### Author Topic: How to Test for Independence  (Read 20636 times)

0 Members and 1 Guest are viewing this topic.

#### Mike

• Veteran Member
• Posts: 796
• Thanked: 118 times
##### How to Test for Independence
« on: February 08, 2016, 05:10:07 PM »
A lot of gamblers and system players don't understand that roulette outcomes are independent, and have a hard time accepting the fact that using triggers and waiting for virtual losses or extreme events has no affect on their chances of winning. If this sounds like you, do yourself a favor and learn how to correctly test for independence.

There really should be no need for this in roulette, because it's clear that past spins don't indicate future spins (the same number of pockets on the wheel remain between spins) but maybe numbers will convince you if basic logic doesn't.

This is the correct way to test for independence:

The example in the vid has nothing to do with roulette, but the methodology is exactly the same. Suppose you believe that if  a dozen doesn't appear in the last 4 spins then it is more likely to occur in the next 4 (this is based on an actual example posted by a member recently). Draw up a table like the one in the video, but with these labels:

APPEARS                   DOESN'T APPEAR         TOTAL
------------------------------------------------------------------------------------------------------
MISSING                                  X                                     Y                           Tm
------------------------------------------------------------------------------------------------------
PRESENT                                  Z                                    W                          Tp
------------------------------------------------------------------------------------------------------
TOTAL                                     Ta                                    Td                           T
------------------------------------------------------------------------------------------------------

Now get some roulette spins and look at sequences of 4 spins at a time. For the first 4 spins:

if the dozen is present but missing from the next 4 spins, add 1 to the cell marked W.
if it is present and also appears in the next 4 spins, add 1 to the cell marked Z.
If it is missing in the first 4 spins but appears in the next 4, add 1 to the X cell.
if it is missing in the first 4 and doesn't appear in the next 4, add 1 to Y cell.

Continue like this until you have a reasonable amount of data, then fill in the totals. Then test for independence as instructed in the vid. Please post your results here for all to see.

Feel free to substitute your own bet if you don't like the example of a dozen in 4 spins. The procedure is the same.
« Last Edit: February 08, 2016, 07:04:56 PM by Mike »

#### Real

• Fighting the war on absurdity one foolish idea at a time.
• Hero Member
• Posts: 1668
• Thanked: 272 times
• Gender:
##### Re: How to Test for Independence
« Reply #1 on: February 08, 2016, 06:06:38 PM »
Good post Mike.  Let's just hope it's not over their heads.

I think the triggers guys naively believe that the triggers work because they're not comparing the same number of attempts with a trigger to the same number of attempts without a trigger.  In the other thread, Palestis' post demonstrated that he too doesn't comprehend how to test correctly.   In short their not comparing apples to apples.  I hope they study your post.
« Last Edit: February 08, 2016, 06:39:38 PM by Real »

#### Rourke

• New
• Posts: 106
• Thanked: 25 times
• Gender:
##### Re: How to Test for Independence
« Reply #2 on: February 08, 2016, 06:42:22 PM »
Here's some actual statical data covering 10.000 spins.

The first column "Number" shows how many attempts is made before RED is hit. The second column "Count" shows how many times the event will occur during the 10.000 spins

Number   Count
---------------------
13               1
14               1
12               2
11                5
10               7
9               19
8               34
7               43
6               96
5               191
4               331
3               614
2               1118
1               2147

I agree Mike, that past spins doesen't influence future spins, but you have to take probability into account.

#### Real

• Fighting the war on absurdity one foolish idea at a time.
• Hero Member
• Posts: 1668
• Thanked: 272 times
• Gender:
##### Re: How to Test for Independence
« Reply #3 on: February 08, 2016, 06:45:38 PM »

Quote
I agree Mike, that past spins doesen't influence future spins, but you have to take probability into account.
You have to take probability into account for what?  (I'm not sure as to what you're implying.)

#### Bebediktus

• Mature Member
• Posts: 284
• Thanked: 108 times
##### Re: How to Test for Independence
« Reply #4 on: February 08, 2016, 06:58:32 PM »

 14 1 1,00 13 1 0,50 12 2 0,40 11 5 0,71 10 7 0,37 9 19 0,56 8 34 0,79 7 43 0,45 6 96 0,50 5 191 0,58 4 331 0,54 3 614 0,55 2 1118 0,52 1 2147
I think he have in mind this - in third column is 1118/2147=0.52 then 614/1118=0.55 and so on ...

