Negative Extreme Variance Management and Positive Extreme Variance Exploitation

Started by Albalaha, Jan 22, 2023, 06:22 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


Variance means going away from average or expected value. It can be both negative or positive. In the absence of extremes any silly progression idea like Martingale, Labouchere, Fibonacci,Pluscoup or any of its variants would beat every sessions of any EC bet. In reality, variance would hit any moment to a level that is not bearable and tarnish dreams of earning using tradition mathematical progressions. Although, there is no absolute limit of how much variance could go north or south, we expect a virtual range of 5SD at both sides to be worst possible. Even if a bet goes say 6SD below mean, a well defined strategy could safeguard us still.
           Say, if in an EC bet we get Losses and Wins represented as W and L as this:
It will be an unbeatable case by itself and no sane progression could beat it. Someone using a traditional money management will lose huge. A flat bettor will lose lesser but he might not win it back until goes extremely lucky to get an even better positive variance helping him.
   It is harsh reality that even after an extremely bad start, the session might to balance itself but could remain tilted at one side. Say, after the first stretch, bet gets win and losses like this:
 Although, this stretch is not as harsh as the earlier one but still not good enough to win back all losses.
Say, we get next stretch touching the mean, it could look like:


How many of us will be at net win till these 90 bets?

Howm many of us will be losing the least during this stretch?

What will happen if first stretch type variance re occurs after these 90 bets? Remote but not impossible to happen.

As I analysed both books of zumma baccarat with over 100k hands without tie, I saw that in any 1000 hands, 448 hits were minimum at either side. Say, we start from the worst possible stretch, which strategy will safeguard us?

More later..........



🎁 Paypal me @

-- Victor


The very first step towards countering extreme negative variance is to understanding it. As I said earlier, it has no real limits or stretch till it can continue to be superharsh but in cases of EC we take 5 SD below mean as a benchmark. All experienced players know very well by their experience or observation/discussion that an EC might even go 1/30 or 25/100 but 50/200 is kind of unlikely and 100/400 or even 125/400 is impossible. Anyways, the harshest of harsh is rare too. Normally, we keep encountering 1 to 1.5 SD below mean.
         Remember, we are not trying to predict anything with it. We are only being cautious. Now, after an alert patch coming, we look ahead for the variance going cooler. If it is still more than double of the house edge or worse, we do not enter betting yet. The moment it gets to a comfort range (that the better patch of at least last ten hits being near average/mean), we can start betting ahead.
         These ideas emanate from the concepts of "law of large numbers" and "regression towards mean" while keeping away from "gamblers' fallacy" and using failed oldschool illogical progressions that were deliberately propagated to benefit casinos.
         We may or may not use "pauses". If it is easy to use practically, pauses will always protect from long series of no hits in most sensible manner. BE it LLLLLLL or LLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLL, you are safe if you pause betting after a loss of three steps and resume only at either getting WLW or WW.

More later................


As, I said in the beginning, "regression towards mean" and "law of large numbers" both teach us that in the long run, things are more prone to be near average and in the short run, anything is equally possible. If we see our last 10 decisions in the terms of an EC, whether it is LLLLL LLLLL or WWWWW WWWWW or any mix of Ls and Ws in that stretch, probability of getting exactly that is as much as any mix of them like, LLWLW WLWLW. So, even something that looks as rare to have as one in a million event could have just happened. That is the power of randomness. Randomness brings us all sorts of patches of Wins and losses in an upredictable manner which fails all oldschool progressions like martingale, labouchere etc. In my early days, I was on a casino tour, I randomly picked a number to bet upon as my prime bet and ironically that number did not hit in 400+ spins that I played. When I calculated as to how many chips were sufficient to beat if a number comes in 401st spin. It reached virtually impossible levels and I know that virtual limit of a single number of roulette not hitting is far higher.
     I have analysed billions of spins via different bots and trackers provided to me by the best programmers on these forums. I literally found that instead of chasing a bet indefinately, pursuing it upto 1.5x of its breakeven point is sensible. Beyond it, it is suicidal. That culminated into my first research in extreme negative variance management and pausing betting at 1.5x of break even (for an EC bet it is LLL) and then wait for the tide to go.

to be continued later.....


I better quit this place than speaking to an uncontrolled mob who shamelessly assert on using martingale and always disrupt constructive debates with their gibberish.
 Mathematics without logic will never work in gambling otherwise every math teacher would have made killings in gambling. I asked this fellow TwoUp to not to speak further in this topic and if he want a fair debate over his out of world idea of a three step Marty dare to open a separate topic of his own.I just cited a randomly bad session, he started shouting cherry picking.I gave another session where his method will bleed, now his revaluations have gone.
  Where ignorance is bliss,it is folly to be wise.
 Thanks all who listened peacefully.


