Author Topic: The Bomb Shelter  (Read 3348 times)

0 Members and 2 Guests are viewing this topic.

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #15 on: September 02, 2017, 05:40:12 PM »

Code: [Select]
1  157  7.182725E-03
 2  175  1.518895E-02
 3  180  2.342392E-02
 4  200  3.257389E-02
 5  157  3.975661E-02
 6  177  4.785433E-02
 7  186  .0563638
 8  168  6.404977E-02
 9  202  7.329124E-02
 10  168  8.097722E-02
 11  167  8.861744E-02
 12  159  9.589166E-02
 13  161  .1032574
 14  175  .1112636
 15  164  .1187666
 16  169  .1264983
 17  160  .1338183
 18  137  .140086
 19  161  .1474517
 20  158  .1546802
 21  164  .1621832
 22  160  .1695032
 23  154  .1765486
 24  144  .1831366
 25  141  .1895873
 26  162  .1969988
 27  143  .203541
 28  145  .2101748
 29  165  .2177235
 30  169  .2254552
 31  142  .2319517
 32  131  .2379449
 33  145  .2445786
 34  130  .2505261
 35  143  .2570683
 36  136  .2632903
 37  150  .2701528
 38  133  .2762375
 39  143  .2827798
 40  127  .28859
 41  127  .2944002
 42  121  .2999359
 43  133  .3060207
 44  115  .3112819
 45  122  .3168634
 46  130  .3228109
 47  135  .3289871
 48  135  .3351633
 49  114  .3403788
 50  135  .346555
 51  118  .3519535
 52  113  .3571233
 53  119  .3625675
 54  102  .367234
 55  131  .3732272
 56  104  .3779852
 57  113  .3831549
 58  119  .3885991
 59  137  .3948669
 60  112  .3999909
 61  121  .4055266
 62  116  .4108336
 63  101  .4154543
 64  102  .4201208
 65  116  .4254278
 66  107  .430323
 67  103  .4350352
 68  105  .4398389
 69  118  .4452374
 70  86  .4491719
 71  86  .4531064
 72  107  .4580016
 73  104  .4627596
 74  100  .4673346
 75  100  .4719096
 76  95  .4762558
 77  98  .4807393
 78  112  .4858633
 79  80  .4895233
 80  103  .4942355
 81  100  .4988105
 82  92  .5030195
 83  96  .5074115
 84  85  .5113002
 85  102  .5159667
 86  70  .5191692
 87  86  .5231037
 88  86  .5270382
 89  73  .5303779
 90  86  .5343124
 91  83  .5381096
 92  102  .5427761
 93  91  .5469393
 94  102  .5516058
 95  103  .556318
 96  77  .5598408
 97  103  .564553
 98  77  .5680758
 99  78  .5716442
 100  83  .5754415
 101  71  .5786898
 102  78  .5822582
 103  78  .5858267
 104  72  .5891207
 105  71  .5923689
 106  72  .5956629
 107  82  .5994144
 108  92  .6036234
 109  74  .6070089
 110  58  .6096624
 111  75  .6130936
 112  61  .6158844
 113  90  .6200019
 114  74  .6233873
 115  72  .6266813
 116  74  .6300668
 117  78  .6336353
 118  78  .6372038
 119  74  .6405892
 120  58  .6432428
 121  56  .6458048
 122  75  .649236
 123  65  .6522097
 124  83  .6560069
 125  63  .6588892
 126  58  .6615427
 127  77  .6650654
 128  64  .6679934
 129  54  .6704639
 130  59  .6731631
 131  50  .6754506
 132  66  .6784701
 133  54  .6809406
 134  53  .6833653
 135  39  .6851496
 136  62  .6879861
 137  59  .6906853
 138  61  .6934761
 139  52  .6958551
 140  56  .6984171
 141  55  .7009333
 142  50  .7032208
 143  55  .7057371
 144  64  .708665
 145  44  .710678
 146  50  .7129655
 147  49  .7152072
 148  61  .717998
 149  46  .7201025
 150  47  .7222527
 151  56  .7248147
 152  58  .7274682
 153  42  .7293897
 154  41  .7312654
 155  41  .7331412
 156  49  .7353829
 157  40  .7372129
 158  63  .7400951
 159  39  .7418794
 160  52  .7442584
 161  45  .7463171
 162  43  .7482844
 163  53  .7507091
 164  55  .7532254
 165  55  .7557416
 166  53  .7581664
 167  49  .7604081
 168  46  .7625126
 169  43  .7644798
 170  31  .765898
 171  42  .7678196
 172  32  .7692835
 173  29  .7706103
 174  44  .7726233
 175  42  .7745448
 176  56  .7771068
 177  46  .7792113
 178  39  .7809955
 179  40  .7828255
 180  38  .784564
 181  37  .7862567
 182  43  .788224
 183  41  .7900997
 184  35  .791701
 185  37  .7933937
 186  44  .7954067
 187  38  .7971452
 188  40  .7989752
 189  31  .8003935
 190  41  .8022692
 191  34  .8038247
 192  38  .8055632
 193  29  .80689
 194  24  .8079879
 195  32  .8094519
 196  29  .8107787
 197  37  .8124714
 198  29  .8137981
 199  38  .8155366
 200  29  .8168634
 201  36  .8185104
 202  42  .8204319
 203  44  .8224449
 204  26  .8236344
 205  39  .8254186
 206  33  .8269284
 207  27  .8281636
 208  36  .8298106
 209  24  .8309086
 210  24  .8320066
 211  35  .8336079
 212  29  .8349346
 213  32  .8363986
 214  25  .8375423
 215  29  .838869
 216  32  .840333
 217  29  .8416598
 218  21  .8426206
 219  26  .84381
 220  36  .845457
 221  30  .8468295
 222  31  .8482478
 223  21  .8492085
 224  36  .8508555
 225  17  .8516333
 226  19  .8525025
 227  22  .853509
 228  22  .8545155
 229  22  .855522
 230  16  .856254
 231  23  .8573062
 232  21  .858267
 233  19  .8591362
 234  14  .8597767
 235  20  .8606917
 236  32  .8621557
 237  18  .8629792
 238  25  .864123
 239  18  .8649465
 240  22  .865953
 241  15  .8666392
 242  14  .8672797
 243  17  .8680575
 244  27  .8692927
 245  24  .8703907
 246  26  .8715802
 247  22  .8725867
 248  22  .8735932
 249  18  .8744167
 250  30  .8757892
 251  15  .8764755
 252  23  .8775277
 253  25  .8786714
 254  19  .8795407
 255  14  .8801812
 256  26  .8813707
 257  18  .8821942
 258  10  .8826517
 259  26  .8838412
 260  23  .8848934
 261  20  .8858084
 262  25  .8869522
 263  19  .8878214
 264  27  .8890566
 265  20  .8899716
 266  13  .8905664
 267  21  .8915271
 268  22  .8925336
 269  19  .8934029
 270  21  .8943636
 271  27  .8955989
 272  20  .8965139
 273  19  .8973831
 274  18  .8982066
 275  25  .8993503
 276  16  .9000823
 277  9  .9004941
 278  20  .9014091
 279  19  .9022784
 280  24  .9033763
 281  15  .9040626
 282  19  .9049318
 283  20  .9058468
 284  13  .9064416
 285  18  .9072651
 286  14  .9079056
 287  21  .9088663
 288  28  .9101473
 289  22  .9111538
 290  15  .9118401
 291  19  .9127093
 292  16  .9134413
 293  20  .9143563
 294  17  .9151341
 295  16  .915866
 296  14  .9165065
 297  15  .9171928
 298  15  .917879
 299  20  .918794
 300  10  .9192516
 301  11  .9197548
 302  20  .9206698
 303  10  .9211273
 304  16  .9218593
 305  12  .9224083
 306  16  .9231403
 307  13  .923735
 308  18  .9245585
 309  10  .925016
 310  14  .9256565
 311  12  .9262055
 312  15  .9268917
 313  13  .9274865
 314  13  .9280813
 315  17  .928859
 316  18  .9296825
 317  15  .9303687
 318  11  .930872
 319  17  .9316497
 320  13  .9322445
 321  9  .9326562
 322  12  .9332052
 323  11  .9337085
 324  10  .934166
 325  19  .9350352
 326  11  .9355385
 327  6  .935813
 328  12  .936362
 329  12  .936911
 330  14  .9375514
 331  9  .9379632
 332  12  .9385122
 333  12  .9390612
 334  17  .9398389
 335  10  .9402965
 336  17  .9410742
 337  12  .9416232
 338  9  .942035
 339  17  .9428127
 340  10  .9432702
 341  17  .9440479
 342  8  .944414
 343  10  .9448714
 344  9  .9452832
 345  11  .9457864
 346  9  .9461982
 347  7  .9465184
 348  8  .9468845
 349  10  .9473419
 350  6  .9476165
 351  7  .9479367
 352  5  .9481654
 353  11  .9486687
 354  10  .9491262
 355  6  .9494007
 356  8  .9497667
 357  11  .9502699
 358  6  .9505444
 359  8  .9509104
 360  10  .9513679
 361  7  .9516882
 362  5  .9519169
 363  8  .9522829
 364  11  .9527861
 365  11  .9532894
 366  7  .9536096
 367  4  .9537927
 368  8  .9541587
 369  5  .9543874
 370  9  .9547992
 371  7  .9551194
 372  7  .9554396
 373  9  .9558514
 374  7  .9561716
 375  13  .9567664
 376  9  .9571782
 377  2  .9572697
 378  5  .9574984
 379  8  .9578644
 380  10  .9583219
 381  5  .9585506
 382  5  .9587794
 383  8  .9591454
 384  8  .9595114
 385  6  .9597859
 386  13  .9603806
 387  8  .9607466
 388  8  .9611126
 389  2  .9612041
 390  12  .9617531
 391  5  .9619819
 392  9  .9623936
 393  5  .9626224
 394  9  .9630342
 395  5  .9632629
 396  4  .9634459
 397  3  .9635831
 398  6  .9638577
 399  6  .9641321
 400  13  .9647269
 401  3  .9648641
 402  6  .9651386
 403  7  .9654589
 404  10  .9659164
 405  7  .9662366
 406  8  .9666026
 407  6  .9668771
 408  6  .9671516
 409  6  .9674261
 410  5  .9676549
 411  8  .9680209
 412  7  .9683411
 413  6  .9686156
 414  4  .9687986
 415  3  .9689358
 416  6  .9692104
 417  12  .9697593
 418  4  .9699423
 419  3  .9700796
 420  9  .9704913
 421  6  .9707658
 422  5  .9709946
 423  3  .9711319
 424  10  .9715893
 425  3  .9717266
 426  2  .9718181
 427  4  .9720011
 428  9  .9724128
 429  7  .9727331
 430  8  .9730991
 431  5  .9733278
 432  2  .9734194
 433  3  .9735566
 434  1  .9736024
 435  1  .9736481
 436  7  .9739683
 437  4  .9741513
 438  4  .9743344
 439  6  .9746088
 440  4  .9747918
 441  4  .9749748
 442  3  .9751121
 443  7  .9754323
 444  10  .9758899
 445  2  .9759814
 446  6  .9762558
 447  5  .9764846
 448  2  .9765761
 449  3  .9767134
 450  9  .9771251
 451  0  .9771251
 452  4  .9773081
 453  4  .9774911
 454  5  .9777198
 455  3  .9778571
 456  4  .9780401
 457  5  .9782688
 458  5  .9784976
 459  7  .9788178
 460  3  .9789551
 461  4  .9791381
 462  1  .9791838
 463  4  .9793668
 464  1  .9794126
 465  3  .9795498
 466  2  .9796413
 467  2  .9797328
 468  3  .9798701
 469  1  .9799158
 470  4  .9800988
 471  2  .9801903
 472  3  .9803275
 473  4  .9805106
 474  4  .9806936
 475  4  .9808766
 476  2  .9809681
 477  2  .9810596
 478  1  .9811053
 479  4  .9812883
 480  5  .9815171
 481  2  .9816086
 482  7  .9819288
 483  3  .982066
 484  3  .9822033
 485  8  .9825693
 486  0  .9825693
 487  3  .9827065
 488  3  .9828438
 489  2  .9829353
 490  4  .9831183
 491  1  .9831641
 492  2  .9832556
 493  3  .9833928
 494  3  .98353
 495  6  .9838046
 496  3  .9839418
 497  4  .9841248
 498  6  .9843993
 499  6  .9846738
 500  4  .9848568
 501  1  .9849026
 502  3  .9850398
 503  2  .9851313
 504  7  .9854516
 505  3  .9855888
 506  3  .9857261
 507  1  .9857718
 508  4  .9859548
 509  3  .986092
 510  5  .9863208
 511  5  .9865496
 512  4  .9867325
 513  2  .986824
 514  2  .9869155
 515  3  .9870528
 516  6  .9873273
 517  3  .9874645
 518  4  .9876475
 519  1  .9876933
 520  2  .9877848
 521  3  .987922
 522  0  .987922
 523  2  .9880136
 524  4  .9881966
 525  1  .9882423
 526  2  .9883338
 527  0  .9883338
 528  2  .9884253
 529  3  .9885625
 530  2  .988654
 531  4  .9888371
 532  0  .9888371
 533  2  .9889286
 534  1  .9889743
 535  2  .9890658
 536  2  .9891573
 537  1  .989203
 538  5  .9894318
 539  0  .9894318
 540  3  .989569
 541  3  .9897063
 542  2  .9897978
 543  3  .989935
 544  2  .9900265
 545  0  .9900265
 546  4  .9902095
 547  3  .9903468
 548  2  .9904383
 549  1  .9904841
 550  2  .9905756
 551  2  .990667
 552  6  .9909415
 553  4  .9911245
 554  2  .991216
 ...
 1401  1  1
 
