Another repeaters thread!

[daveylibra]

At least this post is less complicated than Falkors...!

Now regarding repeaters, anyone remember the quote "a number can't hit twice until it hits once, can't hit three times until it hits twice" from a member here?
I was one of the skeptics of his claim to win using repeaters, because I wrote computer programs to simulate what could happen if we bet on them, and the long-term was always a loss.

However, I've looked at it this way... suppose we have, 1 number that is a 3-show, 10 numbers that are 2-shows, & 12 that are 1-shows.
We know that an event will happen in the future, that is, a number will show 4 times. We know that this will happen within a finite number of spins. ( We also know that another number will become a 3-show etc. )
Now, with each bet we make we are betting on this "event" happening. ( a 4-show. )
There is only one number we can possibly bet on that will win if this event happens!
(I think the max. number of spins for a 4-show is 37+37+37+1, ie every number shows exactly 3 times, + 1 more spin. Equals 112 spins.)
We have had 35 spins so far, 112-35=77. A 4-show must be seen within 77 spins. ( Of course, for it to take 112 spins is VERY unlikely.)
So all I am saying is, is it valid to bet on a SURE event happening, until there are two 3-shows, when of course we bet on those 2 numbers?
I suppose similar points have already been made, but can any maths guys show the fault in the logic?

[Kairomancer]

The problem is that you cannot reliably predict which number will become a 4-show.
Any number can take the lead suddenly in just 4 spins.

Obviously a 3-show number has better odds to become a 4-show in finite amount of tries then an unhit number.
However i don't think you can make a profit in the long run with this strategy.

[Kairomancer]

I think it is better to wait until there is at least 4 3-show numbers and playing those with a negative progression and adding new  numbers to the selection as they become 3-show numbers.
It could be a killer hit and run system.
Obviously a large bankroll and a reasonable stop loss is mandatory based on statistics.

[Nimo]

Take Ayk's newest tracker, set it to 44 spins, fill the tracker, play the 2-4 most hit repeaters flatbet.  Just keep adding the numbers and the tracker will always show the highest repeaters which are in the 3-5 hit range by then.

[Scarface]

Makes sense to me.  Similar to my bets.  Great thing about it, is the amount of numbers you cover will always be small

[Wally Gator]

Nimo, why 44?

[daveylibra]

Nimo, does Ayk's tracker allow you to set a 'window' of a set number of spins to look back on? This would be similar to an idea I have.

[Nimo]

Nimo, why 44?

It has something to do with the law of large numbers,deviation and binomial probability.  A 20% cushion added to a set of numbers 37*.2 =7.4, rounded down to 7, added to the 37 numbers equals 44.  This helps to alleviate spikes in variance and gives a more true cycle of numbers.

But what do I know, I'm just a system player 

[Nimo]

Nimo, does Ayk's tracker allow you to set a 'window' of a set number of spins to look back on? This would be similar to an idea I have.

Yes, it is a great piece of coding.  Opens up a whole new way of working with things.

[The General]

What I don't understand is why if you can code are you running such small test samples?

[daveylibra]

...well if we want to run test samples, we should test the same way that we would play.
I (and surely most) would play a small number of spins, once or twice a day.
So if we can test like this, it would be valid. We would have to reset at a set number of spins/stoploss/target.
Then do this hundreds of times.

Now, my bright idea is this....
We track a set number of spins. 44 seems like a good idea. (Thanks Nimo.)
We then play our best 4 repeaters.
What are the best? We calculate like this-
Obviously the more number of repeats, the better.
Eg, If we only have one 4-show, play it. If we have two 3-shows, play them.
if we also have twelve 2-shows, we only need one to make up our 4 numbers.
So, we disregard the 1st spin, recalculate how many 2-shows. Then disregard the 2nd, 3rd, etc until we get the number of 2-shows required.
On every subsequent spin, we do the same, BUT we calculate the repeaters from spin 1 each time. We only start disregarding spins to eliminate 'lesser' repeaters.
Problem is, how could we track this? Difficult. I may be able to write a Basic program for it. RX anyone?

[Bigbroben]

Dave,
Is 44 spins a fixed window?
Could try something tonight about it.

Passion,
Your image of a pyramid can be applied as a dynamic fluid representation of the numbers hitting , hence the word " desintegration".
Ok, an average shape can be found by running many sims, deviations also.  The pyramid tip could "elongate" or go more flat, same for the base, from one session to another.  Interesting.  It can sort of melt, drip, or even get a "hard one" when using it on a rolling basis.
Mmh, will think of something and will see.

[daveylibra]

Bigbroben -

44 spins comes from Nimo, his post regarding binomial distribution seems interesting.
Alternatively, we could start at the 1st 3-show or 4-show or whatever...

[Nimo]

What I don't understand is why if you can code are you running such small test samples?

I don't code, Ayk does.  I just show samples to show how a method plays out.  For long term results I know I have a set requirement to earn a 3/1 ratio as in I need to earn three units per spin played.  Sometimes it will go to 2.42, sometimes it will go to 4.05, but my average is 3.00.  That is what I aim for in my casino with real money. I treat it as a business and those are my projected numbers that I need to reach my quarterly profit level.   I like to test with real money to see if results are as good as they are in simulations.  For the most part they are.  Surprisingly with some systems I test, I get better results with real money.

With proper bankroll and the proper system, it works well.  Here is a sample of over 600 spins with a +$2054 profit, works out to $2.9985 per spin ratio, so close to my average.  There was a large drawdown, but I knew with this method it would recover, knowing I had a $3000 bankroll to start with I used a $1 starting unit.