#### Mike

• Veteran Member
• Posts: 796
• Thanked: 118 times
##### Re: How to Test for Independence
« Reply #5 on: February 08, 2016, 07:00:28 PM »
I think the triggers guys naively believe that the triggers work because they're not comparing the same number of attempts with a trigger to the same number of attempts without a trigger.

Exactly. I think a lot of players just ASSUME that triggers give an advantage because they selectively focus on the marquee -- looking at how often they get a win ONLY after some losses. That's only half the story, if they looked at how often the win occurs regardless of what has happened they would hopefully give up the fallacy.

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1574
• Thanked: 244 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: How to Test for Independence
« Reply #6 on: February 08, 2016, 07:17:11 PM »
 14 1 1,00 13 1 0,50 12 2 0,40 11 5 0,71 10 7 0,37 9 19 0,56 8 34 0,79 7 43 0,45 6 96 0,50 5 191 0,58 4 331 0,54 3 614 0,55 2 1118 0,52 1 2147
I think he have in mind this - in third column is 1118/2147=0.52 then 614/1118=0.55 and so on ...

Hi Bebedictus, are you from Germany??

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1574
• Thanked: 244 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: How to Test for Independence
« Reply #7 on: February 08, 2016, 07:22:23 PM »
A lot of gamblers and system players don't understand that roulette outcomes are independent, and have a hard time accepting the fact that using triggers and waiting for virtual losses or extreme events has no affect on their chances of winning. If this sounds like you, do yourself a favor and learn how to correctly test for independence.

There really should be no need for this in roulette, because it's clear that past spins don't indicate future spins (the same number of pockets on the wheel remain between spins) but maybe numbers will convince you if basic logic doesn't.

This is the correct way to test for independence:

The example in the vid has nothing to do with roulette, but the methodology is exactly the same. Suppose you believe that if  a dozen doesn't appear in the last 4 spins then it is more likely to occur in the next 4 (this is based on an actual example posted by a member recently). Draw up a table like the one in the video, but with these labels:

APPEARS                   DOESN'T APPEAR         TOTAL
------------------------------------------------------------------------------------------------------
MISSING                                  X                                     Y                           Tm
------------------------------------------------------------------------------------------------------
PRESENT                                  Z                                    W                          Tp
------------------------------------------------------------------------------------------------------
TOTAL                                     Ta                                    Td                           T
------------------------------------------------------------------------------------------------------

Now get some roulette spins and look at sequences of 4 spins at a time. For the first 4 spins:

if the dozen is present but missing from the next 4 spins, add 1 to the cell marked W.
if it is present and also appears in the next 4 spins, add 1 to the cell marked Z.
If it is missing in the first 4 spins but appears in the next 4, add 1 to the X cell.
if it is missing in the first 4 and doesn't appear in the next 4, add 1 to Y cell.

Continue like this until you have a reasonable amount of data, then fill in the totals. Then test for independence as instructed in the vid. Please post your results here for all to see.

Feel free to substitute your own bet if you don't like the example of a dozen in 4 spins. The procedure is the same.

I'm just wondering what Kav has to say about it...!

#### Rourke

• New
• Posts: 106
• Thanked: 25 times
• Gender:
##### Re: How to Test for Independence
« Reply #8 on: February 08, 2016, 07:24:02 PM »

Quote
I agree Mike, that past spins doesen't influence future spins, but you have to take probability into account.
You have to take probability into account for what?  (I'm not sure as to what you're implying.)

I'm not implying anything. I'm stating that the probability of getting 13 black numbers in a row is lower than getting 4 black numbers in a row.

#### Real

• Fighting the war on absurdity one foolish idea at a time.
• Hero Member
• Posts: 1668
• Thanked: 272 times
• Gender:
##### Re: How to Test for Independence
« Reply #9 on: February 08, 2016, 07:26:58 PM »

Quote
I'm not implying anything. I'm stating that the probability of getting 13 black numbers in a row is lower than getting 4 black numbers in a row.

I agree.

Would you also agree that AFTER 13 blacks in a row have occurred that the probability of hitting another five in a row is the same as if they had not hit at all?

#### palestis

• Great Contributor
• Posts: 814
• Thanked: 743 times
##### Re: How to Test for Independence
« Reply #10 on: February 08, 2016, 07:45:17 PM »
Good post Mike.  Let's just hope it's not over there heads.