The following users thanked this post: TERMINATOR

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #16 on: September 02, 2017, 05:41:34 PM »
Debt Recovery -- Method (Goal Setting) [cont.]

The above chart shows the number of coup attempts, followed by the total number of progression failures (PLE's) achieved at that coup attempt (generated over 16M spins), followed by the percentage chance of a PLE for that attempt.

For example, let's say we have 9 coups taken so far on a particular number (PMI; Progression Movement Indicator; the number of coups since the last PLE), we want to know the percentage chance of a progression failure (PLE) for our 10th coup attempt.

We see that the 9th coup had a percentage chance of failure (PLE) of .073 (7.3%) and our 10th coup attempt will thus fail .081 (8.1%) of the time.

This chart IS designed to scare you but its also designed to show the awesome power of RNG BiasTM and how we are able to achieve amazing improvements over probability when we are able to accurately identify it.

The point is that with probability working against us in proportion to the length of our recovery, we need to minimize our statistical exposure as much as possible.  As you can see from the chart, this means cutting down the number of coups as much as possible.

One of the strongest ways to minimize our exposure is goal setting.  Here is the advantage of setting goals:

Average Coup: 3 units

SCENARIO A -- Straight debt collection

-886 BU is the debt, recovering at Level 2 which is 4 chips and 12 BU per coup.
That is 74 coups which has a 46.7% chance of failure.  So 4.7 times out of 10 we will fail and be forced to recover at Level 3 which looks like this:

-2658 BU is the debt, recovering at Level 3 which is 6 chips and 18 BU per coup.
That is 148 coups which has a 71.8% chance of failure.  So over 7 out of 10 times we will fail and be forced to recover at Level 4 which looks like this:

-5316 BU is the debt, recovering at Level 4 which is 8 chips and 24 BU per coup.
That is 222 coups which has a 84.82% chance of failure.  So over 8 times out of 10 we will fail and be forced to recover at Level 5 which looks like this:

-8860 BU is the debt, recovering at Level 5 which is 10 chips and 30 BU per coup.
That is 296 coups which has a 91.6% chance of failure.  So over 9 times out of 10 we will fail and BUST OUT!

SCENARIO B -- Goal setting

-886 BU is the debt but we set a goal to recover 75% of it at Level 2 which is 4 chips and 12 BU per coup.
That is 56 coups which has a 37.8% chance of failure.

From there we would recover the rest of the debt, which is 222 BU from Level 1 which is what I call "locking in the profits" because NOW if we take a loss, it will only be for -886 BU instead of -2658 BU, saving 1772 BU.

We can set goals based on different things like the total number of coups (like 27 which is a 20% chance of failure), or different target amounts based on previous debt levels achieved, etc.

Goal setting is where the art comes into the science of debt recovery but we should always look to minimize our risk at the lower levels and experience higher risk at the higher levels.

Its not as bad as it first seems because we have a couple of things working in our favor:

1) After a substantial gap, numbers will tend to perform better (faster speed, less gapping).
2) We have a bet selection that identifies groups of numbers that perform far better than expectation and probability (RNG BiasTM)

Bet Selection -- Terms



HS Ratio

The HS Ratio is the speed at which a number achieves hits.  It is the number of spins that a number is above or below expectation; a negative number means it is below expectation (less hits than expectation) and a positive number means above expectation  (more hits than expectation).  The number represents the actual number of spins that it is above or below, 0 being in perfect accord with expectation.

So looking at the image above, the HS ratio is the first number displayed underneath the cyan DS title.  Right now, we have selected DS 25-30 which is why its highlighted in purple; it has a HS Ratio of +493.  This means that it has experienced 493 LESS spins than expected for the number of coups generated.

PMI

This is the Progression Movement Indicator which shows the number of coups since the last progression failure (PLE).  It is the second number in the list below the DS title.  Again, looking at DS 25-30, we can see there have been 17 coups since the last PLE.  The PMI figure can be used with the chart from the Debt Recovery section to determine our percentage chance of encountering another PLE; in this case it is 13.38%.

PLE Ratio

This is the Progression Loss Event Ratio which shows how many PLE's we have had vs. expectation; a negative number is MORE PLE's than expected and a positive number is LESS PLE's than expected.  The number respresents the number of coups that the selection is over or under expectation, 0 being in perfect accord with expectation.

It is the third number in the list below the DS title and so we can see that DS 25-30 has achieved a +209 PLE Ratio.

Bet Selection -- Criteria

The primary criteria for bet selection is the HS Ratio.  We are looking for numbers that consistently hit within expectation and almost never go far outside of it; think about it, how else are they able to gain such favorable statistics?

