A Distance Tracker experiment

[Kon-Fu-Sed]

Hello all,

If you've read my "Distance Tracker" thread in the "Software Section" you know that I hit a negative month after two positive.
I am, however, a curious person so... I just had to re-write my tracker: I added program-code for bets, losses and hits. And a bank-roll.

As I knew the outcome of the first three months I decided to use the fourth (April 2008) when writing the code.
And I decided to continue the experiment.


So... This post is an introduction and a kind of a table of contents:

If you want to know what the experiment is all about - read the next post.
If you want to know how I found the basis for the experiment - read the third post. More or less a short story.
There's also a short post regarding backwards ingeneering - the fourth post.

If you want to comment - write a fifth, sixth, seventh... post.


The results from this experiment will be posted in the "Test Section".
That thread will be locked as results are results and really nothing to discuss.


What will NOT be posted in ANY thread is the pattern itself and the rules.
Yes; there are IRON-CLAD RULES for the pattern and betting and thus NO HUMAN DECISIONS are necessary.
Those rules and all the outputs (results) from the tracker will be saved. Complete: Spin-by-spin, session for session...

The reasons for only showing bank-roll development in public, are posted in the "Distance Tracker" thread in the "Software Section".
If a member doubt the results, I will gladly let a willing member we both trust (third party) have a look and confirm.
(I guess this will only be necessary if I win a lot - I have not, yet )


In this experiment I use my computer and software so I will run THE SAME IRON-CLAD RULES from start to finish. NO guess-work, NO gut-feelings and NO hunches will be bet on.
Ever.
(And there will be no human errors.)

Because:
I want the results to reflect the potential for this way of analysing the wheel. I want to be able to make conclusions from the results.
To be able to make any kind of conclusion based on results of a test, I HAVE to have DEFINED RULES.


BUT!
BE FULLY AWARE OF THE FACT THAT THE RESULTS OF THIS EXPERIMENT ARE BASED ON [highlight]ONE SPECIFIC PATTERN[/highlight] AND THUS ONLY VALID FOR THIS TABLE AND FOR THE TIME THE EXPERIMENT LASTS.

*** THE PATTERN IS PROBABLY NOT - I REPEAT NOT - APPLICABLE TO ANY OTHER WHEEL AND/OR TIME.


[highlight]But my tracker is...[/highlight]


And that's why I prefer to call this an "experiment" instead of a "test": Each wheel must probably be analysed for its own, unique pattern ("fingerprint").
And something tested in a proper way should be possible to replicate for the same kind of results by anyone anytime anywhere.
This wheel's pattern will probably not give the same kind of results on another wheel.


Welcome to my "lab"!
KFS

[Kon-Fu-Sed]

THE IDEA behind this experiment:

My "Distance Tracker" makes nice patterns measuring distances between the ball's release-point and resting-point.
We can call this "mesuring the physical power".
I do not measure distances between hitting numbers - this is very important to remember.

In the case of a ball that is released when the last hitting number passes, it may be the same thing, but this is not always the case.
If there's another number passing when the ball is released, this number has to be used as release-number - NOT the last hitting number.


So instead of just simply use the numbers as streams of random independent numbers, we map the spread of the physical power of the ball-release.

It is of course not really true as increased power from whoever throws the ball, will probably make a bigger difference than just a few pockets.
A few revolutions is more probable.
And the opposite for a decrease in power.
(We have to be realistic - at least a bit )


But in my imagination I at least TRY to get free from the usual -2.7% that math and probability dictates - and I've never found any evidence for something else.
This is actual physical powers I measure.
That is true.
So the usual gambling theories (math/probability) should be out of the way.
Are they?
Will the end result of this experiment be different from the usual?
(Worse?)

That's for the experiment to reveal.


What I like, however, is the thought of those "Air-Ball" machines: A cabinet with a real full-sized roulette-wheel and the ball is shot by air.
And shot with good accuracy when the last hitting number passes.
They are mechanical and the air-power is controlled by a RNG.
And those RNGs are not random at all: It's a mathematical algorithm handling it.