I think the triggers guys naively believe that the triggers work because they're not comparing the same number of attempts with a trigger to the same number of attempts without a trigger.  In the other thread, Palestis' post demonstrated that he too doesn't comprehend how to test correctly.   In short their not comparing apples to apples.  I hope they study your post.
I thought you were going to provide a real example to prove your point.
In my pic. (which I repeat below), I show the triggers circled (5 same color EC's), and observe what happens after that. Following all the circled triggers, there are not another 5 of the same EC's showing up.
It changed to the opposite before another 5 EC's appeared to make the group to 10 EC's in a row.
What kind of comparison are you talking about?
How would you compare apples to apples in this case?
In a way I remove all bad apples from the crate and only keep the good apples.
Y should I include all the bad apples for comparison purposes?
Besides, my goal in not to compare. My goal is to place a bet only when the conditions are favorable for me
How does a comparison fit in?
In any case only a real example will give credence to your point. So go ahead and provide one. Just like all good teachers do. Good comprehension comes with examples. Not with theory repetition.
My action is to circle a trigger and see what happens after.
If you were to provide a comparison how would you chose the other leg of the comparison?
(let's not forget that our goal is to create favorable conditions. Betting any EC without preconditions in not good comparison)

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1574
• Thanked: 244 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: How to Test for Independence
« Reply #11 on: February 08, 2016, 07:49:18 PM »

Quote
I'm not implying anything. I'm stating that the probability of getting 13 black numbers in a row is lower than getting 4 black numbers in a row.

I agree.

Would you also agree that AFTER 13 blacks in a row have occurred that the probability of hitting another five in a row is the same as if they had not hit at all?

Yes but doesn't it become 18 (13 + 5 = 18) in a row, thus less possible than 5 in a row??
When you see a streak of 10 successive high, do you perceive it as 10 times 1 or 1 set of 10??

I think the objective observation determines it as a streak of 10, not 10 single results and as we very well know a streak of 10 has much thinner chances than to hit just 1 time.
Besides if there was the same probability the casinos would pay exactly the same for winning 1 time or 10 times in a row for any given EC, but of course this is far from the truth!

If there was a complete independence, it could happen 100 Blacks in a row, it could happen 10 times a number in a row...etc
Your independence has limitations, otherwise there would be not possible to establish odds/payouts for each and every bet selection.

Probably you are confusing independence with variance, but variance doesn't mean that ANYTHING can happen!
EVERYTHING has limitations, every power, every being has limits, it's universal and if you are aware of these limits you may act properly.

Try to see the wider picture...

#### Bebediktus

• Mature Member
• Posts: 284
• Thanked: 108 times
##### Re: How to Test for Independence
« Reply #12 on: February 08, 2016, 08:03:11 PM »
Quote
Hi Bebedictus, are you from Germany??
No...

#### Mike

• Veteran Member
• Posts: 796
• Thanked: 118 times
##### Re: How to Test for Independence
« Reply #13 on: February 08, 2016, 08:07:18 PM »

If there was a complete independence, it could happen 100 Blacks in a row, it could happen 10 times a number in a row...etc
Your independence has limitations, otherwise there would be not possible to establish odds/payouts for each and every bet selection.

SIGH!

BA,

Independence has nothing to do with whether there could be 100 blacks in a row or 10 numbers in a row. As I've pointed out several times before, independence must be about least TWO events to make any sense at all -- you're comparing one event with another to see whether there is any correlation or association, get it?

#### BlueAngel

• I always express my opinion
• Hero Member
• Posts: 1574
• Thanked: 244 times
• Gender:
• Do you want truth? You cannot handle the truth!
##### Re: How to Test for Independence
« Reply #14 on: February 08, 2016, 08:16:34 PM »

If there was a complete independence, it could happen 100 Blacks in a row, it could happen 10 times a number in a row...etc
Your independence has limitations, otherwise there would be not possible to establish odds/payouts for each and every bet selection.

SIGH!

BA,

Independence has nothing to do with whether there could be 100 blacks in a row or 10 numbers in a row. As I've pointed out several times before, independence must be about least TWO events to make any sense at all -- you're comparing one event with another to see whether there is any correlation or association, get it?

In short term it might seem unrelated but in the long term everything falls into place and for one more time do not confuse variance with independence, are not the same.