Deposit Bonus 200% up to $1000

Author Topic: NEW SYSTEM: SINGLE DOZEN  (Read 35421 times)

0 Members and 3 Guests are viewing this topic.

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1937
  • Thanked: 926 times
  • Gender: Male
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #390 on: April 14, 2017, 05:14:00 PM »
Bayes,
Very good results.
Which progression did you use?

Thanks
 
The following users thanked this post: Reyth

TERMINATOR

  • Search YouTube for MANDELA EFFECT
  • Mature Member
  • ***
  • Posts: 263
  • Thanked: 321 times
  • Gender: Male
  • MANDELA EFFECT
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #391 on: April 14, 2017, 05:23:36 PM »
Worst session yet encountered.

Bayes, I cannot follow your triggers in your results doc.

For example, in spin 28, an "l"  at spin 28 is usually placed at the END of the trigger. Correct? So, spin 26,27, and 28 would be the trigger, and you indicate if this trigger won or lost by placing a W or L on line 28 if one of the next 3 spins won or lost. At least this is how your last version of Dr Dozen worked.

Thanks for clarifying.

Also, I would love some more details on your progression. Are you still fine tuning it?
« Last Edit: April 14, 2017, 05:39:38 PM by TERMINATOR »
 
The following users thanked this post: Reyth

TERMINATOR

  • Search YouTube for MANDELA EFFECT
  • Mature Member
  • ***
  • Posts: 263
  • Thanked: 321 times
  • Gender: Male
  • MANDELA EFFECT
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #392 on: April 14, 2017, 06:13:00 PM »
The price you pay for keeping the target modest is that on really good sequences you still only make 1 u every 7 spins, which isn't very efficient, but I have some ideas about how it can be improved when the expectations are running high.

1 unit every 7 spins is better than what I am currently averaging, Bayes. I average 22 units every 200 spins. If you average 1 unit every 7 spins, that would make your average 29 units per 200 spins, which is awesome.
 
The following users thanked this post: Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #393 on: April 14, 2017, 06:35:21 PM »
Which progression did you use?

It's a modification of the progression I used on the "sequence from hell" I posted earlier. The mod is to keep track of the target which increases by one every 8 spins (raised from 7 which I thought was too optimistic - and actually I think it could be raised further, especially since some spins are not bet - I think 10 would be better).

Initially the bank and target are set to zero, and at all times if the bank is greater than or equal to the target the stake remains at one unit. Otherwise, the stake is increase by 1 u per spin, unless, when the profit is calculated assuming this increase it would result in more than 1 u profit if the bet won. If this is the case it's adjusted down so that it results in no more than 1u profit.

i.e.

Add 1 to stake
IF 2 * stake > (target - current bank) THEN
    stake = (target - current bank) / 2

 
The following users thanked this post: kav, Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #394 on: April 14, 2017, 06:57:28 PM »
Bayes, I cannot follow your triggers in your results doc.

Yes it is a bit confusing if you forget that having started an attack, it continues for 3 spins regardless of when the win occurs. Also note that it doesn't look for a trigger until 6 spins have been recorded. Also the start of the losing sequence begins after the 27th bet not the 28th (the numbers are shifted down by one because the column headings are in line one on the spreadsheet).

So starting from the beginning the first bet is on line 7 and is on H. This is won on line 9 but there is one more spin before looking for a new trigger, which is found on line 10 and is H. However, this bet is aborted after 2 spins (line 12) because of rule #3, and there is then one more spin to complete the 3 spins before another trigger is looked for. If you continue in this way you should see that it all works out by the time you get to the 4 losses in a row.

Rule #3 does present a bit of a problem and I never thought to look at the next spin and to record a virtual loss or win. I suppose you're right in that it does throw off the stats, but then you always lose 2 bets, so it doesn't seem right to record it as an actual win, because you never place the winning bet.

Regarding the progression, I haven't implemented in the tracker because I'm planning to write a more generic application, since it can apply to any odds. So it's probably better to explain it in a new thread. I'll do that tomorrow.
 
The following users thanked this post: kav, Reyth, TERMINATOR

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1937
  • Thanked: 926 times
  • Gender: Male
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #395 on: April 14, 2017, 07:41:29 PM »
Bayes,

Thanks for your reply I understand this description

Quote
Add 1 to stake
IF 2 * stake > (target - current bank) THEN
    stake = (target - current bank) / 2

What I don't understand is why there are instances that your total is negative yet you bet 0

A summary would be very useful, because we are brainstorming on various ideas and it gets confusing.

