### Author Topic: The Bomb Shelter  (Read 6100 times)

0 Members and 1 Guest are viewing this topic.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### The Bomb Shelter
« on: August 23, 2017, 09:30:00 AM »
I am going to try this method with a new selection and a twist.

1) Bet DS 16-21
2) Unit size is 2 for 50 coups
3) Drop the unit size to 1 (half-unit) until the progression loses (sun deck)
3a) Begin recovery on Level 2 with a unit size of 4 units

I think this method will give me an even greater edge and I will spin far fewer times.

In fact, I am calling this one:

.

Use statistics as an air raid siren & take shelter from wheel fallout!

NOTE:  This method has undergone a great deal of transformation since this post was first made.  Feel free to review the current method HERE (overview) and HERE (latest method)

1 1 1 1 1
2 2 2
3 3
4 5 6 7 8
10 12 14
17 21 25
30 36 43
52 62 74 (99.16%) -886

1   886      886
2   1772   2658
3   2658   5316
4   3544   8860
5   4430   13290

Average Coup: 3 units

Expectation

1=7
2=13
3=19
4=25
5=30
6=37

Definition of Terms

PLE = Progression Loss Event which is when our progression fully fails after 27 spins.

PLE Ratio = This is the amount of coups until a progression loss is statistically expected; a negative number means we have less coups than expected and a positive number means we have more coups than are expected.

PMI = Progression Movement Indicator which counts the number of coups since the last PLE.

HS = Hits to Spins ratio which is positive for more hits than expectation and negative for less hits than expectation.  When this figure grows dangerously high (higher than the length of the progression), it can trigger an early PLE.

BSE = Black Swan Event which is 2 or more PLE's that appear together within less than a 50 coup span.

BSE Protocol = A time when we spin the wheel without betting, tracking specfic events until a certain statistical criteria is met.  See the Betting Rules below for when and why the BSE Protocol is implemented.

Betting Rules

1) If there have been 2 PLE's within the last 50 coups, do not bet while spinning the wheel, until 50 coups have been achieved without a PLE (BSE Protocol).

2) If the HS ratio is +27 or above, do not bet while spinning the wheel until it has corrected (become negative or low positive).

3) Switch to a different DS if the PMI or the PLE Ratio is at +150 or above.

4) Start a completely new session (comp & wheel) if there are no betting opportunities available.

5) If a next door neighbor (adjacent DS) has a BSE while betting, cease betting and enter the shelter until it passes (BSE Protocol).

6) When starting a fresh session from scratch, a) Wait until a selection has a PLE before betting & b) Only take 25 coups at Level 1 (PMI) instead of 50.

7) If our neighbor has a PLE, we must either reset the session or enter the shelter until a KoH with 2 strong neighbors appears.

DEMO_VIDEO_1
DEMO_VIDEO_2

+304 <=== goal achieved

Dropping down to 1 unit size bets (sun deck).  Let's see how far we can get!

+450 <=== benchmark achieved

Ok, so by estimating the average coup amount, this should be about 100 coups; I did do several snap backs and so it is problably a bit less than that.

At this point I need to design the softwares so I can track the data points.

mmkay?

So we continue onward -- let's see how far before the blast hits!

Hit +600 which we can see here is about 35 coups.  At one point, the HS ratio was 1 spin from triggering the BSE protocol but it reversed itself from that point and now stands at a dangerous looking 20 (27 will invoke the protocol; i.e. the length of the progression).

We also were only 4 spins from a PLE at one point.

The sun deck is pretty safe, with favorable statistics even if we take a PLE but the only reason I am not in a BSE protocol is because I haven't had a PLE yet and I feel like I just want to see how far we can go.

-443/+190(?)[0]L13

Ok so we finally got hit and it was on the sun deck just like we wanted!  So now the statistics are extremely favorable for us and we will drop to Level 2 which is a unit size of 4 units and bet until we recover or hit 50 coups, whichever comes first; its amazing but full recovery is expected in only 28 coups!!!  This is the AWESOME power of The Bomb Shelter!

Level 2 Recovery.  Target: +475

+292/+482 (-40)[17]L17

Ok, so already scrapped it back to over 75% of the debt!  I post here to illustrate a point where we can lower back to Level 1 and a unit size of 2, so we can lock in our profits and recover the rest of the debt in greater safety.

Because my first around 100 coups were without data tracking, I think the conservative thing is to go back to Level 1 now, because the true statistics must suggest that a PLE is more likely than we see (the PLE ratio is more advanced than currently showing, in that more than 100 coups were achieved before the PLE hit).

Level 1 Recovery. Target: 506.33

-886/-957 (-28)[0]L61

Ok first thing I need to note is that the PLE ratio is not displaying correctly; it should be lower than -100 but is only showing -28 -- I know why this is and its just changing a single figure in the calculation.

Here we can see that we scrapped back to +562 which is only 71 units from full recovery.  So it appears the conservative route was the best choice here.

Based on the more than favorable start, the PLE looks more or less due at this point (the PLE ratio would be +0 or greater, if I had been tracking it).

Because of the greater efficiency of The Bomb Shelter recovery system, our losses are much less than we would have normally experienced (because we gained our profit twice as fast and then retreated to lock in that profit); we would be at -1914 instead of -957.

We now have a BSE and must implement the BSE Protocol to insure that we aren't going to get successively swarmed.

Because I must take the softwares down to correct the problem, I will lose the data calculations and we will perform the BSE from scratch.

-1772/-2729 (-98)[0]L20

OMGOSH!  This is the kind of sequence that DESTROYS systems!  But we are safe and toasty warm in our Bomb Shelter!  Had we bet this sequence normally from the beginning, our debt would be -3686!  But here we are nice and comfy at only -957.

We need for this BSE to pass before we can continue betting.

Ok, the BSE has passed.  The HS ratio is not good at +17 which signals a need for some smaller correction (later hits).  The PMI is relatively safe at +50 but only because of the PLE ratio which is actually -48.

I think this warrants 25 coups to recover some debt which will bring the PMI to 75.  The HS ratio is in a downswing but if it turns back and goes the wrong way, I will bail out at the peak & hit the shelter until it corrects.

