I have run a 30 million single zero RNG test, 361 086 games (whenever a number repeats, a new session is started)
2 9799
3 19199
4 26712
5 33099
6 36817
7 38117
8 37424
9 34613
10 30241
11 25616
12 20579
13 15855
14 11481
15 8103
16 5429
17 3428
18 2057
19 1187
20 665
21 372
22 165
23 79
24 35
25 10
26 3
27 0
28 1
MEAN 8.308740854
MEDIAN 8
Max number of spins before a repeat is 28.
Here is another chart:
SPIN Theoretical Actual
2 2.702702703 2.713757941
3 5.405405405 5.465331766
4 8.108108108 8.04365108
5 10.81081081 10.83876925
6 13.51351351 13.52189131
7 16.21621622 16.18831224
8 18.91891892 18.96393589
9 21.62162162 21.64408232
10 24.32432432 24.13372065
11 27.02702703 26.94577394
12 29.72972973 29.63181615
13 32.43243243 32.4432167
14 35.13513514 34.77510223
15 37.83783784 37.62886598
16 40.54054054 40.42141315
17 43.24324324 42.83929018
18 45.94594595 44.97157849
19 48.64864865 47.15931665
20 51.35135135 50
21 54.05405405 55.93984962
22 56.75675676 56.31399317
23 59.45945946 61.71875
24 62.16216216 71.42857143
25 64.86486486 71.42857143
26 67.56756757 75
The left number is the spin number before a repeat. The least amount if 2 of course. The next number is the theoretical percentage of hits. The number after that is the actual based on my tests.
So for example first line:
SPIN Theoretical Actual
2 2.702702703 2.713757941
For spin 2 (betting 1 number, it will always be 1 less than the spin number) the theoretical percentage to hit the 1 number is 1/37 or 2.70% The actual one in the test is 2.71%. As you can see, the theoretical is very close to the percentage I got from the 30 million test.
What does that tell us? It tells us that there is no simple advantage to waiting and then betting. For example waiting for 10 unique numbers in a row, then betting for or against them.
I have been reading up on Dyksexlic's posts about this so called Pigeonhole Concept, and can not understand how he can beat every session.
In 30 million spins the max number of unique numbers was 28. This is the progression you would need for that situation in order to profit at any spin when a repeat happens:
SPIN BET/#
1
2 1
3 1
4 1
5 1
6 1
7 1
8 1
9 2
10 2
11 3
12 4
13 6
14 10
15 16
16 27
17 49
18 93
19 186
20 393
21 885
22 2124
23 5461
24 15123
25 45369
26 148480
27 534528
28 2138112
29 9621504
Must be something more to it, if in fact he was telling the truth. :thumbsup:
Irish
2 questions
What RNG was used?
Could it be done same for actuals.
If the random seed was not randomized or significantly changed every 1/2
million spins then the same random sequences where repeated 60 times
approximately. What platform, operating system, and computer language
was used also matters for these type of tests.
Quote from: crackers on June 28, 2012, 06:19:06 PM
If the random seed was not randomized or significantly changed every 1/2
million spins then the same random sequences where repeated 60 times
approximately. What platform, operating system, and computer language
was used also matters for these type of tests.
It wouldnt take much to have a file of actuals of that size by downloading a year of German actuals on RX set to go into 1 file
Quote from: Turner on June 28, 2012, 06:02:59 PM
Irish
2 questions
What RNG was used?
Could it be done same for actuals.
Yes it can be done for actuals, but I really see no point. The stats will be very similar.
Quote from: LuckoftheIrish on June 29, 2012, 04:18:52 AM
I really see no point. The stats will be very similar.
Good job Alexander Fleming didnt say that :thumbsup:
Its allways best to quote the results rather than guess or assume what the results probably would have been.
Quote from: Rdick on June 30, 2012, 05:38:04 PM
I thought Fleming was a chemist...!
Hmmmm...it was a shot at comedy. Fleming didnt guess .
Tough audience!