Like Forest Gump would say, "A HS Ratio is, as a HS ratio does!".  We can very quickly tell if a selection is not performing to its standard; i.e. slowing down in speed.

Generally speaking, we are looking for the selection that is performing the fastest; the highest HS Ratio.

What we are also looking for is its neighbors to also share a fast speed.  Looking again at DS 25-30, we see that the neighbor 22-27 has a +184 HS Ratio and the neighbor 28-33 has a +116 HS Ratio.  Note that DS 1-6 & DS 31-36 only have one neighbor, which is actually better!

The reason neighbors come into play is because each immediate neighbor shares 3 numbers with our chosen selection.  In a sense, the neighbors are a rating of our overall selection; if both neighbors have a high HS Ratio, that means that our entire selection of 6 numbers is performing very well.

A classic sign that a number is soon to slow down is when one of the neighbors starts losing its HS Ratio.  Also, PLE's tend to spread their negative effects to their neighbors; it is very common to see PLE's & BSE's (Black Swan Events; PLE's occurring less than 56 coups of each other) cause their neighbors to do the same.

Let's look at DS 4-9 for a moment.  It is the second fastest number on the board with a HS Ratio of +462.  It has 810 more coups than expectation (PLE Ratio)!  It has also been 243 coups since the last PLE (PMI)! 

This DS is truly peforming fantastically but since we already know that, we are only concerned about whether it is likely to have a PLE or not.  When we take a deeper look at its neighbors, we start seeing cracks in the armor, specifically DS 1-6.

Two things stand out immediately about this neighbor.  First, it has a NEGATIVE PLE Ratio, which means that it has 121 LESS coups per PLE than expected -- it is plagued by  PLE's by 121%!  Secondly, the red "B" next to the PLE Ratio indicates that it has had (at least) two PLE's in the last 56 coups.

These ugly things about the immediate neighbor of DS 4-9 are very likely to cause it to have a PLE.  I would rather enter the shelter (bet virually, just spin through) than actually bet this DS.

Sure, DS 1-6 has an incredible HS ratio of +326 but the HS Ratio is not the only determiner of the likelihood of a PLE.  Even the best numbers on the board will crack eventually and even though they will likely recover very quickly, that doesn't change the fact that we lost money during that short period of time.

Let's briefly examine the other neighbor of DS 4-9, DS 7-12.  Everything looks fine here but did you notice a subtle warning sign nearby? 

Yes, another BSE with DS 10-15, the neighbor of DS 7-12.  Normally, this will not stop me from betting the main DS (4-9 in this example) but as far as the future is concerned it is a strong sign that DS 4-9 is likely to encounter more problems shortly.

Betting Method -- Technicals

1) Select the DS you wish by moving the purple cursor with the "Q" and "W" keys; "Q" to move the cursor to the left and "W" to move the cursor to the right.
2) Spin the wheel in your casino.
3) Select the spun number with your mouse from the selection of grey "buttons" at the bottom of the output screen
4) Review the output results:
a) If you received a hit, you will hear a cash register sound and the progression will reset
b) if any number had a PLE, you will hear a "bubbling" series of tones and you can see which selection experienced this by searching for the one with a PMI of 0.
c) if any number had a BSE, you will hear a "steady & repeated" series of tones and you can find that selection by looking for the red "B" next to the PLE Ratio & the PMI of 0.
d) If the progression requires you to change your bet, you will hear a mono-tone beep and the "Bet:" number in the bottom right will indicate the number of units required.
e) If you wish to review the chart for the HS ratio of any number, simply select the proper number with the purple cursor and press "C".  The chart screen will show up, showing the entire HS Ratio history for that number.  You can also change the selection from the chart screen, using the same keys

Betting Method -- Method

Terms
KoH -- King of the Hill.  Generally speaking, this is the fastest DS on the board; the HS Ratio is the highest.  Sometimes a number will be "heir apparent", where the HS ratio hasn't reached the fastest but due to recent performance and the overall statistical picure, it is most likely the actual KoH to be; sometimes we should bet the "heir apparent" due to the risk of a PLE on the current KoH.

QoH -- Queen of the Hill.  This is a very fast DS that is contesting for KoH but is not performing as well.  We have to be very cognizant of this number and its neighbors because it is the number one threat to our selection (KoH).  This is due to the fact that if the QoH heats up, it can overtake the KoH and cause a PLE by "hogging up all the hits"; this is called "the changing of the guard".

Rival Princess -- This is a 3rd DS that is contesting for the QoH spot.  This is not always present, but when it is, it can be quite dangerous.  Just like with musical chairs, we don't want to be without a seat when the music stops, we want the other guy to be!

The March -- This is the section of the progression that runs: 4 5 6 7 8.  If our betting goes beyond this point (10 and beyond), our selection status as KoH is potentially in jeopardy.

Betting Method -- Opening

To start we should choose a DS at random and just simply bet it normally.  We should make our first few/several/x amount of coups at Level 1 and the rest from the sun deck (half unit).  I personally am using 3 bets at Level 1 which has a 2.34% chance of failure.

We need to use the sun deck at some point, so the chances of our recovery getting out of control is minimized due to the 4:1 ratio on the losing bet amount.

We should monitor the hits we receive, if they are within or close to expectation, we should stay on the DS.  If our HS Ratio goes strongly negative in relation to the rest of the board, we should switch as soon as we take a coup.

We should keep switching to the fastest DS that has the best neighbors.  We are hunting for the KoH that will outperform probability for many hundreds and even thousands of spins.

Eventually we will identify a clear KoH, the HS Ratio is obviously higher than every other DS and its neighbors are at least close to satisfactory. 

We should also be able to identify our main threat(s) in the QoH & "rival princess" (if there is one).

Betting Method -- Standard play

Betting guidelines

The KoH should always hit within or close to expectation; as long as we are hitting within or close to expectation, nothing else concerns us.  Expectation is 7 spins. 

To make sure we are cognizant of any threats, we need to monitor every spin and mentally sort them into the following categories:

Irrelevant -- DS's that pose no threat.  They will be slow and not connected to any threat.  Sometimes an irrelevant DS will become relevant due to a very large move.  We should check after every coup to see if this is happening.

QoH Leeching --  This is when we get 1, and more importantly, 2 or more hits on the QoH before our selection hits.  It indicates that the QoH is outperforming our selection and could indicate an impending "changing of the guard" (COG).

QoH Neighbor -- The QoH has allies and they are planning a rebellion!  Strong neighbor performance is not in our favor.

Rival Princess Leeching -- The same as QoH leeching.

Rival Princess Neighbor -- The same as QoH Neighbor.

Our Neighbor -- We have allies too!

By mentally sorting these categories we can tell, we can feel the pull, if our number is being heavily opposed to the point of potentially causing a PLE.

Normal behavior is getting a hit in The March that is immediately follwed by at least 1 (more than 1 is preferable) hit within expectation.

Anytime we get two hits outside of expectation in a row, it is cause for concern and we should know the reason why, having mentally identified which threat(s) are leeching our hits.

Anytime our HS Ratio is surpassed by another selection it is grave cause for concern.

Anytime there is a BSE with one of our immediate neighbors, it is grave cause for concern.

Anytime one of our neighbors has a PLE and its HS Ratio goes negative or we just went deep in the progression, it is cause for concern.

If we have reason for concern and feel it is no longer safe for our selection we need to take one of the following actions:

1) Switch to the QoH/Rival Princess (QoH Apparent)
2) Enter the Sun Deck (bet 1/2 units)
3) Enter the Shelter (bet virtually, spin through)

Sometimes we will switch between QoH/KoH, back and forth until one side finally wins -- one side WILL finally win and we must strive to always be on the dominant side or at least on the Sun Deck or in the Shelter when the battle is over.

General Betting Rules


GREEN

KoH status assured.  No worries if a YELLOW hit preceded this.  If a RED hit preceded this, proceed with caution as another GREEN hit is needed to maintain KoH status.

YELLOW

KoH status is strong but an immediate GREEN hit is required.  Something is most likely wrong if a second YELLOW hit is obtained or if RED-GREEN-YELLOW.

RED

KoH status is in serious question.  Without an immediate GREEN hit, this is most likely no longer the KoH.  Strongly consider switching numbers, hit the sun deck or take shelter if there is no GREEN hit, immediately following this.  Two of these in a row is no longer KoH.


Let's be clear, the goal is to dodge the bullet when it comes, or at least take a flesh wound [Sun DeckTM]. ;)



When it stops looking like a KoH, we need to minimize our risk.       
« Last Edit: September 05, 2017, 09:32:24 PM by Reyth »
 
The following users thanked this post: TERMINATOR

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #17 on: September 03, 2017, 02:21:45 AM »


-443/+2658 (+36)[0]G78

This was not a bad run with 136 coups.  I saw the QoH forming and knew that was why we started slowing down but by the time I realized it was so serious as to be dangerous to my selection, I was already in the final progression.



This is what it looked like just before the PLE.  Two things were obvious:

1) The number had slowed down
2) The reason for the slow down was the emergence of the new QoH from the rubble

Technically, we have a new KoH in DS 10-15 and here is what it looked like:



I believe this is a freak move that will not last but DS 13-18 cannot be denied as the "heir apparent" to the throne here.



I think I should switch to DS 13-18 because of the very strong move by itself and both its neighbors that caused the PLE in DS 31-36.  I expect this strong positive movement to continue.

