ZScore comparsion

[ophis]

I want to compare multiple systems according to their z-score.

In short zscore is calculated using Bets and Wins.

My question is:

If i will compare 2 systems then should they be compared by:
last 100 SPINS or last 100 BETS?

[F_LAT_INO]

I want to compare multiple systems according to their z-score.

In short zscore is calculated using Bets and Wins.

My question is:

If I will compare 2 systems then should they be compared by:
last 100 SPINS or last 100 BETS?

Ophis mate,
I think last 100 bets,as unbet spins wouldn't be of any value.

[ophis]

Yes bet lets say that:
system A in 100 spins have 87 bets and some wins.
system B in 100 spins have 44 bets and some wins.

You can still compare them..... and you will see which one have better strike rate in 100 spins.
But i'm not sure if this comparison will be correct or do I need to compare all systems regarding Bets not spins.

because system B need almost double the amount of spins than system A....

so system B would have to spin 200 time to be able to compare it with system A?
so how to compare them "in the fly" as u spin?

i think i need some rest because i cant get my head over it just now damit.

so any math guy?

[GLC]

Ophis,

I think we need to base it on bets for the z-score. 

I have always had trouble with the matrix systems because we had to track for such a long time before we could make a bet and we used a martingale which made it appear that the system was winning gang busters because the losses were such long times apart.  I reality, they weren't that far apart if we only considered the actual bets made.

We can then calculate the units per spin ratio to give us an idea of how long it will take to win (or lose) a reasonable amount.

George

[dangle]

A spin has no effect on the system or the outcome of that system unless you wager money on it.  The probability for each bet only applies to real risk, whether it's initial or compounded.  It does not apply to "tracking", "virtual bets" or anything else imaginary and made up.   

The z-score gives you a nice number you can use to gauge the degree of luck involved in the process of gathering your results.  To get those results, spins really mean nothing at all.  You would need a few thousand placed bets at least before you can start to get an idea if the system has merit or not.   

[Bazeegar]

Z score should be calculated on placed bets.

If you are using progression, I am not sure about the relevance of the Z score. I think you need to test for great number of placed bets compared to flat bets. You can plot Z score with each placed bet and your z score will change dramatically if you are using progression. I had tried to do it with flat bets. Please refer this thread link:://rouletteforum.cc/general-discussion/hit-and-run-vs-mathematical-probability/msg49895/#msg49895 Reply #30.

Bazee

[ophis]

I think you don't understand me.
i know how to calculate zscore, I know its based on bets.

my question is:
comparing 2 systems should I choose
last 100 spins or last 100 bets


let say it have been 200 spins and system A have 100 bets placed and system B have 150 bets placed,
then should I just compare those 2 zscores of last 100 spins (each system will have different amount of bets)?
OR
should I take spins containing last 100 bets from each system then compare them regardless of amount of spins each system had spun.?


I need to be able to compare systems in real-time running systems at once.

[Bazeegar]

Sorry I misundersood.

should I take spins containing last 100 bets from each system then compare them regardless of amount of spins each system had spun.?

I would do this way

[maestro]

has to be spins,,, and whatever system shows good run of wins(positive z score) for different sets of 100 spin then it must be worth playin

[ophis]

i'm asking because i'm about to release for further testing Global tracker....
program that will run multiple systems at once and choose on which system to play to catch rising trend of different systems. something like pattern breaker tracker but with more systems and with auto system select mode.

[dangle]

Compare the last 100 bets from the two systems regardless of how many spins it took to make those bets.  The number of bets really needs to be the same in this instance, whereas the spins don't.

Blank spins that are not wagered on are only pertinent for caluculating how practical the system is to play.   

Also bear in mind that every type of bet follows a probability pattern, you can see that well in the so-called "law of series" for even chances, so unless you have a system that is mathematically credited as a winner, what you are attempting to do may produce results that are surprising and unpredictable.   

[ophis]

hmmm not sure if u noticed when testing Patterns Tracker that depending on type of spins either PB or PF is winning. (not always ofc)
same story with CWB by flat_ino different mod works on different type of spins.

i want to make program that will choose which system to play.

[dangle]

I understand what you are trying to develop.  The problem is fundamental when using any tracking methods. . . when should you start playing and why?

There is no "sweet spot" when you can enter the game and expect to ride an upward trend.  When you do enter the game, the trend is just as likely to bomb out immediately as it is to continue.  It's the same in reverse, if you enter into a downward trend and expect it to go up, it's just as likely to keep going down.  With this in mind, you actually don't need the z-score, you can just use a percentage instead, which would easier to interpret.

Criteria for entering the game is crucial for your program, I imagine many variants would need testing before investing money in it.

[esoito]

A quick interruption, here:  The Z-score is important. Please be sure to include it, Ophis.

As to your question, why not PM Mr.Ore about it? He's 'way up there' in mathematics.