Triplet Dozens

Atlantis · July 28, 2017, 06:57:27 PM

TRIPLET DOZENS
===========

The state of a triplet (three dozen results) can be:

same eg. 222 or 111 or 333
repeat eg. 122 or 331 or 121 etc...
different eg. 213 or 312 or 123 etc..

Idea: wait for 2 consecutive triplet lines consisting of of "same" and/or "different" then play for next line to be a "repeat" line after the first 2 results are in...

For example if first 2 results are 1-2 then play d1+d2

For example if first 2 results are 2-2 then play d1+d3

If lose play as above for the "repeat" to occur on next line... (+1u?)

113 (repeat)
333 (same)
231 (different) trigger
322 (repeat) after the 32 bet d3+d2. result=2. won+1
212 (repeat)
122 (repeat)
331 (repeat)
213 (different)
323 (repeat)
133 (repeat)
131 (repeat)
221 (repeat)
132 (different)
312 (different) trigger
122 (repeat) after the 12 bet d1+d2. result=2. won+1
211 (repeat)
322 (repeat)
332 (repeat)
322 (repeat)
331 (repeat)
121 (repeat)
312 (different)
333 (same) trigger
122 (repeat) after the 12 bet d1+d2. result=2. won+1
233 (repeat)
221 (repeat)

(Bear in mind on avg. 18/27 trips of doz WILL BE repeat trips)

This is a grinder, patience and attention needed - so better play on trusted r.n.g. with option: fastspin only.

Can play for fun - LOW STAKES even on 1p roulette in UK. I can play 2 doz@10p units with 1/10th insurance cover on ZERO too for added safety. mathemagician (RKGold) might like it.

Not for everyone - so take it or leave it.

Seems solid though!

A.

AsymBacGuy · July 29, 2017, 01:02:15 AM

Yeah. This is one of the simplest way to play dozens.

We target the last two different dozens appeared then bet them hoping that the third will be silent as long as possible.

Of course it's like choosing to bet randomly two dozens out of three without assessing the silent one.

But acting in this way we have a general picture of what it's going to happen and by a decent bet selection and by a careful money management we might get good short-intermediate term results.

Say we put in our chart a minus sign (-) after a loss and a plus sign (+) after a win (zero/zeroes not included)

Most likely we'll get sequences as ++-+-++--+-+++++---++-++++-++++--+....

We can be sure as hell that itlr consecutive + will be double placed than single + and the same is true about - signs, this time by an opposite fashion. same about ++ vs longer ++ sequences and so on.

If we pnly bet one time whenever any single + (preceded by one or more minus sign registered) or whenever a single - (preceded by one or more + sign registered) had come out we are going to reduce variance.
We are not altering the W/L percentages, but we'll have the same results than wagering every hand (yet paying a minor vig).

Easy to notice that sequences as -+- or +-- will be losers and -++ or +-+ will be winners.
Of course itlr the winners will be double placed than losers.

If we choose to simultenously wager those two features by a 1-3 progression (1-1 and 3-3 bets) we know we'll lose whenever every -+-- sequence will come out. Every other situation will be a winning one.

Notice that sequences as ++------ or +-+----------- (terrible ones for a continuos play) are winning ones. Because we are respectively get our win on the first spot in the first sequence and on the second one on the second sequence.

Again the only losing sequences by adopting the 1-3 progression are -+--

After having tested many many real spins I can assure you that the probability to get consecutive patterns not winning just one spot is very low.
I mean that -+-- consecutive patterns don't come out quite often. More realistically I'd say you need a lot of spins to encounter a situation where you'll find two consecutive losing patterns.

For that matter looking at three consecutive losing patterns in a row is a sort of a "lottery" finding.

Actually the theorical probability odds to get three losses in a row are 1:729 (zero discounted) but in the real world the overall losing probability is well lower as an homogeneous world is just an utopia.
Wonder if we choose to wait and wait and wait the appearance of one or better two such fictional situations. We can only raise our probability of success.
Or trying to take advantage of many other possible profitable situations we should assess before betting.