Level 2 Recovery.  Target: +529.91



+444/+3102 (+79)[79]G134

Well, have I got a story to tell you!  LOL.  Anyway, this is not your typical session and it doesn't appear that it will be unless some things change. 

First DS 13-18 became immediately slow as DS 31-36 made an immediate comeback but was followed shortly by another PLE!  Thankfully I wasn't on it because I had switched out to DS 13-18 and came in with a "fresh progression" that I was doing manually while the PLE occured in the computer's automated progression.

It was like it had a nervous breakdown or something and well, I couldn't believe it but DS 28-33 stepped in and became the KoH!  The hit returns have been amazing, just like a KoH functions, these 9 numbers just dominating the board (DS 31-36 is almost fully caught up).

The only concern is DS 25-30 which is having difficulty, most likely due to its slow neighbor in DS 22-27.  I get this feeling like I will soon be back on DS 31-36 again!

I think its times like these where there is no competition in sight that we get these massive rushes on our KoH selection.

I'm on the sun deck so let's see how far we can go! :D



-443/+2720 (+3)[0]G102

Well once again, my selection slowed down and I saw the rogue DS 1-6 rising up to about 85 and our selection was about 105 and I could tell that this was one of those COG (Changing of the Guard) moments and so I took some spins in the shelter and DS 10-15 took a BSE and so I thought that might release some pressure and DS 1-6 was slightly down in the 70's.  So I began betting and DS 1-6 rushed up so fast while I was betting so that there was no way for me stop it, same as the last one.

Right now the entire board looks very uninspiring.  Obviously our front runners are DS 1-6 & 28-33 but they are both so compromised; both have a very low PLE Ratio and slow neighbors.

The two undefeated DS's which are 13-18 and 16-21 are moving so slow and are surrounded by negative influences.

I just don't see any hope in this session and so I am going to fully reset everything.  I guess things don't always work out perfectly; it was that Rogue Princess DS 1-6 that wrecked everything.

Level 2 Recovery.  Target: +3052

PLE_DEMO




+2143/+3167 (0)[184]G98

After 13-18 took ahold of the board it just ran and ran and ran, with no competition in sight.  I recovered from Level 3 & Level 2 before, at the tail end, DS 22-27 finally started approaching; it was a classical COG moment & I hit the shelter while these two fought it out.

Once I felt like DS 22-27 was the clear winner, I started betting and got about 5 wins for the all-time high.

This charting data is getting REALLY helpful; if it looks like a down-trend on the chart, why expose ourselves with raised chips?  Regardless of the reason, the selection will usually slow down before gapping (PLE).

Back to the sun deck with DS 22-27.



+99/+3266 (0)[226]G127

So DS 13-18 started to make a comeback and I went back into the shelter to figure out which DS I would need to bet.  Its dangerous when they start conflicting because when the wheel chooses the KoH, it can steal all the hits from the QoH, so its best to remain in the shelter during these times.

DS 13-18 became the clear winner and while I was betting it, DS 22-27 finally collapsed with its first PLE.  Right before the PLE it was making a comeback but DS 13-18 kept hitting within expectation and of course was still the front runner HS Ratio wise and so I continued to bet.

The only reason I missed the PLE is because DS 22-27 never actually caught up in its HS Ratio -- this is the reason we always want to know which DS is dominant (KoH) and be betting that DS.

I have uncovered the reigning DS' in 13-18 & 16-21, both neighbors & without a PLE in over 220 coups!

Below is the last 639 spins from both DS:


DS 13-18 (KoH)


DS 22-27 (QoH)


Do you see how when the QoH was having its strongest time, the KoH was in its down time?  Do you also notice that while the QoH was having its progression loss, the KoH was hitting all the while (every upward move is a hit)? 

This is why the interrelation between the KoH & QoH is so important and it elucidates the following important principle:

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.

As long as we stay on the dominant DS, we will eventually land on the last one left without any progression losses.  Before we get there, certain DS's will fight for dominance and it is up to us to do the following:

1) Move to the Sun Deck
2) Move to the Shelter
3) Move to the newly dominant DS (or dominant heir apparent)

If we do these things, we will avoid all PLE's or at least take our losses at a 1/2 unit.

Continuing on the sun deck.



-443/+2868 (+138)[0]G112

I can't believe this.  First DS 7-12 took a BSE and then DS 13-18 & DS 16-21 (neighbors) equalized in their HS Ratio.  The obvious choice is DS 16-21 for KoH because a neighbor of DS 19-24 is superior to DS 10-15 which has a long term negative HS ratio and is neighbor (now) to a BSE.

So DS 16-21 starts off hitting fine along with DS 13-18 which does only very slightly better (1 hit additional).  Obviously we expect for DS 13-18 to start failing or at least for DS 16-21 to start excelling.  Without warning DS 16-21 takes a PLE:



This is just a bizarre event that is similar to a market plunging slightly downward, knocking us out via our stop loss & then taking off in the direction we forecasted.

My options are fourfold:

1) Continue betting 16-21 (Sun Deck or full)
2) Switch to betting DS 13-18 (Sun Deck or full)
3) Enter the Shelter
4) Reset the entire session

I am tempted to do #1 because I think it is actually the stronger DS and I expect an immediate increase in hits.

I don't want to do #2 because of the terrible neighbor.

#3 seems like the best option but I still want to do #1.

#4 is out of the question because I need to study the behavior of RNG BiasTM once the KoH falls.

Trading discipline suggests that when our plan fails, we remove ourselves from the market and await the development of a new plan.  Its greed for short term gains (and the loss of them, i.e. missing getting them) that would have me do #1.

So I will enter the Shelter and observe.  I am in overall profit & I just need to have situational clarity before I start to collect the debt.

Level 2 Recovery.  Target: +3201



Its just insane now.  I watched as DS 7-12 took a SECOND BSE and DS 10-15 continued to grow weaker but DS 13-18 continued to grow stronger while DS 16-21 declined terribly!

Eventually DS 22-27 approached with its neighbors on the upswing but was slammed and I mean SLAMMED back down with machine gun hits from DS 13-18!  And then DS 16-21 has a second PLE!

Look, I am going to call it now, DS 13-18 is doomed and DS 22-27 is the clear Heir Apparent QoH but I simply cannot bet it yet.

I just wanted to document the events and my thoughts because I have to go to work.

Oh, one other thing, when things just don't make sense, we need to "look under the hood" and see what is going on.  This will be the next update on the softwares where I will create a HS Ratio viewer for the individual numbers in each DS; so we can see what the specific causes are for a DS' behavior because I will admit something, my assertion that...:

The reason neighbors come into play is because each immediate neighbor shares 3 numbers with our chosen selection.  In a sense, the neighbors are a rating of our overall selection; if both neighbors have a high HS Ratio, that means that our entire selection of 6 numbers is performing very well.

...is not ALWAYS correct because in isolated circumstances, a neighbor can be doing well because its FAR numbers (the outside street) are doing well, which will have no effect upon our selection; its just that usually there is a sufficient mix of the numbers that the neighbors are a reliable rating of our selection -- but not always!

This new mod will help us understand when neighbors are and when neighbors are not a reliable factor.



Ok, so DS  13-18 finally fell and I started having hopes for DS 19-24 because of the strong neighbor support.  Well, they were crushed by an immediate PLE. 

I can't solidly hold out hopes for anything at this point, and I just need to reset this session.

I think we can identify several stages that every session can go through:

1) Opening -- multiple DS' trade off as the front runner in a rapid fashion
2) Middle (SUCCESS) -- 2 (or 3) DS' have clearly established themselves as KoH & QoH (& Rival Princess)
2a) Middle (FAILURE) -- Too much rivalry has prevented a clear front runner to clearly maintain KoH dominance
3) ??? -- The KoH has finally fallen and there is a continuing future that can be a smooth transition (3a) or a convoluted one (3b)

Most of the time we will see 2 and not 2a.  I am not sure about 3 but I have shown sessions where KoH & QoH domination continues over multiple thousands of spins (as many as 7000+). 

I believe that all sessions will eventually have a successful 3a but the problem is I don't have time to spin through them just to see what happens and I also don't want to be betting in questionable conditions.

So let's get a new session.  After I scrap back the debt, I will update the softwares so maybe we can shed more light on the lack of clarity issues that sometimes present themselves.



+447/+3315 (-48)[34]G73

DS 10-15 quickly dominated the board, thoroughly crushing all opponents.  Don't get me wrong, there were a few Sun Deck Moments and I even went into the shelter for a few spins when DS 25-30 approached.  The overall performance is below.


DS 10-15

This was an interesting way to end a recovery where DS 7-12 rushed up to equalize and since DS 13-18 (KoH neighbor) just had a PLE, I took the opportunity to slide over so that I would have better neighbor support; the final hit came from this DS, which is below.


DS 7-12

Its interesting to note that DS 7-12 did have an early PLE but its subsequent performance could not be denied and presented itself at the perfect time.  I felt that the lower PMI was proper enough compensation for the fast PLE Ratio.

Ok, so off to mod teh softwares! :D

New_Mod_DEMO

Now we don't just have to blindy trust the neighbor ratings!  We can understand specifically the meaning of the behavior and status of any DS.



+185/+3496 (0)[68]G38

This is more than a good time to enter the Shelter.


