Random Thoughts

[Turner]

I've got Turner trapped in a cage like a game of Roulette...

Sorry...in a cage like a game of roulette?
How does that wotk exactly?

[RouletteGhost]

[falkor2k15]

Falkor...no I dont know what it means. What does it mean?

CL = Cycle Length; D = Dozens; L = Lines

Dozens have up to 3 cycle lengths:

11 = CL1
121 = CL2
1323 = CL3

Lines have up to 6 cycle lengths:

44 = CL1
433 = CL2
3423 = CL3
12452 = CL4
513244 = CL5
6123454 = CL6

So I'm simply testing Dozen Cycles alongside Line Cycles! There appears to be several repeats of a Dozen before a Line cycle closes, but I have yet to properly run the test and analyse the data, so cannot comment too much at this stage...

It's now Turner's turn to get revenge on me! 

[falkor2k15]

Sorry...in a cage like a game of roulette?
How does that wotk exactly?

It means that Steve's outcome later may depend on what action you take first... You've only got 3 choices:
CL1 = put Falkor on moderation
CL2 = laugh out loud
CL3 = ...

So depending on what you do could affect what Steve does next:
CL1 = ban Falkor
CL2 = put Turner on moderation
CL3 = warn RG
CL4 = warn thelaw
CL5 = laugh out loud
CL6 = ...

The cage is 4x7!

[Steve]

Falkor, only your post was reported. I'm not aware of any other infringing posts.

I'll take CL1 thanks.  The top one. You had two CL1's.

[falkor2k15]

You two guys are too predictable - like Roulette! 

[Still]

Now when it comes to the topic of stitching together bets, it is also important to understand which combinations are profitable and which ones are not. The combinations which might seemingly give better odds at first sight may not be the ones that will be profitable and vice versa. Taking a simple example.

Red and Odd. If we need to stitch together these two, will you place one bet on red and one on odd or one bet on red and 8 bets on the black odd numbers? Any creative ideas and view points?

I put together a spreadsheet recently that tested one possible application of this idea, to stitch together even chance bets into some other number. 

I tested putting two EC's together to come up with some weird four sided dice...basically, four unique patters made up of:

RB
BR
RR
BB

I'm assuming that any system that works with four elements would work with this binary numbering scheme.

If you bet on all four of these at the same time, you'd be betting 2 units on red and 2 units on black, every spin. No point.

So, i tried to come up with some selection process, and decided to repeat the last two patterns.  That is, whatever the last two patterns were, i'd take the two remaining patterns, and bet on those over the next two spins.  The idea is that the prior patterns have the best chance of repeating over the next couple of spins. 

This test yielded no edge. 

Does anybody have any suggestions for a four elements game?

Likewise, these binary stitching schemes can be built up to represent an 8 sided dice, a 16 numbered wheel, or a 32 numbered wheel. 

If there is a Birthday Paradox, or Pigeonhole Principle that applies to these number specifications, that should also apply with these stitched-together numbers.

The next step up is an eight element game consisting of these pattern/binary-numbers:

RRR
RBR
RRB
RBB
BRR
BBR
BRB
BBB

And so on, doubling up for each 'digit' added.

Now lets see what a successful parlay would be worth.

First bet (of one unit) wins you 1 unit.
Second (parlay) wins you 2 units.
Third parlay wins you 4 units...

...for a total of 7 units, at the risk of one unit to begin with.

So the chances are 1 in 8 but the payout is 7 to 1.

I'm not sure what the exact computations are, but it is comparable to a line bet where the chances are 1 in 6 but the payout is 5 to 1. 

So there may not be any advantage, ALONE, to use these artificially stitched-together numbers, unless...unless there may be some advantage to keeping all bets related to to a basic binary numbering scheme, or EC game. 

Looking at a 32 numbered wheel, the payout would be something like:

1+2+4+8+16 = 31 for a game where there is a 1 in 32 chance of winning. 

Can anybody do the calculations on this?

Is this better than -2.7% ?

Getting creative, i suppose it were possible to make patterns that exactly mimic any/every other bet on the roulette wheel/mat. 

For example, what is the equivalent of one bet on a dozen, parlayed over two EC bets?   

Is that equal to a "line"? idk

And if so, are there any Pigeon Holes that would to lines that would also apply to this parlay?

We are looking for phenomenon that offer 55%+ chances from some phase of a line bet.  I have heard there are such situations. 

Anyways, now, instead of working with unrelated EC's and unrelated line bets, we would be working with the same basic building block elements. 

