KTF

[Still]

Looking more closely at this today. 

If i've got the stats right, there is a 50% chance of getting a non-hit to hit, between spins 11-40.  That is, in a 30 spin window, we expect a non-hit to hit 15 times. 

So we look at the first ten spins, and bet on the remaining non-hit, which might be as many as 29 numbers, starting out, or 27 on average (usually 8 out of first 10 are hit).

Each time a number hits, it is removed from the next group of numbers to bet.  By the end of the next 30 spins, we might be betting as few as 14 numbers, but the average number of numbers bet will be around 21.75, given an initial bet of 27 numbers.  That is the average loss.  The average win, on the other hand, will be 15.25.   

In that case, the average loss over 30 spins will be 21.25 x 15 = -326.   The average wins will be 15.25 x 15 = +228.   

So the flat bet expectation would be to lose -97 units per game...if you were to always play to 30 spins. 

Applying a +1/-1 progression, you would always win (200+) given an equal distribution of wins and losses, alternating each time. 

When the win/loss distribution varies, that's when we see some big losses.  But does the progression prevail in the long run?

Still working on some tests, but it seems the progression is what makes this work, or watching the average expectations at the checkpoints, and betting accordingly. 

Or something else i can't see right now?

[The General]

Looking more closely at this today. 

If i've got the stats right, there is a 50% chance of getting a non-hit to hit, between spins 11-40.  That is, in a 30 spin window, we expect a non-hit to hit 15 times. 

So we look at the first ten spins, and bet on the remaining non-hit, which might be as many as 29 numbers, starting out, or 27 on average (usually 8 out of first 10 are hit).

Each time a number hits, it is removed from the next group of numbers to bet.  By the end of the next 30 spins, we might be betting as few as 14 numbers, but the average number of numbers bet will be around 21.75, given an initial bet of 27 numbers.  That is the average loss.  The average win, on the other hand, will be 15.25.   

In that case, the average loss over 30 spins will be 21.25 x 15 = -326.   The average wins will be 15.25 x 15 = +228.   

So the flat bet expectation would be to lose -97 units per game...if you were to always play to 30 spins. 

Applying a +1/-1 progression, you would always win (200+) given an equal distribution of wins and losses, alternating each time. 

When the win/loss distribution varies, that's when we see some big losses.  But does the progression prevail in the long run?

Still working on some tests, but it seems the progression is what makes this work, or watching the average expectations at the checkpoints, and betting accordingly. 

Or something else i can't see right now?

The probability of winning doesn't change from one spin to the next based on what numbers have or have not hit.

[Bigbroben]

It is really hard to find a proper goal and proper stop-loss with KTF.

It is always a trade off from, say, winning 50u 9 times and losing once 500u, or winning 400 times 5u and losing once 2500u.  Modify the goal and the game winrate changes, modify the stoploss: idem.

Did test tens of thousands of runs, using different +/- progs, flat, exponential progs (*2- /2), different starting spins, different triggers, like play only after first repeater, or play after 1st repeater and for 3 spins, reset at new high, usw...

It is overall a negative resulting game, at least when mechanically applying different values  and parameters.  (Sorry Notto).

Like you said, the ''bet accordingly'' part, or ''feeling'' part is hard to include in a programmed Excel sheet.

Nevertheless, enjoy this attached Little Helper.  Enter Zero as 37 in column B.  Choose parameters in column N. (Game start when both ''Start spin'' and ''start after repeater'' conditions are fulfilled.  To focus on one of these 2, enter ''0'' in the other cell).

P.S.: Can you find Waldo?

[The General]

It is really hard to find a proper goal and proper stop-loss with KTF.

If the system worked then it wouldn't require either a win goal or a stop loss.

[Madi]

If the system worked then it wouldn't require either a win goal or a stop loss.

In other way, if a “ broken wheel” really worked u wouldnt spend time here . More play means more money.

[Still]

Nice spreadsheet/tester.   Thanks.   You have some skills there BigBroBen. 


With a little more time, i would be tempted to automate some testing with the VBA backend, for example, to make a list of results for each session, adding up all the times it hit it's goal (+50) to see if it could ever overcome all the other times it went straight down to the stoploss.   


Am not optimistic though given the basic conditions: a 50% chance in a 30 spin window, with an average loss that is greater than the average win...trying to beat it with a progression, a target (which i tend to agree should not be necessary if it works), or closer observation of some checkpoints (trying to bet on reversion to the mean or on the trend). 

[The General]

In other way, if a “ broken wheel” really worked u wouldnt spend time here . More play means more money.

It's called "multitasking." More play means more money indeed!  But I still need to network in order to keep tabs on my opportunities.  It's gooood for business.

[Madi]

What kind of opportunity do u look by bashing  rules based system?

By the way did u see turbo ? The other day u said.

[The General]

What kind of opportunity do u look by bashing  rules based system?

By the way did u see turbo ? The other day u said.

Sorry, I just hate seeing recycled ideas that are absurd.

No, Turbo is afraid to meet up with me.

[Madi]

Sorry, I just hate seeing recycled ideas that are absurd.

Why dont u leave them so that they can learn from their mistake?

U didnt answer the question actually. What kind of opportunity u look for ? Like people who will find broken wheel for u? Or what?

[nottophammer]

i'll have some more in october, guess you'll have too wait, back to Random.org

[Bigbroben]

With a little more time, i would be tempted to automate some testing with the VBA backend, for example, to make a list of results for each session, adding up all the times it hit it's goal (+50) to see if it could ever overcome all the other times it went straight down to the stoploss.   

You can take the Excel file and run macros: it can run like one test per second. 
I'll soon take time to explain how to run macros in a Excel help thread.

[Still]

You can take the Excel file and run macros: it can run like one test per second. 
I'll soon take time to explain how to run macros in a Excel help thread.

That would be great.  I am hoping this can be done through Microsoft Onedrive.  Otherwise, i'll have to dust off an older laptop that has full Excel on it.  My current laptops don't run Excel except under 5 MB in OneDrive.   Maybe i can find a copy somewhere to run local, but space is scarce on these cheap net books.   

Meanwhile, it would seem to me better to reverse KTF and bet on repeaters instead. 

Starting with an average 8 numbers, there is apparently 15 chances out of the next 30 spins to get more repeaters: = a 50% chance.   Should be interesting as long as the number of numbers staked is less than or equal to 18.    At spin 30, the average number of numbers staked would be 8 + 15 = 23  but the average bet, through the 30 spins  would be 8 + (15/2) = 15 if i'm not mistaken. 

Can you set up your spreadsheet to bet on repeaters instead?  I could,  but you are already a step ahead, familiar with your coding. 

[Bigbroben]

Yes,
I did a ''Anti-KTF'' sheet, I thought it went more along what I liked, like having more and more chances of winning as the nrs fall.

I will enhance it a little then share.

[Bigbroben]

Here,

Rev-KTF.

New feature for rev-ktf: option of rising prog only after a win without new high.  Included infos on firts 4 repeaters, and the ordinal gaps.

Write nrs in the second column, write zero as 37, find Waldo.

No HolyGuac, just another salsa!