This is NOT an ideal situation to be betting and I would prefer not to but because of all the recent increased negative activity and some small leeway in the PLE ratio, I am willing to risk the next 25 coups.

Level 2 Recovery.  Target: 75 coups (PMI)

+80/-877 (-44)[54]G27

WELP!  Ok well that is exactly what the HS ratio did and so here I am back in my shelter until it corrects.  BSE Protocol initiated.

Invoking Rule #3 as the PLE rose to +150% and the selection has become non-conforming (free radical).  This means that PLE's can hit and stack at any time.  Just like financial professionals tell us to stay out of volatile markets (many large swings), this selection is off the rails and will not allow us to bet without great risk.

Switching to DS 1-6 and taking a PLE with the BSE Protocol.

Having to invoke Rule #3 AGAIN since the PMI rose to 150% on DS 1-6 too!

Switching to DS 31-36, BSE Protocol.

>.<  Did you know our Shelter has lighting?  Yes!  Unfortunately I haven't connected the wiring yet but I am working on it now!  Soon, there will be no need to stumble around in the dark, fishing for PLE's -- the PLE's will simply be visible to us!

Bomb_Analyzer incoming!

A very important side note about the REC.  I would like to stay on a single DS for my entire betting career but unfortunately numbers become non-conforming at certain times and it is no longer safe to continue betting them.

The key is that I am only going to switch when the PMI or PLE ratio is at 150% or above (Rule #3); at that point the swings are too great to control the risk.

Level 3 Recovery.  Target: 503.93

+720/-237 (-66)[44]G5

This is really not a bad situation BUT I took a good number of coups PRIOR to this number failing that I lost the data on, so I am going to hit the sun deck and see if I can just grind my way back.  This way I am heavily locking in profits in a statistical situation that is not as good as it looks.

Level 1 Recovery.  Target: 506.33

A very interesting confluence of events where 25 coups (the amount I was planning to stop at) corresponds almost perfectly with a HS ratio of 27 (it is 26) AND I am 1/2 a unit above the original drawdown amount!

So I am going to go into the shelter at this time.  Normally this would be a great time to hit the sun deck but due to the unknown actual PLE ratio and a KNOWN PMI of close to 75, I just think its best to take shelter now and see what develops.

+89/-148 (-41)[69]G26

BSE protocol implemented.

I watched the HS ratio go through the +27 threshold, recover and go back up again.  The second time the PMI was at 137 and it just simply crashed right to the ground.  The ACTUAL PLE ratio here is +27 which is overdue.

I have two choices.  One, I can stay in the shelter and watch this take a quick PLE and jump on as soon as it does OR two, I can simply bet 25 coups on DS 1-6 which has a VERY healthy PLE ratio of -108.

I would prefer to recover this debt now in a safer bet selection and we can watch DS 31-36 from the sidelines.

Level 2 Recovery.  Target: 506.33

+485/+663 (-96)[59]L21

Alot going on with this number.  First of all, the short-term trend, the HS ratio, is -21 which is very favorable.  Secondly, the PLE ratio is extremely favorable at -96 (almost like a fresh PLE).  However our PMI is at 59.

So the crux of this analysis is the degree that this number will conform to the long-term trend of the PLE vs. the static PMI measurement.  The 50 coup benchmark of the PLE is an artificial figure that is designed to limit our statistical exposure, the actual due point of the PMI is 100.

This consideration illustrates how the overall trend of a selection's performance (upward or downward) will determine when a PLE will occur (earlier or later).

Scenarios

1) The trend is still downward -- the PLE will occur prior to the PMI reaching 100 and the PLE reaching 0.

2) The trend is now upward -- The PLE will not occur until either the PMI reaches 100 or the PLE reaches 0.

Right now all 3 indicators are showing that the trend is still strongly upward (the negative HS is calling for strong correction from the slow short-term swing).

Because of this I will stay in the grassy forest area at Level 1 and a unit size of 2 until I hit 25 coups or the HS ratio corrects.  After that I plan to hit the sun deck.

Oh, and how is DS 31-36 doing?  It was AMAZING it has jumped to a HS of +34, slamming 22 coups versus only 10 coups of our DS 1-6!  I say it is seriously cruising for a bruising and should happen shortly!

+84/+747 (-80)[75]L16

So I have been thinking about this situation.  Our normal method of minimizing our risk is to use a PMI of 50 during favorable long-term trends (PLE of less than -50, usually closer to -100).  So this move is objectively aggressive at 75 coups.

We have heard how traders make money in down markets by surfing the upward spikes in the overall downward trend.  I think this exactly what we are doing here but in order to be safe at the same time, we need to limit our overall exposure; this allows us to still be conservative while also making profit.

What we are doing is ASSUMING that the long-term trend is STILL DOWNWARD.  This is the proper perspective, and thus we would be expecting an early PLE at around or even before -50 on the PLE ratio.

So, I am going to hit the sun deck for 25 coups and I plan to enter the shelter after that to wait for developments.

DS 31-36 is STILL off the rails at +52!  At one point it was near +70!  Even if it loses now, it will be unplayable until it has a substantial short-term correction.

You know I just NOW realized how I was actually INCREASING my risk by only playing 1 number!  Playing 6 numbers is like using a mutual fund!

I just thought of a new statistical analysis!

How many times can a BSE be expected to consecutively stack in the following pattern: LLWL?

Followed by WLWLWLWL etc.

This would also include patterns with losses that are strung together like LLLLWLLLWLWLWLWL etc.

The answer to this question may very well show how we have beaten roulette!
« Last Edit: October 16, 2017, 03:06:01 AM by Reyth »

The following users thanked this post: kav, Jake007

#### TERMINATOR

• Search YouTube for MANDELA EFFECT
• Mature Member
• Posts: 263
• Thanked: 328 times
• Gender:
• MANDELA EFFECT
##### Re: The Bomb Shelter
« Reply #1 on: August 24, 2017, 06:32:52 AM »
Awesome NAME for this, Reyth. Can't wait to see your video demo of the BOMB SHELTER. Cool.

« Last Edit: August 24, 2017, 06:39:13 AM by TERMINATOR »

The following users thanked this post: kav, Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #2 on: August 24, 2017, 06:37:03 PM »
Videos posted!

