL2 will over power it's counterpart uniques in the fight for the repeat.

Together, all the above solutions create a workaround for the initial problem of playing for repeats. And that's all for now... feedback?

[ati]

I didn't disappear, I just don't have much time to think about roulette these days. I'm very busy at work, and by the time I get home from the office(7-8PM), I'm usually brain dead.

[Herby]

Quote from: Priyanka on September 16, 2015, 10:58:18 AM
   â€œA dozen on the carpet, a dozen on the wheel, a selection of 12 numbers that changes constantly. Are they different? Do these bet selections result in changes to your predictions or the distribution?”

From a mathematical viewpoint there is no change in the relations of the outcomes of the distribution.
As far as I can remember (and see as a Non Mathematician) Pri argues mathematically.
-> Choose your dozen (lines, etc) as you like but keep the laws of probability in your mind.

[Herby]

This is another type of parallel game that we should ideally avoid as overly complex unless it's absolutely necessary to include them - TBC.

Who decides what is overly complex ?
If one tries to find a way to the impossible it's not a good advice to avoid the complex.