### Author Topic: Observations  (Read 683 times)

0 Members and 1 Guest are viewing this topic.

#### Sputnik

• Veteran Member
• Posts: 606
• Thanked: 528 times
##### Observations
« on: October 22, 2017, 09:59:03 AM »

Harry wrote that some values are more common then others and i want to show one example how to observe variance getting stronger and weaker.

That 68.3% of the time the divergence would be one SD or less. Either side of the MEAN.
That 95% of the time the divergence would be  2 SD's or less.  Either side of the MEAN.
That 99.7% of the time the divergence would be 3 SD's or less. Either side of the MEAN.
That only 0.3% of the time would the divergence exceed 3 SD's

Now i did a small test using singles versus series of three
First you see the number to left that is the amount of singles before a serie of three hit
To the right you see the SD values for each sequence

Code: [Select]
`1 SD 0.02      A3 SD 1.18      B0 SD 0.00      A2 SD 0.73      A8 SD 2.51      C2 SD 0.73      A5 SD 1.82      B0 SD 0.00      A1 SD 0.02      A2 SD 0.73      A5 SD 1.82      B0 SD 0.00      A3 SD 1.18      B3 SD 1.18      B4 SD 1.53      B1 SD 0.02      A1 SD 0.02      A2 SD 0.73      A0 SD 0.00      A8 SD 2.51      C2 SD 0.73      A2 SD 0.73      A2 SD 0.73      A3 SD 1.18      B6 SD 2.07      C0 SD 0.00      A1 SD 0.02      A5 SD 1.82      B0 SD 0.00      A0 SD 0.00      A6 SD 2.07      C2 SD 0.73      A1 SD 0.02      AA 21 within 1 SDB 9   within 2 SDC 4   within 3 SDD 0   above 3 SD`
But it gets more complex then this example - assume i get eight singles and one serie of three then if i next get eight more singles and one serie of three i also reach a value above 3.0 SD
So when you look at the chart you look in both directions, left to right and up to down.

Cheers

« Last Edit: October 22, 2017, 10:00:36 AM by Sputnik »

The following users thanked this post: kav, Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1443 times
##### Re: Observations
« Reply #1 on: October 22, 2017, 04:23:45 PM »
Do the singles become more likely when there hasn't been a series of 3?

#### Sputnik

• Veteran Member
• Posts: 606
• Thanked: 528 times
##### Re: Observations
« Reply #2 on: October 22, 2017, 05:14:47 PM »

I not sure i understand your question - i have measuring the median value that show the 50/50 situation between singles and series of three - is this what you asking me - you want to know the median value?

I can tell you the following

A serie of three has 50% chance to hit before two singles show
Two singles has 50% chance to show before a serie of three show

The middle value or the median value between both events is one single

Cheers

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1443 times
##### Re: Observations
« Reply #3 on: October 22, 2017, 05:47:23 PM »
What I mean is if between say Red/Black, if there is 3.0 SD on a missing series of three in Red, is it more likely that Red gets hits than Black?

#### Sputnik

• Veteran Member
• Posts: 606
• Thanked: 528 times
##### Re: Observations
« Reply #4 on: October 22, 2017, 06:03:41 PM »

I can not answer you question - but i can mention that you can play one side only or using both.

When you get used to use only one side - then you feel like mixing cw and acw spins direction in the same way as using red and black - because then you get two random stream doing the same thing by them self and you mix them together.

Cheers

#### Scarface

• New
• Posts: 123
• Thanked: 101 times
##### Re: Observations
« Reply #5 on: October 22, 2017, 06:15:43 PM »
Can you give an example of a SD1, 2,and 3?

#### Sputnik

• Veteran Member
• Posts: 606
• Thanked: 528 times
##### Re: Observations
« Reply #6 on: October 22, 2017, 07:33:38 PM »

Yes i can give you example ...
Assume you get singles and series of two then you only count the singles

RBBRBRBRRBBRBRBRRBBB here you have nine singles and one serie of three 2.58 SD

I can show you the formula how to calculate the SD using math calculation - it is very easy
Bayes use the Z-Score and is the same thing and i made a chart for the values - so you can skip that part