I demonstrate how to do an SBE on the video.

The following users thanked this post: TERMINATOR

#### Kreuzritter

• New
• Posts: 2
##### Re: The Bomb Shelter
« Reply #3 on: August 24, 2017, 08:52:41 PM »
Where can I find the Videos?

Thx

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #4 on: August 24, 2017, 10:34:03 PM »
In the first post scroll down and you will see the red links entitled:

DEMO_VIDEO_1
DEMO_VIDEO_2
« Last Edit: August 25, 2017, 07:33:34 PM by Reyth »

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #5 on: August 26, 2017, 12:19:11 AM »

Ok, let's see if this thing still has any bugs...

Bomb_Analyzer_DEMO

Sun Deck Demo

+97/+844 (-50)[105]L2

Ok, I got a nice sun tan and its time to go into the shelter and see what develops.  The PMI is calling for a PLE but of course its just a static estimator whereas the PLE ratio itself has produced and extra PLE by 50 coups.

Let's see if the overall down-trend has turned around.

Ok, I am making a command level decision here to write DS 1-6 off at this point, it is clearly very close to off the rails.  I am going to switch to DS 22-27 that just came in.

This is a slow number with a -77 HS.  I don't have any rules regarding this and so I think it will also be a good education.  Obviously the HS ratio is very strong at -89 and so I figure I might take 40 coups or so at Level 1 and then maybe retreat to the sun deck.

+215/+1059 (-50)[39]L138

This was a very nice run and was quite educational.  At first it was quite nerve wracking because the numbers kept going deep into the progression before hitting.  After awhile of this though & they started heating up significantly; at one point I got like 5 hits in a row!  That's kind of what I was hoping for, that I would be on the selection when it heated up.

Well, I am looking at the -50 PLE and just thinking to myself, let's hit the sun deck and take 11 coups from there, just in case.  The very low HS ratio does not prevent the PLE from hitting and I just feel more comfortable staying on "the good side" of the PLE ratio.

Let's take a look at the board for a bit:

DS 19-24 just had a BSE and it is the only one I have seen in almost 1700 spins which shows how rare it is.

DS's 31-36 & 10-15 OMGOSH Rinad would have had a field day with these, staying overnight and walking out with many thousands of dollars!

DS 1-6 is still chugging along and is quite off the rails.

DS's 4-9 & 10-15 are the elephants in the room having no PLE events in almost 1700 spins!!!

Every other selection just sucks, all them being severely overdue for what appears to be a BSE-fest in our near future.

+42/+1101 (-36)[53]L146

I took a few extra hits on the sun deck so I could get my profit up to +1100 which is the luxury afforded by being able to bet half units.  I get this feeling like I should be more aggressive with the sun deck (or at least I could) rather than retreat to the shelter but I'd rather just be conservative and completely avoid losses.

You might not be able to tell, but take my word for it, DS 31-36 just had a PLE (finally!!).  I fully expect a BSE forthcoming.

DS 19-24 is closing in on ending its BSE and looking pretty decent for a bet at that time, but unfortunately will be at the tail end of where I feel comfortable.

Hitting the shelter.  BSE Protocol implemented.

+236/+1337 <=== (-50)[61]L246

Another nice run where numbers ran slow but heated up at certain times; its an ebb and flow when they heat up.

You can guess what happened, DS 22-27 had a PMI over 100 and a PLE ratio of over 0 and DS 19-24 came in and was highly bettable.  So I switched.  I took 50 coups at Level 1 and then about 11 coups from the sun deck to bring the PLE ratio to -50.

BREAKING NEWS.
DS 10-15 has finally fallen!  Of course its highly unbettable, screaming for a BSE.

DS 22-27 did fall and was bettable but wasn't as good as DS 19-24 when it did (and I wouldn't have switched anyway).

DS 7-12 had a BSE and recovered successfully from it -- still highly unbettable though.

DS 1-6 & DS 16-21 are Rinad's long hoped for dream!

Entering the shelter to see what happens out there, in safety and comfort!

(-146)[0]L269

It might have been but 40 spins before the PLE hit.  Awesome time to be betting!

+427/+1764 (-71)[75]L256

I took 75 coups, 50 at Level 1 and 25 on the sun deck.  I think this is still worth 21 coups from the sun deck but I figure why push it; the number hit the PLE early which means it is in a downswing.

DS 4-9 finally fell!  Almost 3000 spins without a failure is close to 18000 units!!!

DS 22-27 had a PLE which was statistically better than DS 19-24 but I stayed where I was.

DS 31-36 had a PLE and it still is almost impossible to tell!

DS 1-6 also finally had a PLE.

The only bettable numbers on the board to bet are DS 19-24 & DS 22-27 all the other numbers are off the rails with overhwelmingly positive variance.

So going back into the shelter for some zzz's.  See you when I wake up!

So why do we bet only 50 coups on the PMI?

Watch until I bring up and then stop talking about the Excel graphic.

But there is another answer.  If we only bet at 1 unit for the first 50 coups after a PLE, the only thing that can defeat us is a statistical anomaly (a BSE), otherwise we win every time.

+379/+2143 (-50)[69]L107

Amazing run!  I hopped off DS 19-24 onto DS 22-27 only a few spins later to have DS 19-24 have a PLE that was statistically better!  I stayed where I was and watched DS 19-24 outperform my chosen DS (speed of hits).  Then to my horror I watched it take a PLE (I would have been on the sun deck)!  I mean it's right next door, sharing 3 numbers, so quite unnerving!

Everything is fine though and going to hit the shelter and await the latest news!

Hey, remember when I said:

Quote from: Reyth
Every other selection just sucks, all them being severely overdue for what appears to be a BSE-fest in our near future.

Ya, well this is what it looks like.  Notice every one of these are totally unbettable and have been getting hits for an extended period far greater than statistically expected.  This is how roulette corrects the situation, with more statistical anomalies.

Our good fortune comes at a hidden price and unless we are able to avoid it, we will get slammed!

Very soon after this photo was taken, DS 4-9 and 7-12 joined in.