 
The following users thanked this post: Reyth

TERMINATOR

  • Search YouTube for MANDELA EFFECT
  • Mature Member
  • ***
  • Posts: 263
  • Thanked: 321 times
  • Gender: Male
  • MANDELA EFFECT
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #396 on: April 14, 2017, 08:20:09 PM »
Thanks for explaining your Results doc more clearly, Bayes.

Rule #3 does present a bit of a problem and I never thought to look at the next spin and to record a virtual loss or win. I suppose you're right in that it does throw off the stats, but then you always lose 2 bets, so it doesn't seem right to record it as an actual win, because you never place the winning bet.

But remember, this XYY test is SEPERATE from the actual game itself. The only thing we are trying to determine, at this point, is if the XYY HITS more times than not, within those 3 spins (if an "X" does not appear before it). Period. It does not matter how many actual BETS we made for that XYY trigger (whether it's 1,2, or 3), because we are not looking at how many BETS we made. We are only looking at how many HITS it made within those three SPINS after the XYY trigger (and the 3 spins prior).

So, that's why we look at the three SPINS (regardless if they were all bet or not).

After these XYY results are in, and we determine that they ARE more profitable, only THEN can we incorporate these results into our actual game. And THEN we can determine whether or not to use the Red Flags when using this Special Trigger.

For all we know, it may be more profitable to IGNORE rule #3 only when we use this special XYY trigger in this specific circumstance, but still use rule #3 on XYY when we do NOT take the previous 3 spins into consideration. But we will not know this if we only take into consideration the spins that were bet.

For example, if we look at 100 XYY bets in which we used rule #3, and made only TWO bets, and it turns out that the 3rd bet (the one we did not bet on) would have actually WON in 80 of those games, then we might consider ignoring this red flag for this special trigger.

But, if we IGNORE this 3rd bet in our XYY Test because we did not BET this 3rd bet, then we will never know!

Quote
Regarding the progression, I haven't implemented in the tracker because I'm planning to write a more generic application, since it can apply to any odds. So it's probably better to explain it in a new thread. I'll do that tomorrow.

Awesome! I look forward to it. Maybe you can elaborate as to what you meant when you said "keep track of the target" and "current bank" when you post this in a new thread.

Thanks Bayes!
« Last Edit: April 14, 2017, 08:36:44 PM by TERMINATOR »
 
The following users thanked this post: Reyth

palestis

  • Veteran Member
  • ****
  • Posts: 668
  • Thanked: 500 times
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #397 on: April 14, 2017, 09:44:45 PM »
Worst session yet encountered. At spin 28 (on the Dozens) there were 4 losses in a row, followed by 2 wins, another loss, a win, then another 3 losses in a row. This took the stake to the maximum of 27u. It did eventually recover flat betting at the max stake and made the target profit of 42 units at spin 336.

This was a negative 3 stdev event.  :P
I am a little puzzled about this:
First there are WW back to back. What trigger the second W comes from?
 Because you can't have either WW or LL in this system in 2 consecutive numbers.
Then there is a W  in the 3rd spin after WW and another one in the 3rd spin after WW..W..W.
I was wondering what triggers they came from. That can only happen if numbers from previous triggers are used.
And the same I noticed in the first L  of the 4 back to back losses.  If the trigger used was 33-20-20 then the L should be marked in the 3rd losing spin. That would be 12-17- 20 the three losing spins of the trigger. That will push the next trigger a little further.
Can you clarify this?
« Last Edit: April 15, 2017, 12:46:41 AM by palestis »
 
The following users thanked this post: Reyth

palestis

  • Veteran Member
  • ****
  • Posts: 668
  • Thanked: 500 times
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #398 on: April 14, 2017, 11:03:11 PM »
@Bayes.
I processed the numbers you worked on in the excel sheet, manually.  But only the range where you observed the concentration of the most losses. I started using numbers from the beginning. After all it is a test.
But I didn't run into 4 back to back losses. ( that is 4 trigger losses).
By back to back losses I always mean an entire trigger loss@ 3 spins each.
So 4 back to back losses translates into 12 lost spins. I only found 2 back to back losses ( 6 spins).
Also when a trigger wins in the 3rd spin, I don't mark the first 2 spins as lost (red dash).
I only place 3 red dashes if the entire trigger loses.
Likewise, I don't mark the first spin as lost if it wins in the 2nd spin.
I consider a win in any of the 3 spins,  as a won trigger cycle.
I wanted to clarify your findings as 4 consecutive losses,  if it was 4 spins or 12 spins.
When you say 4 in a row, I take it as 12 spins (4 triggers @ 3 spins ea.).
Because I get the impression that you count each individual spin as 1 loss.

