+ Register

[esoito]

Interesting to note serious horse racing punters maintain that a method must first be profitable with flat betting before it can be taken seriously, or taken further.

Why would that  principle not apply to other betting arenas as well? 

[XXVV]

@esoito


Sure. That is the basis on which I proceed also. If it can't work flat staking, there is no way that any overlay can overcome and protect from ultimate catastrophic failure or being frozen out as a flat-liner.


So there has to be at least some positive edge demonstrated after a suitable statistical sample.


I try to build on +10% edges and somehow you have to find that cluster.


Random selection of target numbers alone will not do it in the long term, ie +5000 spins say.


Better, 50,000 spins.


Clumping Theory and knowing when to stop at optimum gain ( in my view 3 hits) is a good start.  WF* Theory in clusters of some 22 spins + 'tendrels' indicates +10%. That is what I have commenced a new thread upon, in order to give a stepping stone. and launchpad.


That is all subject to further key testing of course.


Go the Tool !


XXVV

[esoito]

Yes, one of the strong points of The Tool in its previous incarnation was flat betting only.

And it sure was profitable!

[TwoCatSam]

"And it sure was profitable!"

esoito

I've never even heard of it until a few days ago.  If it was profitable, what happened to it.  Was it like a hammer and someone lost it?

Sam
   

[VLS]

Sam, the original Tool had more bugs than a can of worms.

I ended up with a crapola of sub-versions and fixing bugs in one introduced more bugs in the next 

Coding artificial intelligence is a b**ch.

Some versions would have to be reset after each hit, and have the numbers put manually in reverse until halt. Others would stop providing bets and people would feed numbers for quite a while before realizing it went "doo doo".

A whole mess of triggers adding numbers, filters taking them, multiple evaluators according to current conditions, future modelling, a layer of entropy in order to make it fuzzy enough for casinos not to perceive it as a deterministic "a to b" system across players, and the largest challenge of all would be to gain an edge, flat betting, while keeping all participants' state different, even on the same numerical stream, to avoid player collision and hence appear as a group of disconnected people, while keeping all with actual chances.

A b1tch to marshall a multi-player game where it must show like each user is on its own, powered by the same program. When too many players used it at once, it went kaput. Fortunately, halted instead of giving bad numbers.

Even coding an undo button with so many things going at once proved hell. Still is.



I always kept it as a semi-dormant side project. Now it's time to bring it back.

More than a re-write, this is a "re-ordering", so to speak.

The goals are high. Starting with the premise of going 1 to 1 against the casino, doubling a 100-unit bank on every seat flat betting. For each connected player. I don't know of a similar project deployed. Anyway, I'm crazy enough to take and face coding some projects which certainly shouldn't be taken by a solo coder 

Furthermore, it's an officially unbeatable game, you wouldn't think it was matter of stiff rules only and betting black after 5 reds...

[Albalaha]

Flat bet can be profitable only if: you are lucky enough to get more of wins than losses. Since the game is not predictable, you can't expect to win flat bet consistently ever, with any strategy that you make. Do not forget that ball comes to rest in a pocket, randomly and not as per your strategy. You can only expect and speculate, you can't predict.

[VLS]

Flat bet can be profitable only if: you are lucky enough to get more of wins than losses. Since the game is not predictable, you can't expect to win flat bet consistently ever, with any strategy that you make. Do not forget that ball comes to rest in a pocket, randomly and not as per your strategy. You can only expect and speculate, you can't predict.

You can predict the distribution of a set to a certain degree of certainty.

Please see Bayesian analysis and predictive inference.

http://en.wikipedia.org/wiki/Bayesian_analysis

http://en.wikipedia.org/wiki/Predictive_inference

You don't have 100% certainty; but in this field you don't need 100%...

Numerical sets are predictable, the casino is placing money on this fact.

How small can the predictable sample be? This is an entirely new field. We are blessed with heaps of processing power in these digital times to assist us in ways which weren't possible in the past. From the physical prediction standpoint (not 400 years, but as little as 80-90 years ago a predictive computer in a hand-held device was fiction), as well as statistical.

The goal is clear: the smaller the sample, the better chances for your predictions to be actually applicable to the game.

[Albalaha]

Vic,
      All these theories just sound good. Can you apply these approaches to make any "system" or any defined way of playing that can win with flat bet? Answer would be "no".

[VLS]

Vic,
      All these theories just sound good. Can you apply these approaches to make any "system" or any defined way of playing that can win with flat bet? Answer would be "no".

It may have been already achieved

Time and public evaluation will tell.

[Ralph]

In any largerr sample of numbers  0-36  there are more even numbers than odd, but it is not to any help as we do not get paid for all even numbers which hit, betting outside.

[Albalaha]

In any largerr sample of numbers  0-36  there are more even numbers than odd, but it is not to any help as we do not get paid for all even numbers which hit, betting outside.

        Where do u get such superb knowledge from?