### Author Topic: Maths Wiz  (Read 2636 times)

0 Members and 3 Guests are viewing this topic.

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Maths Wiz
« on: July 30, 2016, 04:41:19 PM »
Hi new to the website.

Testing a system with live autowheel data. I have 20,000 spins recorded by myself to test a system which soo far has not lost in 20,000 spins and made a profit of about £4,000 using a bank of £1,700.

Would I would be greatly know the below...

I have worked out the odds of 12 numbers not showing in 10 consecutive spins is

25/37^10 = 0.0198 (approx 50/1)

The odds of this happening twice in a row is

0.0198 X 0.0198 = 0.000392 (approx 2500/1)

But from my data of 20,000 spins it shows on average that 12 numbers not coming out in 10 spins is actually 1/200 and not 1/50!

Would this be because I'm using the same 12 numbers each time...and then waiting for one of my 12 numbers to land before I start betting...instead of playing any of the 37 number of the table!??

Also say for example my special 12 numbers are

1,2,3,4,5,6,7,8,9,10,11,12

If number 7 lands for example, I place my bets on the above 12 numbers and play them over 10 spins and hope one lands so I wins and then start the process again.

Say if number 7 landed and I placed my bets on my 12 number over 10 spins and none won.

Then the next time 7 comes out it "should" win within 10 spins.

The odds says 1/2500 chance of this not happening but again I have played over 20,000 spins and it hasn't happened once.

Based on the above example would the Maths reflect correctly what I am trying to work out in terms of probability!? As doesn't seem right.

Any help would be great...have asked a few mathes forum online but not everyone understands roulette and system.

Thanks

The following users thanked this post: kav, Reyth

#### kav

• www.Roulette30.com
• Hero Member
• Posts: 2227
• Thanked: 1170 times
• Gender:
##### Re: Maths Wiz
« Reply #1 on: July 30, 2016, 05:02:42 PM »
Hello and welcome to the forum.

This is a very interesting question.
My opinion is you should trust probabilities. 1/50 is the correct frequency of the event.
On the other hand, in such rare events it may take more time for results to conform with probabilities.
It could also be that the first sequence you faced were that 1/50 sequence.

Your actual results can be due to variance or faulty rng.

The following users thanked this post: Reyth

#### scepticus

• Hero Member
• Posts: 2581
• Thanked: 573 times
##### Re: Maths Wiz
« Reply #2 on: July 30, 2016, 05:12:41 PM »
It has often been noticed that in any 37 spins there are about 12 numbers that have not appeared. What has been called The Law of the Thirds. I suggest you  use the "missing " 12 numbers but I wouldn't recommend  a progression here. I assume that your profit was  play money ?

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Re: Maths Wiz
« Reply #3 on: July 30, 2016, 05:42:59 PM »
Thanks for the speedy reply guys!

This is not RNG this is live AutoWheel I have tracked this for 20,000 spins (450 odd hours)

Not played this for real money yet....but going 20,000 spins without a loser. That is a second round loser to confirm! Which should be 1/2500

Going to test up to 50,000 spins.

I think if I get to 50,000 without a single loser I'm off to the casino.

BUT...surely because I'm only using one third of the number (only betting off 12 set numbers each time) then the calculation is not correct??

That's what is confusing me!

For example

Hitting a double number is like 1/369

So if I was only tracking one number say number "9" I should get a double every 1369 spins, but if was tracking all 37 number hitting a double it would come a lot quicker!?

So my point is....I'm only tracking 12 set numbers and not all 37 number.

Soo maybe thay why I'm getting 1/200 and not 1/50 as I'm only play one third of all numbers!?

Also if it Loses the first time...im saying It want lose a second time in a row BUT only on the same number....NOT on any other number.

So again I don't think odds of 1/2500 are correct.

Driving me crazy!

#### kav

• www.Roulette30.com
• Hero Member
• Posts: 2227
• Thanked: 1170 times
• Gender:
##### Re: Maths Wiz
« Reply #4 on: July 30, 2016, 06:02:46 PM »
NO.
The 25/37^10 = 0.0198 (approx 1/50) is for 12 specific numbers to sleep for 10 spins - not any 12 numbers. So this is the correct frequency.
The formula for any 12 numbers is (?) 25/37^9.

If you add too many "security measures" like waiting for it to lose and only bet in the next trigger, especially in long odds like these, you will have a very long waiting time.

The following users thanked this post: Gzgzbee

#### Sheridan44

• Mature Member
• Posts: 279
• Thanked: 251 times
• Gender:
##### Re: Maths Wiz
« Reply #5 on: July 30, 2016, 09:40:55 PM »
Hi Gzgzbee  and Welcome!

Here is one avenue you might want to consider.

Rather than waiting for 12 pre-selected numbers..... begin betting on numbers as they occur from the wheel. Like say the first 10-12 of them.  The idea is to catch "hot" numbers. Many times they have an uncanny tendency to repeat for a while.

"Hot" numbers are actually a product of variance, which plays a role in the less familiar "maths". It took me some time to figure out that there is a math from the textbooks, and there is the math in the real world. Although in the latter stages, probability "behaves" as it should in the long run (house edge).....in shorter trials almost anything can happen. The trick is to exploit these "anomalies".
« Last Edit: July 30, 2016, 09:45:48 PM by Sheridan44 »

The following users thanked this post: Gzgzbee

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Re: Maths Wiz
« Reply #6 on: July 30, 2016, 10:04:13 PM »

Loving the forum!