DS 19-24

This selection, though hitting consistently has been a down trend for a significant time along with its neighbor DS 16-21.  Its only now however that a third enemy has approached in DS 1-6, below.


DS 1-6

This is now a serious threat that is likely to cause DS 16-21 or DS 19-24 to have a PLE.



This is the kind of activity that steals essential hits.  To explain this data for DS 1-6:

Using the number 6 as an example, the first number is a red 57; the red means that the number is negative, so its actually -57.  This value is the HS ratio for this number, the number 6, HOWEVER, it is a special ratio because it is based on the number of hits versus the last 27 spins (the length of our progression).  The value actually represents the number of spins (57) that the number 6 has underperformed the ratio of 1:27 (1 hit to 27 spins).

The second number, coincidentally happens to be 6 and is the streak value.  This is the number of times that this number has hit within 27 spins, in a row.  Anytime it drops to 0 it will be red.

The final & third number is 20, this is the number of spins since the number 6 has last hit.  Anytime it climbs to 27 and above, it will be red.

So looking at all 6 numbers in this selection, we can see that EVERY number has hit within the last 27 spins (the bottom numbers are all green) and ALL numbers, except the number 3, has hit within 27 spins more than once; especially notable is the number 6 that has hit 6 times in a row within 27 spins.

Notice how this DS is absolutely on fire in advance of its HS Ratio (the main screen) showing that it has caught up.  This is what I was saying before about how numbers approach our KoH and how there is a "changing of the guard" (COG) moment where our selection gets overrun and we can likely have a PLE.

This is why I am now in the Shelter.

BONUS QUESTION:  Is there another DS that we should be taking special note of here?

First one to answer this question correctly will receive 100 free hits (eyeballs will be present) to the URL of their choice (this must be a non-gambling site that is free from objectionable material or links to such).



The concept of concentration versus dispersion

What I have found is that usually the random flow will concentrate hits in a single area of the board not only geographically BUT ALSO chronologically, so that this concentration will occur consistently over short cycles of spins.

I have named this concentration RNG BiasTM and it can be consistently observed over hundreds and even thousands of spins.


However, at certain times, this focus will disperse and be separated out among several sections of the board; this reduces the concentration on our selection.

It is during these times of dispersion that our selection (the KoH) is being challenged and it will either emerge victorious (conquering King) or vanquished (long live the new King!).


It is identifying these times of challenge and dispersion and then avoiding exposing our bankoll in them, that will take our success to far higher heights than just 99.16%!

NEW THEORY: The strength of a DS selection can be best measured by the total amount of STR positive numbers.

The STR rating is the streak rating which is the second number in the list of 3 on the DS number analysis screen.

IDEA: Track the STR rating for each DS selection & display it on the HS chart so that upward/downward moves can be compared with this rating, to show its accuracy.





+37/+3533 (+16)[101]G82



Long live the KoH!
« Last Edit: September 05, 2017, 11:17:00 PM by Reyth »
 
The following users thanked this post: kav

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #18 on: September 06, 2017, 12:06:10 AM »


+102/+3635 (+55)[140]G125










+218/+3853 (+131)[217]G147

There were several Shelter moments that are much more easily forecasted using the DS Analysis screen.  Whenever the total number of STR positive numbers drops to 2 (or sometimes 3 depending on the competition), hit the shelter until it hits 4 (or at least 3 depending on the competition).

Its alot more difficult to be precisely accurate with just the HS Ratio chart.  I think this is because it is a generic overall rating for a DS as a whole, whereas the DS Analysis screen is the ACTUAL engine of that DS, broken down by individual numbers.

So, long before the HS Ratio properly displays a downtrend (or the uptrend of an opponent), the DS Analysis screen shows the problem as soon as it occurs.



+90/+3943 (+162)[247]G153

The following Shelter MomentTM is brought to you by our friends at South Park:






Look just to the left of our selection to our neighbor, DS 22-27 -- see how every number is red for that selection?  That is the effect of a nuclear blast known as a PLE.

Now check our selection itself (DS 25-30) -- notice how down the left hand side it is all red?  That is nuclear fallout and its the reason that PLE's & BSE's clump together.

Now I can prove exactly why this phenomenon occurs. :)

Toodle-loo from Shelter Land! :D

PLE_2_DEMO



-442/+3538 (+83)[0]G125

It was a very long & profitable run (+665, around 221 coups) with mutliple switches but eventually I got tagged.

Hitting the Shelter until the clear KoH has been decided. :)

Ok, one of the things I just noticed was a steadily decreasing HS ratio on my neighbor DS 22-27 prior to this fall; first 20, then -12 and then -48 (we can see this from the posted screens).  I think this might be a warning sign that we can make use of.

What is the rationale for declining neighbors making our selection more likely to lose?  Just like when the PLE hit with that selection (as I noted above with images), 1/2 of our selection (the entire left side) is affected. BY DEFINITION, if our neighbor is slow, it means THAT side of our selection will not be hitting very frequently.

I need to start monitoring this, I think it was the sign that my selection was in jeopardy.

Anyway, we can clearly see that both KoH & QoH have a slow neighbor and so I am going to reset the session. 

I will make some mods to softwares, fixing the DS Analysis HS ratio to the proper 37 spin expectation, making sure there is no mouse input before allowing return to the main screen from the DS Analysis screen and I am going to install a Shelter Status indicator that will show on the main screen if I am in the Shelter or not and chart it as well so we can see on the chart when I hit the Shelter! :D



Ok, like how stupid can I be (please don't answer that question :blush:)?  I mean isn't it clear we just had a PLE from our neighbor that was followed by a PLE on our selection? 

Shouldn't it clear that our neighbor has slowed down and this then caused our number to be more likely to gap?

So I am creating Rule #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.

As a side note regarding the fact that resetting a session doesn't affect the statistics, where personal permanence is the dominate force, I am sure that is true BUT when we reset the session, the wheel brings about a new KoH and so its not as if there is literally no effect for resetting a session.

Ok scrapping back right now and have gotten back to Level 1 and about 70 BU from the all-time high.  I just wanted to note my latest thoughts about the DS Analysis (DSA) screen:

1) The method may be as simple as just spinning freely and checking the DSA after every coup.

2) The true idea with neighbors really revolves around the structure of the hot numbers within our selection, where BALANCE is the key, as follows:

H...H..........X...X
X...H..........H...H
H...X..........H...H
Balanced

H...H..........H...X..........H...H
X...X..........H...X..........H...X
H...X..........H...X..........H...X
Unbalanced

Where an equal amount of HOT numbers appear on both sides is balanced but an unequal amount is unbalanced; 5 or 6 numbers is always balanced and 3 or less numbers is always unbalanced.

It could be as simple as:

If our selection is balanced, it doesn't matter what our neighbor is doing, our position is safe as long as our HS Ratio remains KoH and the QoH is not approaching.

The reason I am making this note is because I am feeling mentally tired right now and I am not sure if I am thinking correctly.

Just a reminder to myself: PLE's and BSE's group together for a reason.

I think the answer might be that we are trying to make sure we don't have a weak or sleeping street in our selection. ;)



+4100 took down the softwares for tweaking.

At +4231 right now and here is the method that I have settled into that is working very smoothly:

1) Working with an established KoH, spin freely until a coup is obtained.
2) Check the DSA:


6 STR numbers is King Tide, raise a half-unit
5 STR numbers is High Tide, bet regularly
4 STR numbers is Mid Tide, use the following:

a) If the QoH is far away (58+ points HS ratio) bet regularly
b) If the QoH is closer but has less STR numbers bet regularly
c) Otherwise hit the Sun Deck or enter the Shelter


3 STR numbers (or less) is Low Tide, enter the Shelter or at least hit the Sun Deck (I enter the Shelter).

3) Always switch to the highest HS Ratio selection after every coup.





+279/+4510 (+169)[153]G174



So, everything is running very smooth by focusing now on the DSA.  The thought that has occurred to me is that the HS Ratio is the long-term trend indicator for which DS to bet on whereas the DSA STR figures are the short-term indicators for when to bet and when to stop betting.

The HS Ratio can show a strong up-trend but the STR figures can show that this is only 1-3 numbers repeating, which means that at least half of the selection is gapping; i.e. more vulnerable to the LotT gapping the hot numbers in our selection.

I think that this is the way the data should be monitored & prioritized.  However, I get the feeling that I am spending too much time in the Shelter and it makes me want to seek out a way to narrow the rules down so I can take profit during more of the "repeating up-trend" times.

Maybe there can be a rule that if the QoH is 58 points or more away from the KoH, all 4 STR's can be bet along with all 3 STR's that include entries on both sides.

Anyway, since the STR figures are now understood to be the short-term indicators of the highest priority, I should be charting them so I can study them more closely.

I am going to take the softwares down to implement that.



Here are some assumptions that I am currently operating under:

1) A current KoH will not gap (PLE) unless there is a QoH approaching close enough to take the throne, when it gaps.

2) A STR of 5 or 6 will not allow the selection to fail during the next 27 spins.

3) Since the LotT groups these hot numbers on two adjacent streets to produce results that far surpass expectation & probability, it makes a difference if we over-rely on a single street for our results.