And if they are not truly random... Maybe they make predictable patterns?


My idea in short:
[highlight]Is it possible to find a predictable pattern that will give a positive net - unit-wise?[/highlight]
If so: For how long will it last?

Well - that's what this EXPERIMENT hopefully will answer.

First I had to find a pattern - and I did.
I think...


I hope so...
I went through h**l for it.


/KFS

[Kon-Fu-Sed]

SO! I had this idea about using the tracker to analyze the live-dealer table #3 at Wiesbaden. To see if something could be found.
Just curious, that's all.

But I was told that contrary to what I had been told earlier, the dealers at that casino do NOT throw the ball when the last winning number passes the hand. They hardly ever look, I was told.
Too bad, as the release-number is essential for the tracker to be of any value.
(The custom of waiting for the number used to be the standard in Europe, I've been told)

So I came to think of an "air-ball" machine I've seen (the only one I've seen). It really released the ball when the last winning number passed the nozzle where the ball came from. And with a rather good accuaracy.
I happened to notice this once when two zeroes hit.

But that was long ago and far away.
And I never used the machine...


I began downloading the touchbet results from Wiesbaden and used the tracker to visualize them.
Maybe it's the same kind of machine and thus behaves the same?
There is a small probability.
Very small...


I found that looking at the charts with half-closed eyes helped me take away "details" and I thought I had a better picture of what was "happening".
There were "clumps" and "diagonals" and "forks"...

But those charts were 200 - 300 rows each. For each direction!
That meant a LOT of scrolling up and down and that was not useful; it was impossible to work that way. So I printed all the files from January 2008.
27 looong charts. (Jan 15 starts with a blank "result", 31 has a looong sequence of "10"s, and 29 and 30 are missing)
And I cut them in half and put the CC charts away for the time being.


Then I moved all the furniture in my living-room to the walls and spread the charts on the floor.



I began to study what was most eye-catching for me: The "diagonal" direction of the "clumps".
If you look at the charts you will se that many spins are clumped together and those clumps, quite often, make a "diagonal" pattern. Like a rhomb...?
(It's quite easy to spot with half-closed eyes)

After a while I began to realise that those clumps that began with the diagonal in one direction, every so often change it, and that would be just about the time that you have established the first direction.
Or the "clump" was simply splitted into "strokes".

Nothing really conclusive came out of that idea.
(For this time...)


But then I began to "see something" - don't ask me why but this pattern "emerged" during those days I was oozin' around on the living-room floor...
And after studying those 27 CW charts I could say with confidence that I THINK I found something.

At least it worked for that month...

So I wrote the rules.
I'm not posting them in full but for you to appreciate the volume of them, I give you the start of each rule:

Rule #1. DO NOT...
Rule #2. Look for...
Rule #3. The next distance...
Rule #4. Bet the distances...
Rule #5. Stop betting if...
Rule #6. Stop betting if...
Rule #7. If a bet is lost and...
Rule #8. If a bet is lost and...
Rule #9. Stop the session...

That's it. Nine simple rules.


So I checked the month of February and it was positive.

This was, however, done by hand looking at the files on the computer so mistakes were possible. Very possible.
But instead of double-checking what I had done, I started to think of what I MUST do:

The CC spins...
Finding the pattern and doing the first two months of CW spins was not a sinecure, I tell you.

But I'm a lazy bastard... Really... (Or so I've heard )
So I asked myself: "What if the same pattern works on CC spins as well?"
(The pattern itself can be mirrored so maybe it doesn't matter..?)

And believe it or not but the January CC spins ended positive!
Of course I immediately checked the February files as well and... Yes: Positive.

So I checked both directions for the month of March and they both were negative.


Now, as I said: That was done sitting by the computer, making notes on paper, and so the results may not be exact. But close enough.


I just HAD to write a version of the tracker that could automatically find the pattern and make the appropriate bet.
I also added a column to the output so everything - bets, losses, hits, bank-roll, whatever - were noted automatically.
(The output files will be saved as documentation/proof of failure/success.)