We could, for example, be running VdW's on all the EC's and cycles on all the dozens...leading up to the one line bet we are aiming to make. 

Would this help at all?  I don't know. But has anybody thought along these random lines yet?

[Still]

RB
BR
RR
BB

Using the first four patterns above in some sort of parlay scheme, the payout would be:

1+2 = 3 for a game where there is a 1 in 4 chance of winning.   

So, not very good at all. 

But from my previous post, it does appear the odds get a little better, more comparable with European wheel bets...if not better, by the time you get up to 32/1 chances (five parlays).

If there are repeat systems that work with 37 numbers, maybe this would work just as good...or better?

Each session would be stretched out five times longer as it takes five spins to create one number with a 1/32 chance of winning.

Not good if it's not an improvement.  Good if it is.   

[Still]

Let's see if i can build a parlay bet that has a 1/36 chance of winning, and find out what the payout might be. 

By parlaying an even chance bet five times, i find out that i have a 1/32 chance of winning, for a 31 to 1 payout. 

To get a 1/36 chance bet, it seems we will need to add a 32 number bet to the end of a five-step equal chance parlay.   

Let's double check the math as this is my weakest link. 

1/36 = .02777
1/32 = .03125

What do i have to multiply 1/32 by to get 1/36 ?

The answer is .02777 divided by .03125 = .888888

Which is equal to 8/9 or 32/36

So, if at the end of a five step parlay, we put down all 32 in the pot down on 32 straight numbers, then if we win we will have a payout of 35 plus we get to keep the winning chip:

35 (payout) + 1 (risk capital) = 36

This is exactly the same as betting once, on one straight number, and winning:

35 (payout) + 1 (risk capital) = 36

These calculations disregard the zero for test purposes.

If you add a zero to this game, there's a 1/36 chance the zero will come up, which is 2.7%.

If the parlay i describe increases exposure to the zero, then it is a worse bet.  But if the exposure is not increased, then it's exactly the same.

Or is it?

Some people might think this is a more versatile, more granular bet. 

[Still]

To find out if the parlay bet with a 1/36 chance of winning has more exposure to the zero, we may be able to look at the "total action" multiplied by the house edge, or 2.7% in case of one zero.

1 + 2 + 4 + 8 + 16 + 32 = 63 units of "action".

As we can see, the exposure is almost twice as much as coming in and plunking down 32 units on a straight up number. 

63 x .027 = 1.75 units expected loss
32 x .027 =   .88 units expected loss

So, IF this is a more versatile, more granular bet, it's benefits must overcome whatever the house edge is: x 2.  The en prison rule on Euro wheels helps a little, but there would still be an additional .66 +/- units expected loss starting with a base unit if one. 

On the other hand, it would seem that a lot of action would be saved when the targeted number is more often than not disqualified early in the parlay sequence. 

And on the other, other hand, we have to remember that the total action, to bet one (1) unit on 32 numbers all at once, is the same as risking one (1) unit on 32 different attempts to target one straight number.  The "action" exposed in between is mostly house money, parlayed into more house money.

So it does appear to be the same in terms of chances, payout, and exposure. 

The question: is it really more versatile?

[falkor2k15]

Still, you need to bring in a parallel stream, otherwise no edge - as Pri told me many times.

Dozen Cycles + Line Cycles share common numbers. You seem to be picking up one stream of Pri's other parallel games that involves the same payout odds:

"I have tried various things and have not been able to figure out a way to induce dependencies between parallel games. All thumbs down.
There is one last hope left though which am checking now. It goes like this. It is stiching together of bets. While playing quads I have realised that 1-9, 10-18, 19-27, 28-36 forms quads in terms of spins. But the other way to make quads is by combining results of two spins. Like combining Two ECs like Low(1-18) and high numbers(19-36). The combinations are LL, HH, LH and HL. Here I could potentially have two streams one as a stream of quads with teh above combinations and other as a stream of ECs made of L and H. Because they are formed of same elements they are dependent. I am sure there is some playability I can figure out between these two streams and cycles, so working on it."
link:://:.rouletteforum.cc/index.php?topic=17014.225

Above is full context to stitching two ECs - as part of a parallel game with Quad Cycles.

[Still]

Still, you need to bring in a parallel stream, otherwise no edge - as Pri told me many times.

There are as many as three, possibly four assumptions in this one statement, too nuanced to elaborate on, without veering off my current stream of thought.     

