### Author Topic: Cluster DS Randomization  (Read 3782 times)

0 Members and 1 Guest are viewing this topic.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #15 on: December 12, 2017, 10:59:26 PM »
I have also vigorously tested various of his systems and they all lose.  Your point about swings in recovery I have also learned quite well and its why I don't play unless I have the highest probability (practically speaking) to hit at all times.

In my times of deep frustration trying to figure out how he wins, I have questioned his veracity as well.

Applying his recovery techniques with 99.16% chance of getting a hit has worked very well for me.  Along the way of adapting this, I also discovered a very good bet selection technique.

I simply don't feel like going back and trying his systems to see if I can do it with recovery alone or just recovery and bet selection. :shrug: Maybe I can try his methods with my bet selection and see what happens but I just don't feel like doing it right now; I have alot going on with work and am tired and such.

I learned alot from his recovery methods and I am happy with that.
« Last Edit: December 12, 2017, 11:04:14 PM by Reyth »

The following users thanked this post: MickyP

#### MickyP

• Hero Member
• Posts: 1222
• Thanked: 535 times
• Gender:
##### Re: Cluster DS Randomization
« Reply #16 on: December 17, 2017, 03:58:47 PM »
I've spent some thinking about how he has managed to win.

If you take all the known facts at the start of his approach, bankroll, bet selection, chip value etc, is it possible that his early wins simply increased the "lifespan" of the approach. In the beginning he did not have a recovery plan in place.
When systems/strategies are tested over many thousands of spins you look for a series of spins that puts the system/stetergy into a tailspin of continuous losses. When this occurs and a loss is recorded the system/strategy is labeled a failure. A question relating to this specific approach: Are tests undertaken with sufficient bankroll to reach the table limit using a constant chip value and incorporating the D'allenbert progression replicating how Charles plays?

He starts his strategy with a healthy bankroll and capitalised on good fortune to record early wins. In the process he incorporates elements to dampen losses as well as his progression giving him more space to recover. Never does he get close to the table limit. In fact, he never mentions the table limit.
Taking probability into account what are the odds that he has just been lucky enough to move his strategy into an almost "can not loose" position?

The possibility that he has a deal with the casino is very high. He is in it to make money but never withdraws funds. Will the casino allow him to withdraw the money? I wonder!  Foul play is difficult to prove in a random game of luck.
I now put this to bed but will be interested to hear your thoughts.
Can a system or strategy reach a "can not loose" point due to increased bankroll and not reaching the table limit?

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #17 on: December 17, 2017, 06:55:28 PM »
Yes, of course this thought about his early success increasing his ability to absorb the harsher blows DOES improve his overall ability to recover.

Also, the computer tests that I have done only mimic his recovery and no bankroll can withstand the worst sequences that are found there BUT the method is robotic and therefore not all is lost on that front.

The D'Alembert will never hit table limits.

The following users thanked this post: MickyP

#### MickyP

• Hero Member
• Posts: 1222
• Thanked: 535 times
• Gender:
##### Re: Cluster DS Randomization
« Reply #18 on: December 17, 2017, 07:49:15 PM »
Reyth, the D'Alembert can hit table limits if the bet is big enough but in his case at one dollar increments it more than likely wouldn't go that high .
Thanks for the response.

#### Joey

• New
• Posts: 4
• Thanked: 2 times
##### Re: Cluster DS Randomization
« Reply #19 on: December 19, 2017, 11:03:31 AM »
Reyth, r u still holding ur 2000 units using this system??? Actually how many games did u test on using this system? 500 games?

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #20 on: December 19, 2017, 11:19:10 AM »
Welcome to the forum Joey and thanks for asking!

No, what happened to me was the "stop loss" where we must stay on Level 15 and "just keep betting" and drop to Level 14 once we get a hit, got completely blown out.  That stop loss fails and was an example of Charles trying to find a simple way for the "common man" to recover without having to apply the gambling skill required to surf recovery levels.

