+ Register

[MarignyGrilleau]

Thanx Bally6543 for the very nice topic.  

In my opinion you can void house edge using a no zero roulette playing ec bets.
Bally suggests some ways to fight fluctuation:

We don't need to bet every spin.

We can have several methods of play which are interchangeable.

We can be properly funded for our sessions and set workable targets.

I think that the interested player might try and find bet selections with the lowest fluctuation possible. Computers are of great help performing thousands of tests in a few minutes. So we can have data to analyze, make observations and draw conclusions.
Anymore suggestions?
Maybe we can all contribute to a playing model based on this, if anyone is interested in sharing their findings.
Cheers

[TwoCatSam]

Before I risk me behind, let me ask:  Is variance the same as dispersion?

[VLS]

Is variance the same as dispersion?

Variance is a measure of statistical dispersion.

Standard deviation is another way to measure dispersion.


You can check informative articles here:

http://www.quickmba.com/stats/dispersion/

http://www.princeton.edu/~achaney/tmve/wiki100k/docs/Statistical_dispersion.html


They do have a different entry on Wikipedia:

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

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

[spike]

Seriously?
Sometimes i think you make this posts to make fun of us. We were trying to discuss something here, please make a valid contribution,

Thanks for addressing this. There was so much bad and just
plain wrong info in that post that I couldn't deal with it.
Its like, how do address nonsensical ramblings other than
just ignoring them. Good reply post on your part.

[TwoCatSam]

This seems like a good time for me to shut up! 

[Bayes]

Nice topic. 

Ralph makes a good point in that variance can also work for you, that's the rationale behind using positive progressions; make more on the winning runs than you lose on the losing runs, though personally I'm not a fan of this approach.

The simplest way to reduce the variance is to make more bets, because things tend to even out. The classical mathematical "solution" to negative expectation games is to stake your entire bankroll in a single bet (in the case of roulette, the bet should be on an EC). For some reason, mathematicians see the HA as the big bogeyman and ignore variance completely. 

But if you do what they suggest, it's very easy for the variance to go 100% against you (you lose that one bet, and thus your entire bankroll), but the more bets you make, the higher the chance that fluctuations (variance) will not be so severe. It might seem an obvious point, but it seems to be overlooked by many because of their preoccupation with the house edge.

On the other hand (unfortunately there's always a trade-off) the problem with making more bets is that you're giving the HA more of a chance to bite, so it makes sense to choose a game with the smallest HA you can find. And if you're going to make more bets you need to be adequately funded and not use aggressive progressions, as Bally suggested in his first post.

The harder way is to find bet selections which result in lower fluctuations, which is more fun, and can definitely be done.

[MarignyGrilleau]

An empirical way of speculating on fluctuation.
Based on Point and figure charts (Bayes), Something that worked for Sam, Pattern Random Vs Breaker....
Observe an Even Chance bet. Wait till there is an imbalance and then bet for correction within the same limits.


Random.org (300 spins) 29/11/2012

[attachimg=1]


W W LLW LLW


With Sam's Money Management it seems to work. Of course, the better the bet selection, the better the performance.

[Gizmotron]

Nice graph. It validates my premise that a global effect exists. You have a cluster of streak styled directions salted with flat styled back and forth. Sometimes they travel in your favor and sometimes they are against you.

[Bally6354]

Thanks guys for all the great replies. 

I always used to think that there must be some kind of 'HG' bet out there waiting to be discovered. I know different now. Variance can kill any system on any given day.

It's interesting reading between the lines from some of the more experienced members on here. My opinion is we are mostly saying the same thing.

It seems the trick is to be able to go with the flow and try and get the entry and exit points right. (not to be confused with HAR)

That flow can be very dynamic but I do agree with Gizmo in that there is usually a global trend happening. Although not seeing the wood for the trees can be a real problem. Developing a kind of detatched outlook is probably a good idea and one way of doing that is just to try and make intelligent bets (ie. timing).

I am convinced we can cut variance right down to the bone. Doing this certainly gives us much more of a fighting chance.

[Gizmotron]

Bally, I find it so easy to see it all in real time. It's my charting system. I can see the trends. I don't keep a track for the effectiveness trends. The graph above is the actual effectiveness line. It's an interesting idea to attempt to keep a trend path for the effectiveness. That process might benefit me with better self control.

[MarignyGrilleau]

Thanks guys for all the great replies. 

I always used to think that there must be some kind of 'HG' bet out there waiting to be discovered. I know different now. Variance can kill any system on any given day.

It's interesting reading between the lines from some of the more experienced members on here. My opinion is we are mostly saying the same thing.

It seems the trick is to be able to go with the flow and try and get the entry and exit points right. (not to be confused with HAR)

That flow can be very dynamic but I do agree with Gizmo in that there is usually a global trend happening. Although not seeing the wood for the trees can be a real problem. Developing a kind of detatched outlook is probably a good idea and one way of doing that is just to try and make intelligent bets (ie. timing).

I am convinced we can cut variance right down to the bone. Doing this certainly gives us much more of a fighting chance.

Present change is a constant, Statistical propensity is another one.
As it is obvious, no bet selection climbs or descends the graphic abruptly without hovering. There are only three movements: imbalance, correction and hovering.

[spike]

The harder way is to find bet selections which result in lower fluctuations, which is more fun, and can definitely be done.

But its the only way to defeat variance. You want variance
to be flat, no ups and downs. Players get all excited over
winning streaks, but they also mean there will be losing
streaks. A flat variance means you control the game and
its not controlling you. People would be far better off putting
all their time into bet selection than trying to find new ways
to fool the game with progressions and money management.
If you have flat variance, the need for money management
doesn't exist.

[Gizmotron]

That's a pretty amazing statement Spike. No wonder we are so different. I deliberately seek out  explosive variance. It's the only reason I go to the casino. Of course there's going to be crashing downturns. You don't have to let them take you out. This is funny. I spend the whole time trying to survive flat variance. I do this so that I can get to wild swings. It's so dang easy for me. Every spin tells you what condition you are currently in. You can take every huge downturn and turn it 180°, all you need is anything that continues. This is an amazing moment in forum history . I had no idea Spike was deliberately seeking flat variance. My stuff must sound like Martian to him. In fact it would be very difficult for me to attempt to read a flat variance. I can't actually say. I do everything I can to avoid it.

But its the only way to defeat variance. You want variance
to be flat, no ups and downs. Players get all excited over
winning streaks, but they also mean there will be losing
streaks. A flat variance means you control the game and
its not controlling you. People would be far better off putting
all their time into bet selection than trying to find new ways
to fool the game with progressions and money management.
If you have flat variance, the need for money management
doesn't exist.

[Bally6354]

[attachimg=1]

[AMK]

JL's PB Double Loss Strikerate 594/8 (74/1) 




Spike, if I remember correctly I once saw a method tested over 1 million spins that basically did not fluctuate at all. Stayed right above even, dropped down under a little then right back to a little above even etc  A member on vls forum showed the graph, rjeaton1
I can't find the thread now, might never find it again. The conclusion for this method was that it could not be used to produce signification profit etc


Would this kind of test result be useful for you to work with and use your insights on?