QuoteYes, I'm aware that "evening out" is in terms of ratios rather than absolute numbers, but more fundamentally, regression to the mean can't be exploited by waiting for prior losses because (as Albalaha has pointed out) the wheel has no memory. That seems to be a contradiction because at the same time he is saying that virtual losses do make a difference to his bottom line. Albalaha, can you clarify? 
I did not want to speak further after someone tried to disrupt the discussion entirely with as poor approach as a three step Marty but I would again warn against mixing the concepts of regression towards mean and gamblers' fallacy.
  We should not expect any corrective or clumping wins after a very bad stretch as it is nothing but fallacy.
 First of all, we try to make the worst of worst go virtual by my extreme negative variance management.Secondly,with applying pauses we erase all stretches of long runs of losses too.Thirdly, we need to apply a very smart MM that can handle the worst streaks and win thereafter in just average sessions thereafter.If we still get a very tricky stretch, we lose the least. A money management catering this need will work. If nothing works still, a stop loss will kick in.


Quote from: TwoUp on Jan 28, 2023, 01:37 PMThe "edge" here is not beating the casino math (which is alive and well), it is establishing sufficient distance between failure events, the power of compounding and a shrinking risk footprint.

So if reducing volatility leads to reliability, one way to do that is covering more numbers, but this doesn't really help because although gaps between losses are longer, the losses are proportionally greater when they occur. Or do you mean reducing volatility within a location? In that case I assume you mean with respect to wins & losses, not bankroll. This would therefore make it an issue of bet selection to reduce bunching of wins and losses so that they occur in closer conformity to the averages. If that could be done somehow it would mean that the house edge remained intact but your progressions wouldn't bust so often.

I assume this is what Albalaha is aiming at when he talks about extreme negative variance management and applying pauses and restarts.

I'm not convinced such strategies would work but I'm willing to give them a try. There's always something new to learn.


Quotereducing volatility leads to reliability

That is what is intended by extreme negative variance management. The term "extreme" itself denotes that it is virtually unbeatable with any sane approach. Martingale with unlimited table limits and infinite chips could beat any session but that is practically most imprudent thing to try. Sorry marty fans but marty makes you lose for sure, in the long run and one big win will wipe all the previous gains plus more. One oversmart fella came up with a three step marty as panacea for one of the sample skewed session I came up with. I made all those out of topic discussion in a separate topic from this, instead of deleting them. So marty lovers could continue there, without hesistation. Please avoid unrelated discussions here. Either do that in a separate topic or PM each other.

  Coming to the issue ahead, always remember that extremes are rare and can not re occur in succession due to the sequential probability it carries. Extreme of extreme could happen with a very small bet like a single number. For example, a single number reportedly hit in 678 spins and on one two reported occassions a number hit 7 times successively. So, smaller the bet, higher extreme it could have both positively and negatively. In my one topic in I pointed out that in zumma book of american roulette number 3 kept bad for 15k spins and its average hit rate was over 45 spins. On the contrary, when I analysed Zumma books of baccarat having over 116k hands the worst hit was 30/100 and 450/1000 if we played Player.

So, to manage extremes or to say avoiding extremes, always play either EC bets or atleast 1/3 bets like dozens/columns.


It is also imperative that we think of playing and winning in the long run to earn a net profit than trying to do some magic and get lucky instantly. Have you ever heard of a strategy wherein one tries to beat 450/1000 case? Probably Never. This irrational thinking made the game unbeatable. The moment you have a strategy which beats or even sustains 450/1000 situation gracefully with least harm that should be recovered in the next thousand trials of average hit rate, you have beaten extreme negative variance forever and have a grail of your own.
 Do you know that even if you have 100 chips bankroll, a casino expects you to lose all by approx 4000 trials? Mostly people buy chips, play for 3-4 shoes, lose all and come back empty handed. This happens to 90% of gamblers atleast. Get over this, be a pro player who plays to earn slow and sure or else quit gambling forever.


One more good way to handle extreme is to gradually carry forward losses till regression towards mean finally works, not by getting or expecting compensatory wins(which might happen too) but getting only average hit rate after the exreme bad run. If you are lucky enough to get compensatory wins too, it should help you more but even if it doesn't you should still not just recover your losses but win a good profit too.


This is not true - I have simulation software that pinpoints out all the 3, 4, and 5 STDs using real and true playing models for EC.
You can get back to back with a small or medium, correction that will not let you recover or make a significantly partial recovery.

And it doesn't take many POINTS for the regression to drop significantly.
That results in two to three or maybe four POINTS of regression.
Where at least one vanishes as you need some confirmation of a change towards regression.

You talk much and show little with significant working methods.
No staking plan and no identifiers for real true regression towards the mean or real examples.

This topic is a waste of time, in my opinion.



Don't waste your time on this topic, go ahead with your own revelation. Go and try martingale. That should help you.