While I wrote the code, I had to test it so I used the April files for that purpose.
And when the program was complete it had revealed that the month ended positive.
For both directions (well...).

I decided to run all days from January 1st through the tracker and post the results.
I will keep on until...
Well...

Time will tell.

/KFS

[Kon-Fu-Sed]

Oooh yes...
Backwards ingeneering!

Well, I understand if you think it is. I did use the January CW results to find a pattern and bet-rules that won in January.

Yes.

But explain to me, please, the February and April results.
And why the CC spins seems to get the same kind of results from the same pattern even though they were not involved in the backwards ingeneering at all...

You believe whatever you want to, but this is the truth: ONLY the January CW spins were used to find the pattern I use for this experiment.

No other CW spins were used.
And no CC spins at all.

(There's a plan... )

/KFS

[ChickenDinner]

Great work as usual! [smiley=thumbsup.gif]

I look forward to seeing how this develops.

Cheers,
CD

[TwoCatSam]

KFS

Thank you for all this hard work.  Let me pose a question:  If I know for a fact the dealer pays no attention where the ball is released, does this render your program unworkable for that dealer/wheel/casino? 

I'm thinking it does.

Sam

[Kon-Fu-Sed]

Hi,

@ Sam,
You're correct: The tracker needs the release-number to work. If you have it there's only one problem:
In a gambling (not charting) situation you have to see it, calculate the number(s) and bet...

Because of that, I strongly suggest a dealer that throws the ball at the last hitting number


@ CD, and all,
May was a disappointment...

CW Started well below 0 but recovered - phew.
CC dropped way below 0 because of a 191u loss and then never recovered.

More to come
KFS

PS. One month takes like three hours to complete.
(Lots of addings and stuff)
So there will probably not be a daily post.

[VLSroulette]

Dear Kon-Fu-Sed,

Perhaphs you may want to use the wheels from Smart Live in this interesting experiment:

nolinks://nolinks.smartlivecasino.com/en/rouletteHome.aspx

You have a live wheel where presenters are picking numbers from last one spun AND an automated airball roulette which uses last number to initiate the spin.

It may be time-consuming to gather the actuals, but shall you find something worthy, you won't have to go to casino Wiesbaden as you can bet based on your conclusions online.

Kind regards.

Your friend,
Victor

[Kingpin]

KFS,

You have done it again - some great stuff 

Following this of course.

Best Regards
Kingpin

[Kon-Fu-Sed]

Hi all,

Thanks for all encouragement!



And thanks for the link Victor!

You are right; it might show useful in the future.
And Victor: I just love that chart-tag!

Half the fun in doing this is to see the charts.
Well...
Seeing the June results, I think I'll end this experiment now.

The CC result went dooown: Total net is now -946u. (But they were never in this for real anyway - although they started very nice)
The CW result was -12 for June... Still at +1161u though.

As I mentioned in a previous post I have some plans regarding the CC spins. In short it's finding a pattern on a short rolling basis.

But I have to have a way of finding the patterns...
The ocular inspection method I used for "pattern recognition" was way too tough for me.
Those bars danced in front of me at the end.
(That's why I didn't spend time to find a separate CC pattern)

So I'll go put my thinking-cap on, relax and see what shows up.


This is not the last you'll hear about this.

That's a promise!
KFS

[mistarlupo]

Kon-Fu-Sed,
Just wanted to say thank you! Really great effort, much appreciated.

I wish you good luck with your further tests, we are all on your side, bud!
We'll follow your posts with great interest, as usual...

Regards,
m

[alarian]

Maybe you should clarify what CW and CC charts are since it's sort of impossible to understand much without knowing what that means 

[lucky_strike]

Well i hope my spelling is correct

CW clock wise
CCW counter clock wise

Cheers LS

[alarian]

Hahaha... The most obvious is always the place to look *flaps the palm of his forehead* duh!

[ironguy]

hi Kon-Fu-Sed

your distance tracker seems to be interesting but I don't find it in download section
have u got a link for it

regards

[ironguy]

-