Wow.  So I just watched from my shelter as my next door neighbor (DS 19-24) got blasted by 2 BSE's in a row (x spins apart but without intermission)!  I would have been on the first one for a loss of -886 and been inside for the 2nd one and would be in the shelter right now waiting for it to clear.

Now, I just watched a PLE hit my DS, from the safety of my shelter.  Since it shares 3 numbers with DS 19-24, I want to make sure it is the right thing to start betting.

I have no statistical explanation for why the BSE hit, it is the chaos that we are all guaranteed to experience due to the random walk & the law of large numbers.  However, it IS a statistical anomaly and therefore is inherently limited and I believe more likely to break than to continue.

I actually think its even better to bet DS 22-27 right now because of this freak event and I expect a greater amount of fast hits.

Let's see where it goes.

Bizarre BSE Confluence

Alright, I have had some time to think about this and it is now clear to me how inconsistent it is to bet a selection where 50% of it is under BSE assault and by the Betting Rules, the other 50% of the affected area must stay in the shelter.

Therefore I am adding Rule #5:

If a next door neighbor (adjacent DS) has a BSE while you are betting, you must cease betting and enter the shelter until it passes.

Ok, so clearing these BSE's from the shelter...

Level 2 Recovery.  Target: +2077

+234/+1647 (-492)[75]L323

So ya unbelievable but DS 19-24 took not four, but FIVE BSE's in a row!  By the time it cleared, it was superior to DS 22-27 and its neighbor (DS 25-30) has a BSE going as well, so I switched and scrapped 25 coups at Level 2.

Interesting note, DS 16-21 has recovered and is now slightly bettable.

I am taking shelter to see how things shape up.

+655/+2302 (-178)[80]L71

Man that number heated up!  Check out the move from -323 to -71!!

I took about 35 coups at Level 2 and hit my target and so lowered back to Level 1 for about 15 coups and then I  scrapped 25 more coups (75 total) on the sun deck and was so close to the all-time high that I just went the rest of the way; that's the beauty of the sun deck!

When starting a fresh session from scratch, a) Wait until a selection has a PLE before betting & b) Only take 25 coups at Level 1 (PMI) instead of 50.

DS 25-30 has been fully restored and is highly bettable!

DS 16-21 got nailed by a BSE while it had favorable statistics; this appears to be an over-correction blast from the past.  Quite a scary notion...

Good night from Shelter-Land! zzzz....

-886/+1436

Wow, can't even believe this.  There might be grounds for this correction due to the large HS move prior?  Maybe we can make an indicator for this?  I will keep this in mind...

Clearing it from the shelter.

Unfortunately, I left my comp unattended for about an hour and came back to find it had updated itself, losing all the spin data I had.  Therefore I am starting a fresh session and of course keeping the same bankroll info and such.

This might be interesting for hit and run fans to see how the break in sessions might affect the statistics!

Level 2 Recovery.  Target: +2101

+550/+1986 (-51)[50]L19

I took 25 coups at Level 2 & 25 coups at Level 1.  This Rule was implemented due to the lack of pre-existing statistics to work from.

Btw, I was thinking about the large HS move some more and the idea of a HSMI has re-presented itself.  This would be a movement tracker that would count the difference in the HS score over a rolling [certain number] of spins.  If that value would excceed [a certain amount], the BSE Protocol would be implemented.

So anyway, back to the shelter to await events.
« Last Edit: August 29, 2017, 05:18:39 AM by Reyth »

The following users thanked this post: TERMINATOR

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #6 on: August 29, 2017, 07:49:28 AM »

-443/+1931 (-205)[0]L134

This was a full recovery where I hit the target about halfway through, lowered to Level 1 and hit the all-time high.  I then went to the sun deck to get 25 more coups but got only like 5 or so before it failed.  No reason why it would do that but it has increased its downward trend.

This is why we have the sun deck so we can minimize our exposure while taking additional profit beyond the 50 coup mark (PMI).

It might be hard to notice without a chart (I'll be throwing one up soon), you can see how my recovery inclination is set upward where my losses are successively increasing in profit amount; this is normally accomplished by setting a 75% target goal and then lowering the chips to lock in profit.  The sun deck allows me an increased ability to do this.

I have also been thinking more about the HSMI.  The PLE ratio has a movement indicator to make sense of the fluxuating expectation performance (PMI) and so should the HS ratio.  The movement indicators is the way we can make sense out of expectation which is so influenced by variance; the movement indicators are static measurements.

Here is the HSMI criteria:

If the HS ratio moves more than 166 points in a 50 coup period.

166 points is (hit expectation[6.17] * progression length[27]).  The reason for 50 coups is because it is the maximum length that I will leave my chips fully raised.

I am planning on taking the softwares down and implementing the HSMI, where if any number has this large HS ratio move, it will display an "H" next to the HS ratio; this way I can monitor this indicator to see if it is relevant in predicting PLE's/BSE's.  First I need to go to sleep though.

Level 2 Recovery.  Target: +2374

-1772/+415 (-289)[0]L142

Such is roulette.  Now what is required is patient, skillful profit recovery.  Clearing this BSE from the shelter.

Level 3 Recovery.  Target: +1552

+474/+889 (-271)[75]L185

After the BSE cleared, I took 25 coups at Level 3.  That is as far as I am willing to go, as this move is more than aggressive enough.

Interesting note that DS 13-18 took a PLE while I was recovering.  I consider a neighboring PLE a positive event as long as it doesn't turn into a BSE.

Watching the direction of the trend from the shelter.

+690/+1579 (-180)[27]L233

Target achieved.  Took 27 coups.  Now dropping to Level 2.

Level 2 Recovery.  Target: +2176

+397/+1976 (-132)[75]L198

Took 50 coups at Level 2 & 25 coups from the sun deck.  Hitting the shelter to see what happens.

+208/+2184 (-199)[18]L331

Target achieved.  Dropping down to Level 1 for the rest of the debt or 50 coups, whichever comes first.

Level 1 Recovery.  Target: +2374

+209/+2393 (-167)[50]L311

Interesting phenomenon where DS 28-33 & 31-36 have been taking repeated BSE's and it feels like they are the "scapegoats" keeping them away from me on DS 16-21.

