Patient dozens

[ZeroBlue]

For the above suggestion here is a test made over Bayes "Hell Session nr.1"



    33
    22
    28   323
  0
    4
25       312
1
21
18
19
    8
25      323
    13
    26
    22
16
9
34       333
    29
    28
19
3
    11
9
    8
    17
    33    333
    28
30
    24
18
    6
    29   332
32
    22
    13
1
    11
3
    33   322
    22
14
    17
27        313
12
    35
    8
19
    28    331
30
    2
    22
    28    321    Trigger- From now on we will bet against the formation 311 (sleeper) to happen.
19
    8
    28    333 WIN
34
    35
    24
    26   333 W
    29
    35
    22
    29   312 W
1
    22
    33   313 W
    2
34
    22
    6
12
9
    33   321 W
    24
7
    15
    31   321 W
19
1
    17
    11
19
    20
32       322 W
    15
    15
34       322 W
19
    13
34       322 W
    13
    28
    20
    33   313 W
9
    31
    15
    35  321 W
    13
    6


These are several attacks displayed until a loss but we should aim (in the beggining) for one Win and restart.
Why?
...not because hit & run 


but because the frequency of hits in a window.
for dozen 3 we got our trigger much sooner than for dozen 2
10 virtual games Vs 16 virtual games until trigger.


If you run a simulator and get frequencies for how many games we need to complete a cycle of all 9 formations.
Then we can get optimal attack windows.



hope someone runs it in Rx or so.


Zer0Blue

[Drazen]

Very good idea Zero, thanks. Yes we should have different approaches to test and to conclude, which goes better.

Regards
                    Drazen

[Proofreaders2000]

A modified Pattern Breaker for dozens.  Interesting

If the strike rate is high enough on the first trigger of the original version, maybe a one time moderate to heavy bet might work...

[LuckyLucy]

As I stated in Zeros thread, its the consistnacy thats the issue I find with covering two dozens or columns. Surely there has to be a way to place consistant bets and win...hmmm thinking cap on.