Definetely the game is EV-, but only whether we consider any outcome equally probable anytime, everywhere and anyhow.

That's nonsense.

as.

Atlantis · July 30, 2017, 03:03:55 PM

Very interesting. Thanks for that AsymBaccGuy.

Another thought that springs to mind is to record a "rolling history" of the last 27 'triple' results (same or different or repeat)
When the 'repeats' are lagging behind a little begin betting for repeat to occur.

For instance if the number of repeats = 16 or less then bet for repeat. Keep betting for repeat until total repeats increases to 17 or above out of the 27 history results; in which case wait until total repeats falls back to 16 again before jumping in to bet....

Can be done manually - as I did when I tested - but would be easier with a program to count and update the "triples".

This also seems to work well. I chose 16 as my trigger, but could be set lower if desired. (in which case you get less bet opps)

eg:
1 = same (3 of same doz)
2 = repeat (2 out of 3 doz are the same)
3 = different (All 3 doz are different)

Only when I get 16 or less out of 27 repeat trips do I bet for repeat.

No probs getting to +10 on fastspin RNG today doing that in abt. 500 spins.

A.

Sputnik · July 30, 2017, 03:33:05 PM

Quote from: AsymBacGuy on July 29, 2017, 01:02:15 AM
Yeah. This is one of the simplest way to play dozens.

We target the last two different dozens appeared then bet them hoping that the third will be silent as long as possible.

Of course it's like choosing to bet randomly two dozens out of three without assessing the silent one.

But acting in this way we have a general picture of what it's going to happen and by a decent bet selection and by a careful money management we might get good short-intermediate term results.

Say we put in our chart a minus sign (-) after a loss and a plus sign (+) after a win (zero/zeroes not included)

Most likely we'll get sequences as ++-+-++--+-+++++---++-++++-++++--+....

We can be sure as hell that itlr consecutive + will be double placed than single + and the same is true about - signs, this time by an opposite fashion. same about ++ vs longer ++ sequences and so on.

If we pnly bet one time whenever any single + (preceded by one or more minus sign registered) or whenever a single - (preceded by one or more + sign registered) had come out we are going to reduce variance.
We are not altering the W/L percentages, but we'll have the same results than wagering every hand (yet paying a minor vig).

Easy to notice that sequences as -+- or +-- will be losers and -++ or +-+ will be winners.
Of course itlr the winners will be double placed than losers.

If we choose to simultenously wager those two features by a 1-3 progression (1-1 and 3-3 bets) we know we'll lose whenever every -+-- sequence will come out. Every other situation will be a winning one.

Notice that sequences as ++------ or +-+----------- (terrible ones for a continuos play) are winning ones. Because we are respectively get our win on the first spot in the first sequence and on the second one on the second sequence.

Again the only losing sequences by adopting the 1-3 progression are -+--

After having tested many many real spins I can assure you that the probability to get consecutive patterns not winning just one spot is very low.
I mean that -+-- consecutive patterns don't come out quite often. More realistically I'd say you need a lot of spins to encounter a situation where you'll find two consecutive losing patterns.

For that matter looking at three consecutive losing patterns in a row is a sort of a "lottery" finding.

Actually the theorical probability odds to get three losses in a row are 1:729 (zero discounted) but in the real world the overall losing probability is well lower as an homogeneous world is just an utopia.
Wonder if we choose to wait and wait and wait the appearance of one or better two such fictional situations. We can only raise our probability of success.
Or trying to take advantage of many other possible profitable situations we should assess before betting.

Definetely the game is EV-, but only whether we consider any outcome equally probable anytime, everywhere and anyhow.

That's nonsense.

as.