Here you've brought in a parallel stream of lingo, running parallel to this thread, which, walking on the edge of knowing and unknowing,  brings no edge.

You have assumed at least a couple of things about what Priyanka has said in the quoted post. 

First of all, no edge was mentioned in the posted text, beyond "i'm sure there is some playability i can figure out", in regards the idea of running two parallel streams.   

Second, you are conflating games with streams.

Dozen Cycles is a holy game.
Line Cycles is a highly worshiped game. 
VdW on Dozens is a revered game (notice i spell Dozens with a capital D). 
VdW on EC's is a beloved game. 

In the quoted post, it was all "thumbs down" on the unholy alliance of running parallel GAMES, as beloved as those game combinations may be to our assumptions.

Instead, hope was given to running parallel STREAMS of equivalent bets, one stream made up of the usual bet, and the other made up of a custom combination of EC's and whatever else it takes to make an equivalent bet. 

In theory, both of these similar bet streams will lose in the long run, just as both "Parando Paradox" bets lose in the long run. 

The overall suggestion, going back to the first time "Parando's Paradox" was mentioned in this thread, is that perhaps there is a way to switch between these streams at opportune times, with either a third stream, or another game as a switching mechanism.

If this is the case, it may not make sense to run two parallel switching mechanisms, if that's what is meant by "all thumbs down" on running parallel "GAMES".   

Dozen Cycles + Line Cycles share common numbers.

And so?

Combine these at your own risk according to your faith. 

Currently, i am only building up equivalent bets. If there is a way to run these equivalent (losing) bets in parallel, and then switching between the two at opportune times, that would be a bonus. 

Right now, i am only building equivalent bets. Has anybody in this thread ever done that?

[Still]

Quads.

This term is used by Priyanka one way, and used by most others another way.

The most commonly understood use is a four-number bet that pays out 8 units for every unit put at risk.

Priyanka's "quad" is made up of THREE STREETS. 

A street pays 11 units for each unit put at risk. 

If you bet on three streets, the risk is three units to gain 11. 

Two of the units will be lost, so the payout will be 11-2=9 units  (at the risk of 3 units).

9/3 is the same as 3/1, which is also the calculated reward:risk ratio for the aforementioned parlay bet made up of two EC's:

RR
RB
BB
BR

Like the quadrants that Priyanka has sometimes spoken of, these each have a 1 in 4 chance of winning. 

Theoretically, anything that would work on any 1 in 4 chance, would work with these. 

Looking back at Priyanka's "Speedy Gonzales", it would seem possible to make a 1 in 4 chance float for a million spins, albeit with a progression and some serious draw-downs with continuous play. 

Whether there is any relevance to that i have no clue, except it has been suggested that at one point, Priyanka had put some hope in the idea that running parallel streams of quads, and switching between the two, might offer some better odds. 

Again, i am suggesting, perhaps for the first time in this thread, that there is a difference between parallel streams, and other (dependent) games that might function as switching mechanisms. 

The idea is to run parallel streams, not parallel switching mechanisms.

[Still]

Parallel

Using the aforementioned "quads" as an example, how might we run two parallel "streams" (also using the aforementioned parlance)?

First we could observe that it takes twice as many spins to accomplish anything using the EC format (RR,RB,BR,BB) than it does to accomplish something with Priyanka's original version of quads.   

So the "cycle", whatever it may be, will be twice as long. 

This seems like a mismatch to me that is not equivelent. 

What seems more equivalent to me would be to run two consecutive streams of the one "quad", against one stream of the twice-as-long-cycle "quad" made up of EC combinations. 

In some weird way, then, both cycles would be equally long, and so, more equivalent. 

No clue how this could be useful, but it does seem to be a new way of thinking about these concepts, possibly more in line with what was MEANT to begin with.

[Priyanka]

The idea is to run parallel streams, not parallel switching mechanisms.

Still - interesting read. And you are so right in the above statement. Parallel switching mechanisms should and will fall into the house edge and expectation hole as they will be independent bets. The idea is to run two parallel streams and inducing some kind of dependency between those streams. If they run as two parallel streams without any dependency then it will again succumb to the law of large numbers. 

An example of what I mean by dependency is below. Again the key word is example. Taking the example of repeaters, if we run two streams one containing dozens and other containing lines, we can easily come to a conclusion that if repeater has not happened in dozen then it has not happened in lines as well. So that is a dependency as lines are a subset of dozens. Likewise there are other aspects we could use to create dependencies between two parallel streams.

Very interesting read Still.