There is definitely a phenomenon about "side-by side" BSE's but whether that also affects the chances of a BSE elsewhere on the board, I simply don't know...

There also seems to be this pattern where the whole board is wildly positive above expectation and only a very few are below expectation on PLE's.

I have also noticed how there will be maybe 2 and then 1 DS that will go for thousands of spins without a progression loss; DS 4-9 went 499 coups as the last DS standing before it fell; that would be about +1500 units, with the loss at -443 from that amount.
« Last Edit: August 30, 2017, 07:00:57 AM by Reyth »

The following users thanked this post: kav

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #7 on: August 30, 2017, 12:24:18 AM »
As a side note here, I was just reading a post by Jerome and realized that this system can be played with faster & greater profits, less expensively and with a higher win rate by just playing the EC's:

1 2 4 8 16 32 64 128 (99.52%)

1   255      255
2   510      765
3   765      1530
4   1020   2550
5   1275   3825

1 unit    =    10 chips [8](6)(4)(2)
sun deck  =    5 chips   (4)(3)(2)(1)

Average coup: 10 chips (8,6,4,2)

To achieve a higher average coup, we need to play with 10,8,6 or 4 chips as the unit size.

I can't for the life of me think why I should play DS's here when EC's are better in every way, including the maximum possible number diversification?

Ok, when multiplying the bankroll amount by 4, I see it'a actually more expensive.

643 = progression
1036 = max
« Last Edit: October 31, 2017, 07:43:58 AM by Reyth »

The following users thanked this post: TERMINATOR

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #8 on: August 30, 2017, 07:19:59 AM »
I must process the following fact when I am awake enough to think:

The chances of a BSE is exactly the same as for a PLE event; it is no more rare than a standard progression loss.

I need to figure out the impact of this information on the system as I am playing it now and exactly what changes I need to make.

Things that must remain the same:

1) Sun Deck -- this will now be used as a "do or die" platform for betting.
2) Shelter -- this will now only be used for a BSE.
3) PMI -- this counter is necessary for knowing when to hit the Sun Deck and when there is a BSE.

No longer necessary:

1) Tracking the PLE ratio -- it still shows the speed of PLE's so leave it in

Questions
Do I need to track the HS ratio? -- why not it shows the speed of hits
Do I need to worry about bet selection? YES!
How relevant are the BSE's showing up on the display? EXTREMELY!!! (see below)
Do I need to track the whole board? YES!
Should I track the HSMI? -- why not it might work
What else should I be thinking of?

New strategy! When a BSE hits, it poisons the DS('s) next to it.  Spin until a BSE hits and it limits the choices available.
« Last Edit: August 30, 2017, 08:04:24 AM by Reyth »

The following users thanked this post: kav, TERMINATOR

#### dobbelsteen

• Hero Member
• Posts: 1560
• Thanked: 544 times
##### Re: The Bomb Shelter
« Reply #9 on: August 30, 2017, 12:48:17 PM »
Reyth why do you use tools a very few members have. Text as well as the pictures I can`t understand. This is killing any discussion of the topic. Abbreviations without explanations are killing. Such threads are only for insiders.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #10 on: August 30, 2017, 01:53:49 PM »
Quote from: Reyth
Definition of Terms

PLE = Progression Loss Event which is when our progression fully fails after 27 spins.

PLE Ratio = This is the amount of coups until a progression loss is statistically expected; a negative number means we have less coups than expected and a positive number means we have more coups than are expected.

PMI = Progression Movement Indicator which counts the number of coups since the last PLE.

HS = Hits to Spins ratio which is positive for more hits than expectation and negative for less hits than expectation.  When this figure grows dangerously high (higher than the length of the progression), it can trigger an early PLE.

BSE = Black Swan Event which is 2 or more PLE's that appear together within less than a 50 coup span.

BSE Protocol = A time when we spin the wheel without betting, tracking specfic events until a certain statistical criteria is met.  See the Betting Rules below for when and why the BSE Protocol is implemented.

Betting Rules

1) If there have been 2 PLE's within the last 50 coups, do not bet while spinning the wheel, until 50 coups have been achieved without a PLE (BSE Protocol).

2) If the HS ratio is +27 or above, do not bet while spinning the wheel until it has corrected (become negative or low positive).

3) Switch to a different DS if the PMI or the PLE Ratio is at +150 or above.

4) Start a completely new session (comp & wheel) if there are no betting opportunities available.

5) If a next door neighbor (adjacent DS) has a BSE while betting, cease betting and enter the shelter until it passes (BSE Protocol).

6) When starting a fresh session from scratch, a) Wait until a selection has a PLE before betting & b) Only take 25 coups at Level 1 (PMI) instead of 50.

My betting methods don't appeal to very many people because of the higher bankroll requirements (I play .01-.10 tables) and my methods are only for online play.

Anyone can request a copy of any of the software that I am using, in this thread and obviously provided free of charge for all.

What are the chances of gaining a coup after 50 successful attempts? 64.43%

What are the chances of getting a PLE within 50 coups? 36.57%

The chances are very similar to a 2 dozen bet or a single number for 37 spins.

What is the average earnings for 50 successful coup attempts? 150 units

The bizarre thing is that the odds of getting a PLE earlier in the 50 spin window are slightly higher than the 50th spin and slowly decrease as it gets deeper in the 50 coup range; I think THIS is why we see the vast majority of numbers wildly positive vs. expectation.

Its like the longer we go, the greater chances we have to keep going!?

I mean its slight like the difference between 8500/1M and 5500/1M -- .85% vs. .55%

I will attach the distribution so there is a record.

My assumption that there is any kind of increase in statistics due to a 50 coup window (1/2 expectation) is incorrect and is very slightly the opposite of what I expected.

Its almost as if it would just be simpler to do the following:

Take a single coup at 1 unit (2 chips) and then immediately hit the Sun Deck for 1/2 unit (1 chip) and keep betting the Sun Deck until it loses.

0  8631 0.008631
1  7207 0.015838
2  7914 0.023752
3  8210 0.031962

So betting 3 times at 2 chips is slightly better than playing 36 numbers at once for a single spin.  We should rarely see it fail and then every other bet is at 1/2 a chip (sun deck) even though we have a bankroll for a full 2 chip failure.