One thing that i test with random bits is one stream by it self, for example singles versus series using red and black by them self, the distribution is the same and you don't mix red and black together.
You more i test and you more i get used to use black for it self and red by it self playing even money, i also start to feel that mix them together give more erratic results and more chaos.

Le me explain - as red by it self as random stream is a victim for the law of series and same for black, then what happens when you mix them together, well my opinion is that there is a tendency difference. For example is i count singles versus hit of series of three and higher, then if i reach 2.5 SD with 12 black singles and 2 series of two blacks, then there is a strong dominance at the red side - imbalance, even worse if i get more singles before regression towards the mean.

This does not occur when you mix them both together as you get perfect balance between zig zag distribution and only aim to encounter larger series to hit - so in one way you have singles versus larger series using both red and black together - but using only black you get both tendency of strong side imbalance and the expectation of larger series to unfold.

In one way is the same way using clockwise direction and anti clockwise direction for advantage players - but this solution is for system players.
I would say same theory apply using two dozen, why not pick only dozen 2 & 3 as your strong side against dozen 1.

CHeers

Kattila · July 30, 2017, 04:21:15 PM

Why not create your own EC s or dozens by splits or streets ? Also
can create singles and series using groups made by streets or splits.

Example *forced* series or singles :

Street/ groups and patterns

2 a or a
6 b a
11 a b
3 b b
11 a b
4 b b
7 a a
9 b a

Can attack against series or against singles.
Bet for max. 4 spins for W or L4. Next attack
after new create/trigger . Use possitive progression.

Example *forced * dozens ( forced into Pattern s ):

Splits / groups and patterns

5 a or a
17 b a
3 c b
8 a b
11 b c
2 c c
17 a a
12 b a
6 c b
8 a b

can attack two groups( last two to repeat) or can attack only
one group (for example the second last group, or 3rd position).
Big chances to hit other positions(gaps/ distances) , not the created
pattern formation.

Atlantis · July 30, 2017, 05:48:24 PM

Quote from: Atlantis on July 30, 2017, 03:03:55 PM

Another thought that springs to mind is to record a "rolling history" of the last 27 'triple' results (same or different or repeat)
When the 'repeats' are lagging behind a little begin betting for repeat to occur.

For instance if the number of repeats = 16 or less then bet for repeat. Keep betting for repeat until total repeats increases to 17 or above out of the 27 history results; in which case wait until total repeats falls back to 16 again before jumping in to bet....

Can be done manually - as I did when I tested - but would be easier with a program to count and update the "triples".

This also seems to work well. I chose 16 as my trigger, but could be set lower if desired. (in which case you get less bet opps)

eg:
1 = same (3 of same doz)
2 = repeat (2 out of 3 doz are the same)
3 = different (All 3 doz are different)

Only when I get 16 or less out of 27 repeat trips do I bet for repeat.

To get more bets can use Part 2 of the idea - which is basically the opposite....

If repeats = 21 or more bet for SAME or DIFFERENT to happen.
eg: 1-1 bet for d1
eg: 2-3 bet for d1
eg: 1-2 bet for d3

This is a one dozen bet (after first 2 results)

However , if repeats drop below 21 then STOP betting for same or different until repeats count = 21/27 again.
When repeats drop to 16, implement Part 1 - betting FOR the repeats. (2 doz bet)

A.

Atlantis · July 30, 2017, 08:40:50 PM

Tried it with parts 1 and 2 and won +30u
I found a way using RX to track the "triples" by setting the "spins to calculate graphics from" to 27 and using number buttons 1,2 and 3 to act as the same, repeat or different results accordingly. Once I had the 27 results, updating after each triple, I could watch for the triggers.
So far never placed a doz bet higher than 3u.

The below attached gfx indicates a bet on a single doz (to result in a same or different outcome) after the first 2 spins.

A.

AsymBacGuy · August 02, 2017, 12:02:52 AM

Interesting points.

I wish to add my two cents in order to possibily improve our probablity of success.

