100 spin observation test

[Anastasius]

Looking at some sets of 100spins final outcomes  very common to see at least a 5 hitting number or two and also sometimes a few 7 hitters.

What about betting any 4 hitters until end of 100 spin set then re set. I think would have decent hit rate and would narrow down a way to predict a winner.

[Firefox]

Looking at some sets of 100spins final outcomes  very common to see at least a 5 hitting number or two and also sometimes a few 7 hitters.

What about betting any 4 hitters until end of 100 spin set then re set. I think would have decent hit rate and would narrow down a way to predict a winner.

You're going in the right direction  vis a vis repeaters. It's  recommended to review at least 800 spins to see if there are any trends.

[Anastasius]

I agree fox ... im also looking at huge samples . The law of large numbers keeps it very steady theres definate ways to predict. Im certain the fallacy of past spins is a fallacy in itself

Short term past spins dont matter  but long term they do matter. If u know what i mean.yes about repeaters

[Firefox]

Well, long term they might matter is probably more accurate 

It's surprising how random numbers performing well can get in the way, even over 1000's of spins.

If you can get hold of a book by Russell Barhardt called "Beating the wheel" you'll find it very helpful. Laurance Scott wrote some even better material and The General from this forum is a big expert on this.

[Anastasius]

As long as each number is written honestly even one spin a day all from different wheels the law of large numbers  will prevail.

In the early spins  many numbers up to 25 can battle to hit the most but later in line with law of large numbers only a few battle it out and new ones arise.

[Firefox]

In the 1000 to 5000 spins zone, it's still quite likely that random numbers can show well. After that number of spins  or even during, they may well rotate the numbers ring, so you'd have to be on the ball. Bias could still be there though depending on what it is, or could move around. Or rotors can be swapped between wheels or rotor and stator moved to another location.

But obviously mean is n/37 where n is number of spins and you want to be looking for a number(s) about 3 standard deviations  positive where standard deviation is SQRT(npq). n is number of spins, p=1/37, q=36/37

[Anastasius]

Fox do u think theres a limit once u get to about 500 spins. If u have 1 number that is the most hit . Then 1 or more numbers is equal as the most hit (for example 3 numbers have now hit 20 times and all 3 are most hit)

What do u think at the 500 spin mark is the max amount of numbers that are all equal  before one of them wins and becomes a new highest number?

. im up to spin 500 approximately the 30th repeater   ...

[Firefox]

After 500 spins, mean is 13.5.

SD is 3.62, so 20 hits is just over 2 x SDs out.

At 1000 spins mean is 27 SD is 5.1 so if your numbers maintain their lead in proportion they'll nearly 3 xSD. The SD is a function of the root of the spins whereas the mean is just proportional, so the significance changes as the numbers get bigger.

All I can say is stay on it and see if those numbers keep ahead or other ones come through.

Attached some graphs of what a sector bias looked like after 1000 spins.

These numbers were 2.5 to 3.0 SDs positive. This was caused by spin conditions, not wheel bias but it comes to a similar thing.

See link:s://:.rouletteforum.cc/index.php?topic=25908.0

[Anastasius]

Ffox . I like what your saying. ..

Can u just tell me in laymans terms one answer.

1. The maximum number of numbers that can all be in the most hit position after 1000spins before 1 becomes a new most hit number.. for example after 1000spins 3 numbers all hit 30 times then one hits and has now hit 31 times. Then later 2 have hit 31 times and one hits and becomes 32 times... do u understand my example?

[Anastasius]

I know that about 1/3 are above SD. Hitting above average but im not playing that.

In thr start about 22 numbers can hit twice before one number  becomes the first triple up....but later only a few can have hit 20 times before one becomes 21.. do u know how many is the max?

For example about 34 can hit once before forst double
22 can hit twice before forst triple
15 can hit 3 times before first 4

Etc

[Firefox]

Max number of numbers in the most hit position after 1000 spins is half of them.

Ie 18 numbers with about 33 hits, 17 numbers with about 21 hits. Not likely but possible. For every number up there will be one or two down and vice versa.

My graph is typical. There will be spikes but 12 numbers up, 12 breaking even, and 12 down is typical.

[Anastasius]

I only  have a max of 3 numbers as most hit equal after 29th repeat. For each level from 1st hit (just good luck)?

So what repeat number is it at just 2 numbers max...thats the time to bet ...unless the 2 numbers dont hit 200 times

[Firefox]

Best advice is look for 2.5 to 3 SD off the mean at that kind of level. Disregard any other numbers.

I'll post a simplified test later.

[Anastasius]

Your graph is showing half above average

Can u just bet the 18 most hit numbers and keep a running total?

[Firefox]

Personally I wouldn't do that.

Most of those are normal chance random numbers which will disappear as soon as they come hot.

The idea is to stay on it for a long time and bet only the ones which you are sure  are consistent front runners. 3 SD out.

With modern wheels that may well be none!   Or you may get 1 or 2 that are 1/36 or 1/36.5 numbers weak bias Negating or canceling some or all of the house edge, but not giving you an edge either.

The General is the man for this though. He can tell you much more.