The only other consideration is to avoid being next to a BSE or a HSE (if implemented) by changing our bet selection at those times.

As far as what that selection should be, I think it should be the fastest number on the board that isn't threatened by a BSE/HSE and we shouldn't switch numbers until we become thus threatened.

EDIT: I have also realized that our bet choice should also consider its immediate neighbor(s), making sure that we aren't next door to a slow number.

Bomb_Shelter_UPDATE

+191/+2584 (+569)[76]G473

The new system is working great.  As in the update video, I chose the wrong DS with 7-12 and eventually it had a progression loss & slowed down and as you can see its neighbor DS 10-15 is a strong downward pull, being slow.

So I scrapped it back up to 75% of the debt on Level 2 and then switched to DS 4-9 for the rest of the debt at Level 1 and it hit much better.

So now I will just hit the Sun Deck for as long as I can on DS 4-9.

Let's do a quick board review with the new system:

The bottom row is dominated by slow numbers, where DS 28-33 is just barely acceptable and DS 25-30 is the selection of choice there.

Similarly, the upper right corner is completely slow with DS 7-12 being compromised by its slow neighbor DS 10-15.

Clearly the area of choice here is the upper left corner, where DS 4-9 is the best choice, being the fastest selection on the board and having 2 strong neighbors.

The speed of a number is shown primarily by its HS ratio (the top number).  Secondarily, one can look at the PLE ratio to see if it has been hit much by progression losses (the bottom number).  The middle number simply shows how long its been since the last PLE and is not necessarily relevant in our bet choices.

+74/+2658 (+347)[257]G239

Ok so I am documenting this profit level for two reasons.  First, I have reached an important benchmark, I have earned in profit the cost of 2 full recovery levels!

Secondly, that I have switched numbers.  There is a pattern when a number slows down where it will take a later hit in the progression (all selections will do this at some point) but then the next hit will also be later.  Meanwhile, another selection will hit at least once during each of the slower hits.

I saw this pattern with DS 1-6 and so I simply decided to switch.  DS 4-9 is hitting steadily but now more slowly, so it still makes a great neighbor.

Looking at the board we can see that the bottom row is now transforming into a hot section with DS 22-27, 25-30, 28-33 all becoming good selections.  DS 25-30 is still the bet of choice there with 2 strong neighbors.

I should have probably moved down there instead but so far DS 1-6 is doing fine.

The HS & PLE ratios are almost like "RNG wheel bias" detectors.

Take DS 1-6 for instance, as of right now we are showing an HS ratio of +285 (285G), that is 285 less spins for the number of hits we have.  The PLE ratio is showing +381, that is 3.8 times more coups than is statistically expected.  Of course we see that the PMI is 291 which is 291 coups since the last progression failure.

This is clearly an RNG "wheel bias".  What this actually means is that this group of 6 numbers is consistently, over a long period of time (the last 4556 spins), hitting SIGNIFICANTLY faster than expectation; again and again these numbers return to being hot in short cycles.

DS 4-9, 7-12 are also good examples.

So we can see that the entire section of the board, numbers 1-12 are consistently hot.  So this is like a hot sector on a wheel.

Let's look at DS 22-27, 25-30 & 28-33 for a moment.  Sure they are hot now, as reflected by the positive HS ratio on all of them but DS 22-27 & 28-33 are quite low on their HS ratio at only around 50, which means they are only recently hot as opposed to consistently hot.

DS 25-30 is obviously consistently fast with a 341 HS ratio BUT its PLE ratio is only 87 which means that even though it is a fast selection, it has been plagued by progression failures (consistent large gaps).

Therefore these DS's are not a "RNG bias" but are probably just a normal statistical flow.  For them to become a "RNG wheel bias" they will need to show stronger numbers across BOTH the HS & PLE ratios.

Btw, just a side comment here.  I hope you noticed how DS 13-18, 16-21 & 19-24 (remember they are still connected!) are thoroughly corrupted with DS 16-21 having a progression failure just 10 coups ago.  The order of events was a PLE in 16-21, followed quickly by the BSE on 19-24 and then the BSE on 13-18.  We need to avoid this section of the board (DZ 2) like the plague!

+1664/+293 (+199)[66]G454

Ok, time to update on my progress.  It started out with a progression loss on DS 1-6 and I was noticing that DS 25-30 was heating up and hitting more frequently.  Around 15 coups later it had a BSE.

So I decided to switch to DS 25-30 and I switched back and forth from DS 1-6 & DS 25-30 based on the switch pattern I mentioned above.  After awhile DS 25-30 had a PLE but started heating up so much that I gained many hundreds of units in a short time.

Soon after that DS 1-6 had a PLE, followed by DS 4-9 and then DS 1-6 had a BSE.  This is clearly a "changing of the guard" regarding this section of the board.  The kewl thing was that it was in a sense "announced" beforehand when the PLE first hit on DS 1-6 while DS 25-30 was hitting very well.  Now DS 25-30 is the fastest number on the board.

I have hit my 75% goal based on the 508.43 debt figure that you can see on Level 3 of the recovery tier.  So its time to go back up to Level 3 and I will set a new target of 520.04 which is 75% of the Level 2 figure of 523.91.

So far this is the worst downswing in my entire history of playing this method, the total swing was -4205.

Level 3 Recovery.  Target: +2004

What is the maximum number of coups that we can expect before a PLE hits? 1401

What is the % chance of a PLE hitting at any one time based on the number of coups? below

13 coups = 10%
27 coups = 20%
43 coups = 30%
61 coups = 40%
82 coups = 50%
108 coups = 60%
141 coups = 70%
189 coups = 80%
276 coups = 90%
357 coups = 95%
419 coups = 97%
544 coups = 99%
1401 coups = 100%

Also attached for full details.

This chart is very scary.

I know that using the shelter will lessen my profit before the next BSE hits but it just seems like the sanest path to only bet during the most favorable times for PLE failures while in recovery.

+1740/+2033 (+209)[17]G493

Target achieved.  Now moving up to Level 2.

The bias continues!  After 7025 spins we still have DS 4-9 & 25-30 dominating the scene; DS 25-30 has clearly taken over as the hottest section on the board with 2 extremely strong neighbors.