Ok, so now we can see the performance of the STR figures for each DS on the same chart as the HS Ratio performance.  So far this been an excellent way to view PLE's to find out what the STR was at the time the failed progression started.



+293/+4803 (0)[170]G122

A classic example of KoH vs. QoH.  The situation is a bit dangerous as they are close to each other; DS 1-6 can easily overrun DS 31-36.  This is a good time be inside the Shelter until things become more clear; one of these will fall first and the goal is to be on the other one (or in the Shelter) when it does! :D



+104/+4907 (0)[220]G173

Mmkay?  Total dominance.  DS 1-6 had its PLE and now stands 99 points below our selection.  This is a very strong time where it CANNOT overtake our selection within 27 spins.



Furthermore, the security provided by all 6 numbers being active in our selection, makes this a wonderful time to be betting.

This pattern appears again and again and I feel like the potential here is so great for a very strong betting scheme.
« Last Edit: September 12, 2017, 06:50:00 PM by Reyth »
 
The following users thanked this post: kav

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #19 on: September 13, 2017, 07:00:57 PM »




+414/+5321 (0)[445]G314

DS 31-36 is absolutely crushing it!  We can't make stuff like this up!  This is plain as day, in your face, thwarting of expectation & probability, focused exclusively on 2 adjacent streets, over a period of thousands of spins!

I just earned in profit, 3 full recovery levels!!!


I haven't had a PLE for 1341 BU!  That is roughly about 425+ coups!!!  It only takes 37 coups (at Level 2) to recover the debt made from the Sun Deck!  This is roughly 11 TIMES that amount!!!

When we combine the fastest long-term trend for 2 streets with fast short-term trends for the numbers in those streets, WE WIN!



Making a note here because I am tending to forget this important concept:

It would be good to monitor the percentage of how our 6 numbers are performing in the LotT versus the rest of the board, like a rating.





+196/+5517 (0)[585]G380





Some thoughts:

1) The STR figures are like "hot streak detectors" and "cold streak detectors" where they indicate by their presence or lack of it, the chance of a streak good or bad.
2) Having the long-term HS Ratio for the DS as a whole is additional protection towards our success because when "instant dispersion" occurs without warning, the wheel is still favoring our selection overall (assuming no QoH takeover).
3) Maybe we shouldn't stop betting the KoH at any time because when we stop betting we are bringing the eventual loss closer & therefore minimizing our profit BUT:

a) Chances are we will get plenty of profit anyway because our selection gaps less
b) We have a chance of missing a PLE which is invaluable
c) IF the STR ratings prove to be a better indicator of a PLE than random, we are operating with an advantage



It finally had its first PLE.  Here is the data dump on the last spin before it fell:





Here is its status immediately after the 27 spin progression failed:



Here's the great thing though, before it fell there were only 3 STR figures (the entire right side 34-36 had fallen asleep) and so I was in the Shelter! :D

It feels great to have ZERO debt! :D

627 straight coups (that's a 0.0050325 event)!  That is approximately +1881 BU (exclusively from the Sun Deck with no raising)!  My actual profit?  +1581 BU! :D

So, 1881-1581=300 BU and the cost of a PLE is -443 BU!  So we showed a profit of +143 BU AND we avoided any risk of multiple successive failures during recovery! :D

We can see from the DSA that DS' 1-6 & 4-9 were primarily responsible. 

What can we say, its impossible to complain with 6.27 TIMES EXPECTATION for the number of coups!  It ran 349 LESS SPINS than expected for the hits generated.  It took 526 MORE COUPS than expectation for its single PLE!  At over 3500 spins, it is a CLEAR example of RNG bias! :D

RNG Bias is a statistical fact.  More typically, we will see around 200 straight coups before a PLE.

I have a choice here to either reset this session or continue to play, where obviously DS 31-36 is the uncontested KoH.  I might as well learn now how it goes, and so I am going to test it by continuing on.  I won't start betting until the STR numbers come in sufficiently.

The second PLE hit just about at expection and here is the data dump for the bet right before it did:





Here is the data for right after the progression failed.



Once again, the STR numbers pointed out the bust by only showing 3 numbers at the start of the failed progression, so I was in the Shelter!




So, 105 coups from the Sun Deck would be +315 BU and the loss would be -443 BU and so we saved 128 BU because of the Shelter!  My actual profit?  +176 BU!  Much better to be +176 rather than -128, right? :D

The STR numbers indicate to us immediately when there is a problem and it is kind of counter-intuitive because normally after a single deep run in the progression I would be strongly inclined to take that second run because its likely we will get an updraft as compensation -- but the STR numbers counsel us to immediately STOP betting because we don't have any actual support! O_o

Once again, DS 31-36 is the clear KoH and so I must continue betting it to see what happens next! :D





+78/+5815 (+624)[94]G298

Well, have to reset because the casino ended the session.  This will cause the random flow to assume a new character.  I will attempt to fix some things in the software as well.

Look at the bias though!  Over 4750 spins with a clear and uncontested KoH!!  Once again, like always, the pattern emerges; 624 more coups for the 2 PLE's received & 298 LESS spins for the total coups received!

Look at the rest of the board, nothing comes even close! :D

Open_Demo

The pattern emerges once again.  After the conclusion of this video, the casino closed the session before any more spins could be recorded.

RNG_Bias_Demonstrated

Open_Demo_3





+423/+6238 (0)[152]G150

This is how it has turned out so far since the last video.  Of course, we can see the clear RNG Bias has persisted now for over 700 spins.  DS 13-18 is the QoH and has approached closely.  Because our selection only has 2 STR numbers, I will continue on, spinning from the Shelter.

I haven't had a PLE for 2258 BU!  This is roughly around 725 coups!  This is an 0.002379 (1 in 420) event! :D



Just wanted to make this note:

The STR figures perform an analysis of the random sequence.  If significant statistical dispersion exists, they will indicate it clearly.







+41/+6279 (0)[214]G158

Just documenting that the KoH has vanquished all enemies, being the last man standing without a PLE.  This phenomenon has persisted for over 1100 spins!

The use of the STR figures has greatly increased the accuracy of my play, where I am able to take coups despite enemies nearby, identifying the safest times to do so.





+57/+6336 (+178)[94]G148

Documenting the PLE of the former KoH seen above.  On approximately spin 1325, DS 13-18 overtook DS 10-15.  It tried to rise back up and came very close but on spin 1573 it fell; it went above 270 coups before falling, which is typical.

It is known that when the KoH & QoH are battling, one must fall and the other must continue and so some caution must be applied (the rules about 4 STR numbers).

Here is the data dump right before the KoH fell:





Here is its status right after the fall:



Once AGAIN, it was only 3 STR numbers and I was in the Shelter!  As you can see it went to slightly over expectation in the number of coups (110).

A thought has occurred to me that after the original KoH falls, maybe the next KoH's don't go for as long statistically?  I mean I have no idea but I am wondering if once the KoH falls, its better to start a new session for a new statistical focus?

One thing is clear, DS 13-18 has not performed overall as well DS 10-15 has; the PLE ratio difference of close to 100 points.  However, I think that it is always best to stay with the current KoH.

I am going to continue with DS 13-18 as it is still the KoH.  I have to learn how things go after the first round of PLE's somehow.

Documenting for the first time ever, a BSE for the KoH!





Here it is right after it fell.



If you guessed that I was in the Shelter, you are correct!  But still, this is quite shocking.  This is clearly due to the obvious dispersion in both its neighbors and DS 25-30; these simply have stolen its hits.

This is why the STR figures are SO important to our play, they automatically detect dispersion and our rules make sure we follow their guidance.  I have mentally lost count of all the PLE's I have dodged because of their warnings; I haven't had a PLE for 2414 BU!

The amazing thing is that it is STILL the KoH!

I am going to keep spinning and see what I can learn...



Just illustrating here how sessions can sometimes manifest.  Even though I have been able to take coups with the STR rules, it is slow going and I am not sure it is as safe as it normally should be.  I am considering making a rule about ending sessions that devolve into dispersion...

Compare the above session to a normally focused session below:



I have been spinning through and betting selectively with the STR ratings "advice" and I am noticing that the sequence "cannot make up its mind" which is the KoH, even though one will take a strong lead it will then be replaced by the other which catches up and this causes selections to go deeper into the progression than normal, during a properly focused session.

I think its dangerous to bet during times of dispersion even though the STR ratings still seem to work.  I am going to follow my instincts based on what I have seen and reset this session.

I believe that eventually, the random flow will resolve itself after enough spins but I have seen some wild gapping and I would rather not risk testing funds & time to find out how dangerous it really is.



Just a quick note on the new session, I find it refreshing because we have a clear KoH and the hits come fast.  I think its probably best to not be betting during objectively dangerous conditions; the key is always to be clear about what roulette is actually doing, right now in the present and make the best decisions based on that.







Just noting the new session.  A classic KoH & QoH battle.  Not sure which one will be the long term KoH but 31-36 is battling quite hard as QoH.





Ok, so here after 600 spins, we have a clear case of dispersion with DS' 1-6, 22-27, 25-30 & 28-33.

Why is this dangerous?