BUT

There is still hope for testing along the following lines:

1) Switch sessions instead of selections (the dark blue areas) which is always optional
2) Tracking the hottest DS and always betting it
3) Implement actual Recovery Levels 2-5, where the unit size = the Recovery Level
4) Trying to determine a method of betting/empty spinning based on certain statistical criteria
« Last Edit: December 19, 2017, 11:23:04 AM by Reyth »

#### Joey

• New
• Posts: 4
• Thanked: 2 times
##### Re: Cluster DS Randomization
« Reply #21 on: December 19, 2017, 12:18:51 PM »
I personally tested all his system for more than 200 games. None of his system are workable in a long run. Question is: Is this Cluster workable? R u still using this Cluster DS? Will it burst?

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #22 on: December 19, 2017, 12:57:28 PM »
Yes it bursts but what is missing is going to Level 2 (Recovery) once we bottom out on Level 15 (Progression).  I still think it can be further improved with starting a new session and only betting the hottest DS.

The following users thanked this post: Joey

#### YoLo

• New
• Posts: 45
• Thanked: 13 times
##### Re: Cluster DS Randomization
« Reply #23 on: December 19, 2017, 02:26:23 PM »
But reyth, how do we recognize hot ds without using any programs like you use. That is what i am trying to do, a little tweek is needed from the original system by charles is what i am developing. Cos all of his system is failing from what i l've tested so far.

The following users thanked this post: kav

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #24 on: December 19, 2017, 08:23:12 PM »
If I was at a casino I would have a sheet of paper that had 11 rows and every time a DS hit I would make a mark on that row.  We can start betting from no history or the last 10 or w/e the marquee shows.

The following users thanked this post: kav

#### MickyP

• Hero Member
• Posts: 1222
• Thanked: 535 times
• Gender:
##### Re: Cluster DS Randomization
« Reply #25 on: February 03, 2018, 08:23:09 PM »
Reyth, have you put this system testing to bed or are you still trying to understand his wins?

I think this is good study material as it shows why machanical system playing fails.

#### MickyP

• Hero Member
• Posts: 1222
• Thanked: 535 times
• Gender:
##### Re: Cluster DS Randomization
« Reply #26 on: March 15, 2018, 08:03:26 PM »
Yes, you guessed right. I haven't put this to rest yet.

The thread "secrets in the numbers" inspired me to take another look at this thread again.

Reyth, you mentioned an important issue; finding a hot double street to bet on. If bets are randomly switched to a potentially hot double street and the expectation of hits is met then the system will prevail. Add a rule or two like reset on a new high; this will prevent the progression from getting out of hand. See it as a built in control mechanism to avoid having to depend too much on a recovery plan.
« Last Edit: March 15, 2018, 08:05:20 PM by MickyP »

The following users thanked this post: kav, Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #27 on: March 16, 2018, 01:28:29 AM »
The stronger our bet selection and wagering plan, the less successive negative variance we will experience.

This progression uses a total of 56 bets for only a little more than what I use for 27 bets.  I would need to do some statistical testing to see how often we hit that 56 bet limit.

Right now, it has to beat 99.1% for a total outlay of 443 units.  Just for an initial benchmark, the difference between 27 bets and 56 bets is:

99.16% vs. 99.99534%

The goal is to create a wagering plan that will better the statistics for a smaller cash outlay.
« Last Edit: March 16, 2018, 01:36:55 AM by Reyth »

The following users thanked this post: MickyP

#### MickyP

• Hero Member
• Posts: 1222
• Thanked: 535 times
• Gender:
##### Re: Cluster DS Randomization
« Reply #28 on: March 16, 2018, 05:16:03 AM »
@Reyth, are your stats above based on Charles' current method of betting with the static shift of the target double street?
If so, the improvement required may be reached by switching the target as soon as a potentially hot target is identified.
If you still have say 3 spins to go before the change but have identified a new (hot) target; why play the remaining spins on the target in play? Move when you want to and stay on the target if forecasts indicate a continued trend.
Following potentially hot double streets may produce the required results and will keep the progression from rising into recovery mode.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1614 times
##### Re: Cluster DS Randomization
« Reply #29 on: March 16, 2018, 06:04:50 AM »
Hot numbers alone is simply not enough to stave off the destruction.  The entire system has to be focused on it.

The first thing I need to do is find out what the straight bust out rate is; i.e. I need to calculate how often it busts out by not achieving a profit before exhausting the 55 bets.

Then I need to figure out a way to adjust it that will increase that win rate.

The following users thanked this post: MickyP