Right now I am only taking 27 coups (20% chance of failure) and spending the rest of the time in the shelter, waiting for PLE's from fast numbers.

Anyway, going to take 10 more coups at Level 2 and hit the shelter again.

What is "RNG Bias"?

It is the "king of the hill" of the whole board.  It is a DS that will hit far faster than expectation, have far more coups without a loss and its neighbors will both share similar statistics.  This phenomenon will continue for hundereds and thousands of spins.

There will always be a king of the hill (KOH) in every spin sample.  Changing sessions will change the KOH.

A partial reason this phenomenon occurs is that the RNG doesn't provide perfectly balanced distribution; some numbers must be losing more than expectation and others will thus be winning more.

It is the same "force" or reason that causes the entire board to be not perfectly balanced , which causes the hits to stack faster and in greater number and for a longer period in a particular DS more than any other.

I scrapped it back to Level 2 and there was a progression failure.  I scrapped back some more and then fully reset the session.  I am now back at Level 2 with a target goal of +2627.

Every time we reset the session we get a new KOH and here it is clearly DS 31-36, its neighbor being quite strong.

There is also a Queen of the Hill (QOH) in DS 22-27.  I can't be sure but I think we are in the middle of a "changing of the guard" where the Queen & King swap positions.  I think this because DS 31-36 had 2 outside of expectation hits in a row while DS 22-27 grew faster.

I will observe from the shelter to decide which one to bet with.  I think DS 31-36 will win out here because of the slow QOH neighbor in DS 25-30.

One of these 4 DS that are left, are likely to go multiple hundreds of hits without a PLE.

Level 2 Recovery.  Target: +2627

+624/+2627 (0)[267]G200

What I predicted is exactly what happened -- first the weak QoH neighbor had a PLE followed immediately (within 5 spins) by the QoH herself.  After that the KoH just kept hitting and hitting.

At a certain point the QoH tried to make a comeback and it was clear that she was winning the battle, so I hit the shelter and watched the KoH drop to 160 HS.  I decided to wait for the numbers to prove to me that the KoH was going to maintain its dominance and not have a PLE by reaching 180 -- this happend while the QoH declined, which is a sure sign that dominance is in place.

As long as the KoH remains dominant, it CAN'T have a PLE and if it does have a PLE, there will be a reason.  In the above case I was protecting against the "changing of the guard" where King and Queen trade positions.

Now we have a new QoH it seems, in DS 19-24.  Now its a battle for who will be the last DS standing and again I predict it will be DS 31-36 due to the weak QoH neighbor DS 16-21.

Raising to Level 1 to achieve the full debt recovery.

Level 1 Recovery.  Target: +2834

Full Recovery & DEMO

Sorry for the long length but I go into depth on my thought processes while playing and scrap back all the debt.

+210/+2837 (0)[48]G72

This was a real battle royale trying to stay with the flow of the random sequence.  This is typical of early sessions where the roles have not yet been clearly defined and various "factions" are fighting it out for KoH!

Unless its neighbors come on strong, I predict that DS 31-36 will become the ultimate KoH but clearly DS 10-15 is now holding that title.

Hitting the sun deck to see what happens.

Video Update

+163/+3000 (0)[107]G82

My prediction came true, DS 10-15 collapsed into a pile of dust, DS 1-6 hasn't succeeded and DS 31-36 is the KoH!

DS 1-6 has had its toll though.  Because of its almost constant sniping, it has held my HS ratio down.  This is a scenario where the QoH has a rival princess; a 3rd selection that contests for the Queen spot.
« Last Edit: September 02, 2017, 11:30:03 PM by Reyth »

The following users thanked this post: TERMINATOR

#### TERMINATOR

• Search YouTube for MANDELA EFFECT
• Mature Member
• Posts: 263
• Thanked: 328 times
• Gender:
• MANDELA EFFECT
##### Re: The Bomb Shelter
« Reply #11 on: August 31, 2017, 07:12:49 AM »
Thanks for keeping us updated on your progress, Reyth. Seems like you're actually improving your method to the point you can minimize losses and increase winnings. I enjoy reading your posts and learning what new things you've discovered.

I like your idea of a 75% recovery method before going back up a level. Is this what D-Alembert talked about? Or is this what you learned on your own?

You also said this new system is so easy you could play it in a brick and mortar room. The BSE's being the only thing you really can't track there. I wonder if they allow players to keep track of roulette numbers on cellphones...then we might be able to keep track of BSE's.

Anyway, I think it would be great if your new method could be played not only in B&M's as well as online, but could also be used for a much smaller bankroll. Your earlier post mentioned it could be played on EC bets, but I guess that didn't work out. But hopefully you could somehow have an alternate method where a player would only be risking hundreds instead of thousands of units.

Anyway, keep up the great work, Reyth!
« Last Edit: August 31, 2017, 07:42:57 AM by TERMINATOR »

The following users thanked this post: Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #12 on: August 31, 2017, 04:58:39 PM »
Ya, great points Term.  I am a progression fanatic which is also frowned on by many folks.  The reason is because I don't feel comfortable betting unless I have at least a 99% win rate.  Because of this, the cost for 5 full levels of recovery will always be thousands of units no matter what you bet.

You mention the IDG guy, well his system required 7500 units for the full 5 levels.  He didn't discuss the concrete details of his recovery method but I know that sometimes he would stay at a lower level until the debt was collected.

But I have been all over his systems with a fine tooth comb and they don't work reliably in and of themselves.  This means that his recovery methods are the key to his success but he doesn't discuss them!  He just says that he "scraps" the debt back over and over.

So I came up with the 75% figure based on the expected average coup amount versus the debt.  We can actually customize our recovery based on whatever criteria we want, setting target goals based on number of coups or some percentage of the debt.

I think locking in the debt is another key to recovery -- we have to always keep in the back of our mind the potential loss during the recovery process -- this is the sign of a successful recovery system.

The goal is to create an ebb and flow during the most chaotic times where we go up to Level 2 but back down to Level 3 and then back up to Level 1 and back down to Level 2 etc.  We DON'T want to go down to Level 2 and then down to Level 3 and then down to Level 4 and then down to Level 5 -- we are trying to protect our bankroll.