1. Z-Score 0,02 -  1 singles contra 1 serie of three
2. Z-Score 0,73 -  2 singles contra 1 serie of three
3. Z-Score 1,18 -  3 singles contra 1 serie of three
4. Z-Score 1,53 -  4 singles contra 1 serie of three
5. Z-Score 1,82 -  5 singles contra 1 serie of three
6. Z-Score 2,07 -  6 singles contra 1 serie of three
7. Z-Score 2,30 -  7 singles contra 1 serie of three
8. Z-Score 2,51 -  8 singles contra 1 serie of three
9. Z-Score 2.58    9 singles contra 1 serie of three
10. Z-Score 2,70 - 10 singles contra 1 serie of three
11. Z-Score 2,89 - 11 singles contra 1 serie of three
12. Z-Score 3,06 - 12 singles contra 1 serie of three
13. Z-Score 3,22 - 13 singles contra 1 serie of three
14. Z-Score 3,37 - 14 singles contra 1 serie of three
15. Z-Score 3,52 - 15 singles contra 1 serie of three
16. Z-Score 3,66 - 16 singles contra 1 serie of three
17. Z-Score 3,80 - 17 singles contra 1 serie of three
18. Z-Score 3,93 - 18 singles contra 1 serie of three
19. Z-Score 4,06 - 19 singles contra 1 serie of three
20. Z-Score 4,18 - 20 singles contra 1 serie of three
21. Z-Score 4,30 - 21 singles contra 1 serie of three
22. Z-Score 4,42 - 22 singles contra 1 serie of three
23. Z-Score 4,53 - 23 singles contra 1 serie of three
24. Z-Score 4,64 - 24 singles contra 1 serie of three
25. Z-Score 4,75 - 25 singles contra 1 serie of three
26. Z-Score 4,86 - 26 singles contra 1 serie of three
27. Z-Score 4,96 - 27 singles contra 1 serie of three
28. Z-Score 5,07 - 28 singles contra 1 serie of three
29. Z-Score 5,17 - 29 singles contra 1 serie of three
30. Z-Score 5,26 - 30 singles contra 1 serie of three
31. Z-Score 5,36 - 31 singles contra 1 serie of three
32. Z-Score 5,45 - 32 singles contra 1 serie of three
33. Z-Score 5,55 - 33 singles contra 1 serie of three
34. Z-Score 5,64 - 34 singles contra 1 serie of three
35. Z-Score 5,73 - 35 singles contra 1 serie of three
36. Z-Score 5,82 - 36 singles contra 1 serie of three
37. Z-Score 5,90 - 37 singles contra 1 serie of three
38. Z-Score 5,99 - 38 singles contra 1 serie of three
39. Z-Score 6,07 - 39 singles contra 1 serie of three

Here is the result from 300 trails using red and black:
At left side you have how many singles hit before a serie of three show and at the right side the SD value:

1 SD 0.02
3 SD 1.18
0 SD 0.00
2 SD 0.73
8 SD 2.51
2 SD 0.73
5 SD 1.82
0 SD 0.00
1 SD 0.02
2 SD 0.73
5 SD 1.82
0 SD 0.00
3 SD 1.18
3 SD 1.18
4 SD 1.53
1 SD 0.02
1 SD 0.02
2 SD 0.73
0 SD 0.00
8 SD 2.51
2 SD 0.73
2 SD 0.73
2 SD 0.73
3 SD 1.18
6 SD 2.07
0 SD 0.00
1 SD 0.02
5 SD 1.82
0 SD 0.00
0 SD 0.00
6 SD 2.07
2 SD 0.73
1 SD 0.02
« Last Edit: October 22, 2017, 07:37:16 PM by Sputnik »

The following users thanked this post: kav, Scarface

#### Real

• Fighting the war on absurdity one foolish idea at a time.
• Hero Member
• Posts: 1668
• Thanked: 272 times
• Gender:
##### Re: Observations
« Reply #7 on: October 22, 2017, 08:20:36 PM »
Exactly what is the point of calculating the standard deviation in the red and black?

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4181
• Thanked: 1443 times
##### Re: Observations
« Reply #8 on: October 22, 2017, 09:11:31 PM »
Clearly he is looking for 3 hits in the next 3 spins.

#### Sputnik

• Veteran Member
• Posts: 606
• Thanked: 528 times
##### Re: Observations
« Reply #9 on: October 23, 2017, 06:42:07 AM »

Yes there is many options - assume i know when the 50% has vanish - then there is no idea to continue betting deeper into the hole - that is equal two attempts - so assume i staking positiv progression 1248 when there is 50% chance to strike at least two or more - then i don't have to waste more then two attempts for each attack - the reason to this is the known expectation

Made a short test for 150 trails and was up 60 units using positiv progression - so this has not only do with winning two in a row or catch a serie of three - when one side chop it can continue to chop and you have pretty good odds winning more then one, two, three, four, five times in a row and beyond

Cheers

The following users thanked this post: Reyth