We do not want to try to win at any cost, we are there trying NOT TO LOSE. It'd be a big accomplishment to achieve a not losing situation since we have to overcome a big -5.26% or 2.7% negative edge.

Thus we need to reduce at most our winning hopes. For every post mortem long winning situation we'll get a proportionally higher amount of losing situations without any exception.

We are entitled to lose, if we break even after any session we really are good players. If we quit as winners most of the time, we are formidable players.

We do not want to be chained to expected values only. Conversely we should notice carefully what the actual values are.
Imo the best play is to try to balance the expected values with the actual values, especially when actual values seem to correspond to expected values.

Since the past results won't affect the future ones, we can infer that an homogeneuously equilibrated world won't exist at all. or, better sayed, that such probability is very very low.

Definetely rare events come out in clusters than they will disappear, whereas most likely events come out in clusters and more probable than not they are interpolated with the less likely opposite events.
Up to a point where the first assumption takes its validity.

Since the world is random (or supposedly random) there's no point to figure out the precise spot where things will shift toward a less likely events direction.

But the sum of such endeavours most of the time will.

as.

alrelax · August 02, 2017, 12:08:56 AM

Asym, reference the clusters..and other similar things......."steal the situation' and pounce in it.

AsymBacGuy · August 02, 2017, 01:10:32 AM

Quote from: alrelax on August 02, 2017, 12:08:56 AM
Asym, reference the clusters..and other similar things......."steal the situation' and pounce in it.

Of course, but there's an important difference about a 50/50 game and roulette dozens.
On 50/50 games, if A comes out B is zero and vice versa.

Over an A, B and C system, if A or B or C come out, two outcomes out of three will be silent per every spin. Not to mention that the probability to be right or wrong won't be fifty fifty as it's obviosuly shifted toward the two dozen chosen or excluded.
There are not other possible betting situations to be taken.
Zero/s disregarded, you'll be right either 33.3% or 66.6% of the time or wrong for the same amounts.
The fact we're going to be payed less than expected shouldn't affect our strategy.

as.

Atlantis · August 17, 2017, 10:47:19 AM

I like this as:
You're not playing every spin.
You can avoid some 0's because of that.
Plenty of time to record and chart play (less chance of mistakes)
Not complicated to understand and easy to spot triggers.
Playing on live dealer SPEED roulette is my preference. (with chat and sound off; I listen to my favorite music while playing)
Gives you time to think and not rushed to make decisions.
I know its just another Double Doz bet; but has a nice feelgood 'vibe' to me; patience required though.

Later maybe I try with separate progression for the columns.

Another way to play is to also record # of repeats and the # of same/diff. If you think an imbalance might be occurring (repeats lagging behind) then think about bet, bearing in mind 18/27 repeats is avg.

EG:
Triplet formations
------------------
R
R
S
R
D
R
S

at this point you have 4 triplet repeats and 3 others = 4/3 to repeat.
I'm tempted to play for repeat to start to pull away any time soon...

Example session just now. First I check the past history marquee and get some triple results before starting play...

R=repeat triple
S=same triple
D=different triple

O=total same/different triples

311 R
212 R
332 R
232 R
312 D
222 S---- 4R/2O
132 D-----4R/3O L-2 -2
112 R-----5R/3O w+2 +0
332 R-----6R/3O w+1 +1
133 R-----7R/3O w+1 +2
121 R-----8R/3O w+1 +3
133 R-----9R/3O w+1 +4 - I stopped here and wait; expecting S or D
113 R-----10R/3O
221 R-----11R/3O
232 R-----12R/3O
113 R-----13R/3O

My gut is telling me expecting S or D to happen pretty soon seeing as I have 13 triple repeats out of 16 triple results I quit here.

(However, could have played for single dozen (S or D) here)

But I decide not to risk it and keep hold of my +4 and come back later for a new session

This is more flexible, individual approach and in the style of adapting and reacting to what the wheel is throwing at you + gut feeling/instinct rather than playing according to the strict, rigid rules of a systematic route.