A hidden secret about the recovery levels is that the ratio of coups:recovery grows smaller as you go down the tier.  So that's why I focus on the LEVEL DEBT of the level above the one we are on, and work with a 75% figure of THAT debt and not the overall debt.  We have to minimize our risk at every step.

What we have in our favor is the tendency for substantial gaps to be followed by a higher hit ratio and for substantial gaps to be relatively rare and therefore not likely to repeat -- again this is why I need a 99%+ win rate.
« Last Edit: August 31, 2017, 09:51:54 PM by Reyth »

The following users thanked this post: TERMINATOR

#### dobbelsteen

• Hero Member
• Posts: 1560
• Thanked: 544 times
##### Re: The Bomb Shelter
« Reply #13 on: September 02, 2017, 09:09:03 AM »
I will try the bomb shelter on a RNG online roulette in the fun mode. Can you advice me. In the fun mode my balance start with 5000 units.

The following users thanked this post: Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4303
• Thanked: 1542 times
##### Re: The Bomb Shelter
« Reply #14 on: September 02, 2017, 03:06:14 PM »
1) I would recommend 13300 BU to cover 5 full levels of recovery

2) You need to track the performance of the DS' and identify the strongest ones to bet.  I would recommend using my software to track it.

3) I think you are asking to illustrate how difficult it would be for anyone to follow me -- maybe true but Bio Brick had no problem in the Bullseye thread:

http://forum.roulette30.com/index.php?topic=648.msg23023#msg23023

However, I wish to admit that this thread is probably too much raw information and not enough information that is simply presented for my exact method.  The reason is because I have been struggling to find the best way to maximize my chances of success and as my understanding has become more complete & accurate, my method has shifted.

So thank you for pointing this out and here is exactly what I do and think when playing:

Statistical Overview

I have found that DS's have a certain stability when it comes to statistical performance, specifically the speed of each selection (how frequently they hit) & the frequency of large gaps (over 27 spins without a hit).  I have called this phenomenon RNG BiasTM and it is an amazing statistical fact.

I have found that every session will have certain DS's that will perform over a long term (thousands of spins) far greater than expectation for the frequency of hits (HS ratio; Hits to Spins ratio; how many hits versus expectation for the number of spins) and that these same DS's will have far less large gaps (27 spins+ without a hit) over the same spin cycle.

I have noticed a phenomenon where a single DS will go for thousands of spins without a single PLE (Progression Loss Event; gap of 27 spins without a hit) and multiple hundreds of coups without a failure to hit within 27 spins (PLE; Progression Loss Event).

I have discovered that most of the time there will be 2 dominant DS's during any spin sequence, where one of them will be superior to the other; sometimes DS "A" will be superior to "B" but then they will switch and "B" will be superior to "A" in their speed (HS Ratio).

I think its important to be constantly aware of the interrelation of these two selections and strive to always be betting the dominant one because eventually either "A" or "B" will fail (PLE) and the other one will not and will continue to perform well.

I believe that the dominant selection actually forces the other selection to fail because it repeatedly grabs crucial hits that the other selection needs in order to survive; this combines with a statistical flow that focuses on the non-dominant DS's to bring about the failure (rare but eventually it happens).

My goal is to identify which DS's are likely to achieve this excellent performance & bet them exclusively.  I have created a method to do this that has several components that work synergistically in our favor.

Progression

This is the statistical foundation of the system and the basis for analyzing the performance of the DS's.

The most important feature of the progression is the win rate.  The win rate determines if we successfully recover when recovering debt.  It also determines whether a DS will perform well, long-term.  Without a proper progression, everything fails when it counts the most.

Here is the progression I use, its a simple Loothog but the length is specifcally designed to produce the best results for the least amount of cost:

1 1 1 1 1
2 2 2
3 3
4 5 6 7 8
10 12 14
17 21 25
30 36 43
52 62 74 (99.16%)

These 27 bets will hit 99.16% of the time.  The cost is based on the unit size which I discuss below.

All DS's are tracked using this progression & the software reports every time a failure occurs (PLE) and every time when 2 failures occur within 56 spins (BSE; Black Swan Event).

Unit Size

I have discovered a method of using the unit size to provide an additional statistical advantage in our play.

A unit is 2 chips.  This allows us at certain times to bet a half-unit of 1 chip.  The specific advantages are discussed in relation to recovery, below, in its section.

Therefore the cost of the progression is 886 base monetary units (BU; cents/dollars/bits/pence etc).

I have called betting a half-unit, the "sun deck" of our bomb shelter and its a very powerful compromise, as opposed to spinning virtually (entering the shelter).

Debt Recovery -- Overview

Eventually we will have a progression failure (PLE) and have a debt of -886 BU.  If we were on the sun deck our debt will be -443 BU.

In both cases we need to recover our debt more quickly than we normally accrue profit, achieving the lowest possible statistical profile that we can, which is why we use the following recovery schedule:

Units...Level Cost...Total Cost
....1................886.......................886
....2................1772.....................2658
....3................2658.....................5316
....4................3544.....................8860
....5................4430.....................13290

The basic way this chart operates, is to drop down a level (unit size) after every progression failure.  So that after one failure with a debt of -886 BU (total cost) we will be on Level 2, betting 4 chips (2 units) and recovering debt twice as quickly as we normally do.

If we were on the sun deck, our debt would be -443 BU because we were only betting half-units BUT we will still attack the debt with 4 chips!  This is the power of the Sun DeckTM and is a very helpful tool in our toolbox.

We are prepared for a full loss at 1 unit but we strive to achieve our losses from the sun deck (half-unit).

If we also have a progression loss (PLE) while we are still recovering on Level 2, we will then drop to Level 3 and start recovering the debt using 3 units (6 chips).

What we want to avoid is to simply use this chart mechanically, without thought.  Debt recovery is an art just as much as it is a science and so we must engage our entire mind to the task as much as we can.

Debt Recovery -- Method (Goal Setting) [cont. below]
« Last Edit: September 08, 2017, 05:13:38 AM by Reyth »

The following users thanked this post: kav, TERMINATOR