Well just tested another 1100 live autowheel spins (24hours)

And £180 profit off using a bank roll of £1700!

Again in that 1100 spins it only lost 4 times and always won on the second round! Mmmmm.....

I have check all the 12 numbers I use to see how many they have come out over 20,000 spins and they range from 510-584!

Which is On average is about right.

20,000/37 = 540

Soo why is the over 1/50 odds soo out!?

As if 1/200 is true then odds of it happening twice is

200x200 = 1/40,000

Soo for every 24 hours I can profit approx £200

There are 37, 24 hours in 40,000 spins (approx 45 second per spin)

37 X £200 = £7400

And one loss would cost me £1700.

The following users thanked this post: Sheridan44

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1611 times
##### Re: Maths Wiz
« Reply #7 on: July 31, 2016, 02:33:22 AM »
Right, what Kav said.  20,000 betting events is a common benchmark for a reliable system but it can take a million or more spins for the results to conform to the actual statistics.  This is one of the reaons for performing tests of millions of spins; eventually the results remain static.

Your system sounds awesome.  Any chance you can clarify it more?  Are you betting a 10 element progression on a single dozen after a trigger?
« Last Edit: July 31, 2016, 02:36:13 AM by Reyth »

The following users thanked this post: kav

#### Bayes

• Veteran Member
• Posts: 688
• Thanked: 566 times
• roulettician.com
##### Re: Maths Wiz
« Reply #8 on: July 31, 2016, 07:31:28 AM »
Hi Gzgzbee,

I concur with Kav and Reyth on this. The most likely explanation is that your results are just due to luck (luck = variance). This can be quantified by standard deviation or Z-score, which is a measure of how far the results can deviate (in either direction) from the expectation or average, which in this case is a "success" every 2,500 spins.

I've just written a little calculator which I'll be uploading to my web site later, and putting in your values results in a z-score of -2.8:

Although your result is fairly uncommon, z-scores can range between -5 and +5 (sometimes even a little more in extreme cases) so I wouldn't get too excited yet and start believing that you've found a "holy grail".

Not saying you definitely haven't, just that more data is needed.

The following users thanked this post: kav, Reyth, Sheridan44, lone wolf

#### Bayes

• Veteran Member
• Posts: 688
• Thanked: 566 times
• roulettician.com
##### Re: Maths Wiz
« Reply #9 on: July 31, 2016, 07:47:42 AM »
It's worth keeping in mind that you wouldn't have made the post in the first place if the event had been "ordinary". All over the world there may be 1000's of people working on roulette systems, and purely by luck, a few will get some extreme results (or at least, what they think is extreme, although mostly it turns out that the results weren't so extreme as they thought, when you do the maths).

No system enthusiast would bother to present their results on a forum if the outcomes were close to expectation, assuming they know what the expectation is (hey guys, I've just invented this amazing system betting on red where I win 50 bets out of 100!  ), but when  "outliers" turn up they may seek some explanation or perhaps even trumpet loudly that they've found a winning system. In the vast majority of cases no explanation is necessary because the system results are within variance "tolerances", as will become evident when other forum members try to replicate the results (and fail).

The following users thanked this post: Sheridan44

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Re: Maths Wiz
« Reply #10 on: July 31, 2016, 09:04:44 AM »

How is best to test this over millions of spins!?

Not interest in RNG....but would be interesting to see if I get the same results.

Anyone know where I can get results from a autowheel!?

Thanks

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Re: Maths Wiz
« Reply #11 on: July 31, 2016, 11:02:51 AM »
Doing an interesting test this afternoon!

I have download 1 Million Live roulette results from the roulette30 website.

Going to test my system out using theses numbers.

So far I have looked at 3000 and I have had 9 first time loser and no second round losers.

Odds say first time losers happen 1/50 and second round loser 1/2500

so far 3000/9 = 1/333.

Early days....but will update you this afternoon/tonight

Should be able to check 100,000 + spins!

So far profit is

£563.00 of a £1700 bank!

The following users thanked this post: Reyth

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4393
• Thanked: 1611 times
##### Re: Maths Wiz
« Reply #12 on: July 31, 2016, 11:07:23 AM »
Wow G very nice!  Looking forward to hearing more...

#### Gzgzbee

• New
• Posts: 7
• Thanked: 4 times
##### Re: Maths Wiz
« Reply #13 on: July 31, 2016, 03:29:21 PM »
Cut short testing...

Got to 4500 spins and it actually lost and 300 on it lost again!

BUT I was told the system only works on the Autowheel machines and not with a live dealer spins the wheel!

Soo back to recording the autowheel for me then haha!

#### scepticus

• Hero Member
• Posts: 2581
• Thanked: 573 times
##### Re: Maths Wiz
« Reply #14 on: July 31, 2016, 06:00:31 PM »
Cut short testing...

Got to 4500 spins and it actually lost and 300 on it lost again!

BUT I was told the system only works on the Autowheel machines and not with a live dealer spins the wheel!

Soo back to recording the autowheel for me then haha!

So who told you that it only works on Autowheel Gz ?
Sounds strange . Did someone tell you about their success with it ?