1) This is no longer the beginning of a session but an established statistical sequence
2) When the sequence brings one selection to surpass another, there is often a PLE with the selection that has been overtaken
3) The maximum number of spins where a KoH can be expected to be overtaken during a single progression run is 56 spins.
4) We have FOUR DS' within range to overtake the KoH at the same time
5) When many selections are hitting quickly at the same time, too many numbers have a higher priority, which means that our selection is more likely to gap

If we look at all of the above named DS' on the DSA we will see that all of them are bettable (with the exception of the fact that none of them are the fastest selection), in that they have at least 4 active STR numbers.

I am sure that we can spin through from this point and the sequence will arrive again at a clear KoH but so far, I see no reason why this is necessary.

Resetting the session.

Open_Demo_4





Once again showing dispersion but this time its DS' 1-6, 4-9, 10-15, 38-33 & 31-36.  When dispersion arises, we will notice that our selection will rarely have the STR strength to bet on; after all, they are "dispersion detectors".

Rather than spin through, I will just reset.  Added a last spin highlighting to the main display, right on the output numbers!

Open_Demo_5

Despite the small dispersion, I will continue to spin through and display the results once things solidify.  No PLE for 2898 BU!  This is around 900+ coups! :D





The dispersion finally subsided after almost 200 additional spins.  The clear QoH is DS 28-33 and everything else is out of range.
« Last Edit: September 19, 2017, 11:01:30 PM by Reyth »
 
The following users thanked this post: TERMINATOR

dobbelsteen

  • Hero Member
  • ******
  • Posts: 1501
  • Thanked: 481 times
Re: The Bomb Shelter
« Reply #20 on: September 14, 2017, 09:02:24 AM »
Is the Bomb shelter playable in the 10 seconds of Mr.J.
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #21 on: September 14, 2017, 02:10:14 PM »
Yes but not INSIDE of a B&M casino because they will not allow computers, right?  A 10 second spin timer is ideal for this system (online wheel).
 
The following users thanked this post: TERMINATOR

dobbelsteen

  • Hero Member
  • ******
  • Posts: 1501
  • Thanked: 481 times
Re: The Bomb Shelter
« Reply #22 on: September 17, 2017, 11:56:31 AM »
Reyth your answer means that you need a computer. A computer without a program or software is not a tool.

"online wheel" does that means a live online roulette?
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #23 on: September 17, 2017, 03:38:33 PM »
Yes a live online casino will allow you to use a computer (with software) from home.
 

dobbelsteen

  • Hero Member
  • ******
  • Posts: 1501
  • Thanked: 481 times
Re: The Bomb Shelter
« Reply #24 on: September 18, 2017, 07:37:11 AM »
I have tried to download the bomb_analyzer ,but that was not possible. The computer needs a special program, I do not posses
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #25 on: September 18, 2017, 01:55:55 PM »
http://www.qb64.net/

Download and upack.  Then put both program files into the same folder and run the .exe
 

Real

  • Fighting the war on absurdity one foolish idea at a time.
  • Hero Member
  • ******
  • Posts: 1416
  • Thanked: 211 times
  • Gender: Male
Re: The Bomb Shelter
« Reply #26 on: September 20, 2017, 03:22:01 AM »

Reyth,

I won't comment on the system.  But I do want to say this:

Your thread is artistic, visually stimulating and it's laid out well.  I can tell you went to a lot of trouble creating it and put a great deal of time into it's layout.  I suspect that you work in graphics/web design and that you do great things, and enjoy your career.  The thread is physically appealing to look at.

Good job!
 
The following users thanked this post: Jake007

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #27 on: September 20, 2017, 03:35:53 AM »
Yes, well eventually I knew you would be forced to comment on it and the ONE thing I know is that you CANNOT dispute the data; it is irrefutable.  Regardless of what you believe, its about numerical fact.

The great thing is that regardless of what you believe, I still make money! :D

Of course I am still testing and have a ways to go but I have never seen anything like this in roulette before, ever!

I know that I am blessed by God. :D







Wow.  So had my first PLE failure on the KoH with a 4 STR; DS 4-9 went until about spin 900 before being overtaken.  I went for 3084 BU without a PLE!  That is roughly 950+ coups (0.000288 or 1 in 3472 event)!

I took this hit from the Sun Deck and so I am only down -443 BU.  Neither number is suitable and so I must stay in the shelter until one of them comes back in.



Making a note here that this PLE occurred when I played a 4 STR and the very close QoH had something like 3 STR; the failure occurred at the same time as the takeover.



Level 2 Recovery.  Target: +7000





Level 1 Recovery. Target: +7064

Ooops.  Accidentally reset the comp, so starting a new session.  Clearly we can see the KoH & QoH (DS 4-9) pattern which has persisted for over 1000 spins!





The basic picture has formed with a clear KoH but obviously early with some disperson regarding the QoH in DS' 4-9, 28-33 & 31-36.

Fixed the DSA cursor and it was a variable conflict as I had suspected.

Let's grab some moar profit! :D



Just an important note:

Long-term bias with the HS Ratio keeps us safe from the Reverse Engineering CurseTM because it keeps us focused on the the selection that is the fastest overall; i.e. we do not chase favorable short-term indicators with the STR ratings.

I think this is a big mistake with alot of systems, where we shoot ourselves in the foot by trying to "trick" roulette into giving us better odds, instead of just working naturally with the odds the way they are; there is a natural short-term ebb and flow that roulette will NOT let us disturb.







Obviously EXTREMELY bad dispersion here with the whole board against us except for DS' 4-9, 7-12 & 10-15.  Rather than spin through for however long it takes to gain clarity, I will simply reset this session.

Documenting the new session where eventually our KoH was surrounded in a time of dispersion.  Here is the status just before the PLE:





Here is right after the PLE:



I was in the Shelter because there were 4 STR numbers but DS 1-6 was at King Tide (DSA Rules).  Once again we see a failure right at a time of QoH takeover and ALSO in a time of overall dispersion; DS' 1-6, 4-9, 13-18, 19-24 & 28-33.

I still was able to gain +104 BU despite the dispersion.  This was a session that I was strongly considering for a reset but I liked how DS' 22-27 & 25-30 were looking; they were tied in HS Ratio with the single street between them 100% fallen asleep!  I kind of had to spin through to see what would happen.

Clearly this is not the ideal situation yet that we normally see by spin 400 and we can choose to either reset or spin through until it solidifies.



The disperson here is continuing and so I am going to reset.  The excellent thing is that resetting completely changes the random focus; we will get a brand new statistical picture to work with!







Here is the new session and at about 260 spins, we finally see the classic picture where we have a QoH in DS 25-30. 

DS 22-27 is barely in range and can be slightly discounted as a threat since it is also neighbor to the KoH (some of its "threat potential" is actually fueled directly by the KoH).



BOMB SHELTER ALERT! An important message from Pales has just come in!


The question is how to play this system in a way that A PLAYER HAS THE ADVANTAGE. The logic is that a crystal clear TREND (in this case 3 rows), can not suddenly disappear out of the sky, as if someone flipped a switch. It could, but it is much more rare to disappear, than to continue for at least ONE MORE TIME. THAT ONE MORE TIME  is all you need to make the damage.

With our system this relates to the long-term trend in the HS Ratio AND the short-term trend with the STR figures; BOTH have to fail for us to get nailed on a tight bet during dispersion.  In other words a long-term LotT hot group of numbers that are currently highly active have to all fail before we lose.







Documenting the KoH slowing down tremendously and the uprising of multiple dispersion candidates in range, most notably DS 13-18.  I would rather not be betting in this environment and so I am resetting.

At least we are able to detect dispersion when it occurs and take corrective action whether it be spinning through until things normalize or resetting.





New session humming along quite well.





Documenting a strong KoH session that has persisted for over 1000 spins AND a time of dispersion.  Normally this would be a betting situation with a 4 STR for our selection but because DS 10-15 (QoH) has a 5 STR and is within 58 HS Ratio, I decline to bet due to the higher risk of a takeover & gap.

The KoH phenomenon is clearly shown with our selection going 2.2 times expectation for the number of coups taken without a gap.





Documenting here what appears to be a possible COG with DS 31-36 (QoH) moving very rapidly and the KoH slowing down. 

Its been a battle the whole way with "minor dispersion" where selections that are out of range (59+ HS Ratio points) hit frequently but none of them individually enough to close into range.

The KoH phenomenon is still clear after 1500 spins!

We can expect that one of DS' 13-18, 22-27 & 31-36 is going to gap & the other two will assume KoH/QoH positions.



Documenting the first ever recorded failure of a 5 STR KoH with no enemies within range (and thus no takeover).

This is right before the PLE:





Right after the PLE:



-443/+7274 (+179)[0]G124

Well, of course it was from the Sun Deck.  This run was for +652 BU, so I can't very well complain, showing an objective profit of +209 BU, despite the loss.

NOTE: The casino ended my session and so I am forced to reset.

Level 2 Recovery.  Target: +7607
« Last Edit: September 23, 2017, 08:27:52 AM by Reyth »
 

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #28 on: September 21, 2017, 03:01:45 PM »
1) The method may be as simple as just spinning freely and checking the DSA after every coup.

2) The true idea with neighbors really revolves around the structure of the hot numbers within our selection, where BALANCE is the key, as follows:

H...H..........X...X
X...H..........H...H
H...X..........H...H
Balanced

H...H..........H...X..........H...H
X...X..........H...X..........H...X
H...X..........H...X..........H...X
Unbalanced

Where an equal amount of HOT numbers appear on both sides is balanced but an unequal amount is unbalanced; 5 or 6 numbers is always balanced and 3 or less numbers is always unbalanced.