Hope this helps.

A.

hexfex82 · August 17, 2017, 12:25:59 PM

Hello A.

i tried it now live. I looked at the flow and played directly after S and D for a R. And 2 times also after a R. All with success. Thx

3   3   3   S
2   2   3   R
2   3   3   R
3   3   2   R
1   3   2   D
2   3   2   R
3   1   1   R
3   1   3   R
2   3   1   D
2   1   1   R
3   1   3   R

hexfex82 · August 17, 2017, 05:49:57 PM

2 additional Sessions:

1   1   3   R
2   2   2   S
2   3   1   D
1   3   2   D
3   1   2   D
2   1   2   R
2   3   1   D
2   2   3   R
1   1   2   R
3   1   3   R
1   3   1   R
1   3   2   D
3   2   2   R
1   1   3   R
1   2   2   R
1   1   2   R
2   1   2   R
3   2   2   R
3   2   2   R
2   3   3   R

Very difficult

1   2   1   R
3   3   3   S
1   3   1   R
3   2   3   R
3   1   1   R
1   1   1   S
3   1   1   R
1   1   3   R
3   1   2   D
2   2   1   R
1   1   1   S
1   3   1   R
1   2   1   R
2   2   3   R
1   1   3   R

Atlantis · August 17, 2017, 06:15:14 PM

Thanks hexfex82,

Yes - the original way waiting for 2 non-repeat trigger is better/safer I think. One bet only. Maybe keep at flat bet too...?
Rather than trying to guess/predict.
However, I think overall something is missing to make this better. I need to think - but maybe a hedge is required and a token bet on the zero too. I am going to try and see if I can convert/tweak it somehow. I'm stopping play for now after a 'hard' session myself.
Thanks for your results. I'll come back to this again as I think something in it.

A.

Blue_Angel · September 21, 2017, 07:31:47 AM

May I suggest a twist?

Since we are looking for 1 repeat within 3 spins, by the way this is most common sequence, to have 2 unique dozens/columns + 1 repeat within 3 consecutive spins, so instead of looking only 1 way why not bet simultaneously for repeats on both dozen and column?

Just an example:

first result of the triplet is for observation only and let's say is number 13, it belongs to 2nd dozen and A column, let's call it 2A.
1) 2A no bet
2) bet 2nd dozen + A column, if we get double repeat the net profit would be +4, if we get single repeat from either our net would be +1, if there is no repeat we lose 2 units, so in this case only we'd proceed to the final 3rd step.
Let's assume that after 13 was number 33, which means:
1) 2A
2) 3C at this point we would bet dozens 2 and 3 plus columns A and C, by winning both we would get +2 units net, thus recovering the -2 from previous bet, but in case we would win only one then it would cost us 1 more unit which would make the total -3, in case of 0 or 1B we would lost 4 units and the total would be -6.
If we have recovered then restart from 1st step, if after 2 bets/3 spins the total is -3 then restart from 1st step, only in case the total is -6 then increase to 2 units each bet spot (both dozen(s) + column(s)

The point is to recover by aiming a double repeat, therefore in case of -6 next bet would be 2 units on dozen + 2 units on column, the double win from both would return +8 units net, if we win either that's +2 and we restart, in case we lose then the total is -10 and on the 3rd spin/2nd bet we would bet last 2 dozens + 2 last columns with 2 units each ( 4 x 2 = 8 units).

The calculation progress in steps of 3 spins/2 bets and aims to recoup in a single double repeat/win.
When the bet is 1 D + 1 C the net from double repeat/win is x4 the unit value, when bet is 2 D + 2 C the net from double repeat/win is x2 the unit value.

We are looking for 1 repeat, from both or either D/C, it should and could happen most of the time, with a mild progression an overall profit will be just around the corner.

If you want a betting example then post from 20 up to 40 numbers.