« Last Edit: April 14, 2017, 11:09:17 PM by palestis »
 
The following users thanked this post: Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #399 on: April 15, 2017, 07:08:01 AM »
Palestis, in your screenshot you've marked the spins starting at line 13 (H M M) as a trigger, but since the 3 spins prior to the trigger are LLL this breaks rules 1 and 2, so this trigger should be skipped?

And also don't forget that the tracker waits until 3 spins have passed after a trigger before looking for a new one, which partly explains why my triggers are different.

Yes, when I refer to losses I mean L = a failed trigger, which equals 3 bets.

Since you've had the tracker for a while now and haven't said that anything was wrong with the bet selections, I assumed it was ok. That was the main reason I sent it to you in the first place.
 

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #400 on: April 15, 2017, 07:14:00 AM »

What I don't understand is why there are instances that your total is negative yet you bet 0

A summary would be very useful, because we are brainstorming on various ideas and it gets confusing.

Kav, this progression is a mod of the one I used in the original sequence I posted, not the redistribution of it you posted in reply # 344. In that first progression (similar to "foolproof" and "maxim") there is no betting of zero units; that was something I added to the new divisor progression I created in response to the sequence in reply 344.

I agree, it's getting confusing. Perhaps we should stick to discussing the bet selection in this thread, which is also confusing.  ;)
 
The following users thanked this post: Reyth

kav

  • www.Roulette30.com
  • Administrator
  • Hero Member
  • *****
  • Posts: 1937
  • Thanked: 926 times
  • Gender: Male
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #401 on: April 15, 2017, 09:10:20 AM »
I see now. What you are doing is this:

Quote
Looking at the last 6 outcomes (w/L) and betting 0 units if there is less than 2 wins (i.e. at least expectation in the last 6 spins), and reverting to the standard divisor otherwise

and this:

Quote
Add 1 to stake
IF 2 * stake > (target - current bank) THEN
    stake = (target - current bank) / 2

Great ideas both.

 
The following users thanked this post: Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #402 on: April 15, 2017, 09:57:55 AM »
Because you can't have either WW or LL in this system in 2 consecutive numbers.

I agree you can't have LL in 2 consecutive numbers, but why can't you have WW in 2 consecutive numbers? This makes me think the way I'm choosing triggers is completely different from the way you're doing it. According to my understanding, this is valid:

M
M
H  Trigger = H
H  win. Now look back 3, the trigger is M
M  win.

So 2 wins in 2 spins??

A reminder of your first post in this thread:

Quote
The trigger is in the form of XXY. (XX are 2 numbers in the same dozen and Y is another dozen).
 Needless to say that XYX and YXX is the same thing as far as the trigger is concerned.
When we see this trigger in the last 3 numbers spun we simply bet the single dozen (Y), for 3 bets

I think we need to go back to basics and clear up any misunderstandings and ambiguity.
 
The following users thanked this post: Reyth

Bayes

  • Veteran Member
  • ****
  • Posts: 688
  • Thanked: 558 times
  • roulettician.com
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #403 on: April 15, 2017, 10:12:29 AM »
@ Kav, I wasn't using both those rules together in the session I just played, because they belong to different progressions, but they could be combined.

Quote
Looking at the last 6 outcomes (w/L) and betting 0 units if there is less than 2 wins (i.e. at least expectation in the last 6 spins), and reverting to the standard divisor otherwise

It might be better to look at things in terms of "cycles", where a cycle for a dozen bet is 3 spins. In that case, the above rule doesn't work because you could have WW LLLL :  Here there was no win in the last cycle, even though over 2 cycles there were 2 wins. So an improved rule might be "at least one win in each cycle. e.g. WLL LWL or WLW LLW would satisfy the rule but LLL WWL or WWW LLL would not.

 

MrPerfect.

  • Hero Member
  • ******
  • Posts: 1372
  • Thanked: 792 times
Re: NEW SYSTEM: SINGLE DOZEN
« Reply #404 on: April 15, 2017, 10:37:12 AM »
To wait for something to happen first need to sertify yourself that this somethING does in fact changes expectation. If not, why to bother?
   Stats trigger for randomly selected group of numbers ( dousen) , when it already does not hit... sounds like a wishfull thinking.