It could be as simple as:

If our selection is balanced, it doesn't matter what our neighbor is doing, our position is safe as long as our HS Ratio remains KoH and the QoH is not approaching.

We are trying to make sure we don't have a weak or sleeping street in our selection. ;)



Here is the method that I have settled into that is working very smoothly:

1) Working with an established KoH, spin freely until a coup is obtained.
2) Check the DSA:


6 STR numbers is King Tide, raise a half-unit
5 STR numbers is High Tide, bet regularly
4 STR numbers is Mid Tide, use the following:

a) If the QoH is far away (58+ points HS ratio) bet regularly
b) If the QoH is closer but has less STR numbers bet regularly
c) Otherwise hit the Sun Deck or enter the Shelter


3 STR numbers (or less) is Low Tide, enter the Shelter or at least hit the Sun Deck (I enter the Shelter).

3) Always switch to the highest HS Ratio selection after every coup.






Prior to and immediately after switching to hot selections, was more of a test of the recovery technique.  Until I implement the DSA method, it shows the efficacy of the bet selection.  After implementing the DSA method, the power of the STR rules is demonstrated.

It may be possible to now play this method with only 5316 units.
« Last Edit: September 22, 2017, 11:12:22 PM by Reyth »
 
The following users thanked this post: Jake007, TERMINATOR

Reyth

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3971
  • Thanked: 1280 times
Re: The Bomb Shelter
« Reply #29 on: September 24, 2017, 04:50:54 AM »




Documenting this PLE which occurred in an early session where a 5 STR KoH completely failed.





Going to reset this session because of dispersion.





Level 3 Recovery.  Target: +7504





Level 2 Recovery.  Target: 577.16






Resetting this session due to dispersion.  I am officially changing the way I open a game to only bet STR 5 KoH's or better until things solidify.





Documenting this PLE which occurred again in an early session with a 5 STR KoH completely failing.





Level 2 Recovery.  Target: +7623





Level 1 Recovery.  Target: +7716





Debt fully recovered.  The total drawdown was -2098 BU.

Due to these early game failures (reverse engineering curse), I am changing the opening rules to only betting 6 STR numbers until solidity is gained.

I am also going to stop resetting sessions because the longer they are, the stronger they get.



NEW STATISTICAL MONITORING IDEA

The short term trends run in small & quickly shifting cycles.  It would be nice and maybe somehow useful to observe these cycles for reference.

A possible form for this could be:

DS NAME
# OF HITS IN THE LAST XX (27) SPINS (color coded)
(scrollable)

This feature is already implemented on the charting & DSA but it is not presented so we can see them totaled all at once, at a single glance.

It is very easy to create a screen to display this information which is already being calculated and stored for the last 8000 spins.

GREY = 0 appearances
GREEN = 1-3 appearances
YELLOW = 4 appearances
RED = 5+ appearances

There are times when the "opposition" "shrinks" (i.e. declines) while the KoH advances and this will be easily shown on this screen.

I am thinking of implementing it on the MAIN screen with a color coding for the DS title, where the background will indicate by color the status of each DS.

This is going to be best.  Maybe some patterns can be discovered that relates to PLE's or up-trends/down-trends.

FURTHER MODIFICATION

There are two status elements that can be monitored:

1) Total actual number of hits -=- FACE COLOR [black(0), white(-3), cyan(4), blue(5+)]
2) Total numbers currently hitting -=-  BACKGROUND COLOR [grey(0),green(-3),yellow(4),red(5+)]


implementing....50% implemented (see bottom part of post below for the other 50%)



Imbalance is guaranteed to be produced by the LotT and because the number choices are limited to 0-36, some of these numbers must repeat more often than others over multiple LotT 37 spin sessions.

Congruently, these numbers that repeat more often are located on the felt which is also a limited field, so some of these numbers must group together in a confined space on this field; these confined spaces will always correspond to any of the "pre-made" bets like DS, Quad, Street & Split.

There must always be a top selection of numbers within a pre-defined section of the felt, even though shifts do occur from time to time; the larger the random sequence, the slower these shifts take place.

By definition, if a selection is appearing more often than expectation, it is not gapping very long or very often.

RNG BiasTM is a guaranteed phenomenon.



Updraft_DEMO





Documenting this PLE where a 5 STR failed on an uncontested KoH in a very strongly established session; the first reported of its kind.  I have no explanation but with all the KoH's I have played, this is obviously rare and of course subsequent to a profitable run, despite this loss.

Level 2 Recovery.  Target: +8104





Level 1 Recovery.  Target: +8189





Sun Deck Recovery.  Target: +8214





Level 2 Recovery.  Target: +8089





Documenting the first "Aged King Collapse" where the QoH & Rivals have surpassed the PLE RATIO of the KoH (HS Ratio superior), which means that the KoH has had PLE's more frequently than slower hitting selections -- this is not normally seen.

This PLE catastrophe of an entire section of the board was witnessed from the Shelter.





Level 1 Recovery.  Target: +8187

Unfortunately the casino closed this session before I could proceed further to see what happens next.





Level 1 Recovery.  Target: +8214





Debt fully recovered.  As it is plainly clear for every session, we see the classic pattern of a strongly established KoH.



The 2098 drawdown was due to back to back early session (reset) failures.  Who knows, we might be able to play this for only 2658 units...












Had to reset comp, documenting progress.  Unfortunately couldn't see if DS 1-6 (QoH) would actually overtake our selection. 

We can see our selection has had a PLE (which I dodged from the Shelter) and all other selections are vastly underperforming in the COUP:PLE ratio.

We can also see that the RNG Bias persisted for over 1500 spins since we recovered the debt.  There is a statistical reason for this phenomenon; we are dealing with three closed circuits:

1) Number results (0-36)
2) Felt locations (11 unique combinations of streets)
3) Random sequence length

*** There MUST ALWAYS be a top performer within all these areas because the LotT MUST ALWAYS produce repeaters and the sequence length represents the LotT just as much as 37 spins does; in this case, 2043 individual LotT sessions, chained together as 1. ***

[cite examples with simple maths]







Documenting this PLE during dispersion where DS' 1-6, 19-24 & 22-27 are all vying for KoH.  This is a rare move where DS 1-6 grabbed a 5 STR from DS 22-27 immediately after DS 22-27 made a similar move.  Treading water with this loss.

Level 2 Recovery.  Target: +8547





Level 1 Recovery.  Target: +8633





Sun Deck Recovery.  Target: +8657





Debt fully recovered.  This is not a typical session and I get the feeling these types of sessions should be avoided somehow. 

It is eery when a faster selection (DS 1-6) outpaces the "PLE KoH" (DS 19-24 with 0 PLE's) when it has a PLE and a BSE (3 total PLE's) and is running ABOVE expectation for PLE's:COUPS.  This is certainly not typical and is not what we want to be betting on.  Thankfully, I dodged all the PLE's from the Shelter, using the DSA Rules.

In fact, I am resetting this session.





This PLE took place in an early session BEFORE the first PLE had occurred and therefore it ended up being our selection! >.<  To be clear, the protocol for early sessions is:

1) No betting until a PLE has occurrred
2) No betting until there is clear solidity in the KoH

Immediately after this PLE I accidentally closed my browser and so resetting session by default.

Level 2 Recovery.  Target: +8551





Level 1 Recovery.  Target: +8632





OMGOSH!  A 6 STR failure where BOTH neighbors had a PLE at the same time!  I don't think I have ever seen this before.  Even though an event like this must be quite rare, it simply is not possible to avoid this.  What can be avoided is what led up to it which is the early session losses that occurred due to impatience.

Since this was High Tide, I was betting 3 chips (1.5 units) and therefore I must drop from Level 1 to Level 3 to gain the advantageous recovery ratio by betting 6 chips (3 units); this is a recovery ratio of 6:3 (2 to 1) and only requires 54 coups.

Level 3 Recovery.  Target: +8227





Level 2 Recovery.  Target: +8473






Resetting this session because there are too many rivals.

Level 1 Recovery.  Target: +8616





Noting here that the KoH is performing very well (uncontested) DESPITE the very low PLE Ratio and the "cursed" left section of the board.  At a particular point I was forced to accept that this was genuinely the KoH and started betting it at 6 STR.

Sun Deck Recovery.  Target: +8658





Debt fully recovered.  The total drawdown was only -1404.





This how DS 7-12 turned out.  It had another PLE which I dodged from the Shelter and it almost just had another one with a 4 STR with all competition at 3 STR or less; it went to the last bet.

This has brought me to think that it is not proper to play KoH's that have a negative PLE Ratio; even though the hits will deliver quickly, so will the PLE's.  In other words the risk is just too high.

So the new rule is:

Do not bet a KoH that has a negative PLE Ratio; either spin through or reset the session.

I am again noting that there are selections on the board that have 0 PLE's, where the KoH has had multiple (at least one is enough) PLE's.  This may ultimately become the rule, where if the best PLE ratio is not also the highest HS Ratio, we must spin through or reset.

Accidentally closed the session at +8720.
« Last Edit: November 03, 2017, 08:49:57 PM by Reyth »
 
The following users thanked this post: TERMINATOR