Variance control vs. Edge from flat-betting

[falkor2k15]

Here's an interesting thought for what it's worth (or what little time we have left): If we can control variance then it follows that we still lose from flat-betting and without gaining any edge (and often we lose from progressions too); here's an example:
HLHLHLHLHLHLHLHLHLHL

Above we lose if we just play H or just play L. Same with the following: RRBBRRBBRRBBRRBB

So in order to gain edge we don't flat bet - we wait for the average number of losses to hit first before betting when we expect a win - above would be one loss for HL and 2 losses when betting for B.

Variance control also indicates that progressions would eventually start to win as opposed to breaking the bank - but that it's still 2nd place to waiting for average losses.

So let's say we wait for an average of 2 Reds before betting black but we get a 3rd Red:
RRR...

If there was this bias in the variance - as opposed to the equality of Red vs. Black - then we would stop betting and begin waiting for a virtual win before re-attempting.

So how might we gain control over variance? Maybe track dozen cycles + line cycles
D1+L1... bet 2+3
If win:
D1+L1, D2+L3... bet D3+L1+L3.

First find out how often that stitched cycle wins vs. loses and then only bet when you expect a win - otherwise you have to wait for the next sequence - always based around when both Dozen cycles and Line cycles are on spin 1 and in sync, etc.

[SWEET]

Hi Falkor,
With due respect.

Its true that, variance kill whatever progression and our little br.

The worst ever recorded, EC bet,
"only69red/200", that could only won with labby, star from(1), won at 67hit, need progression that snowballed to hundreds units, and just impossible to play.

Even only30hit/100, also hard to play.

There never recorded yet, any hit less than 30/100 in any bm casino, thus if we see streak of 10 losses, then there will be at least 30hit in next 90spins.
(if you not believe this, research any real data, after 10 or more losses, then there be at least 30hit, before the 100th spin.)
Thus, only 30hit/100 spins, could applied, as highly unlikely, or near impossible, to hit in next 100 any EC roulette/baccarat.

Now, if next 100, has only 30hit, how are we going to turn the negative variance, into hit within math expectation?, and also, not to turn next 100 spin, that has within math expectation, into negative variance!!!?
(I mean, in there, say 45hit in 100spin, our method should not make it into losing negative variance!)

Some may say, that "follow the last or 2nd-last", can turn negative variance into advantage, but it also proven that the ftl and ft2ndl, also hit by long 30/100 negative variance too!!!

[SWEET]

So, your main purpose now, is to look, for a bet selection, that ALWAYS hit WITHIN MATH EXPECTATION!
(What is , math expectation?
eg.
in next 100spins, there may, from45 to 55hit, or any EV 27%,35/37, ecart, or any #@+-#, that
the math boys must nodding head agreeing to ..)
If you can have,
a math selection,
that can produce,
hit within math expectation, 40+%
when next 100 has worst variance, 30/100,
or within math expectation "50/50minus edge", then any $#@& progression, can win.

[SWEET]

Some math professors, here, screaming their lung out that, we cannot win, because -EV2.7, edge,
bla,bla,bla, with any progression.
ok, now  ,if the 46hit/100.
that more than -2.7,
it easy to win with any simple progression.

If they still denying, then,
let play a game that has 50/50, no EDGE, 1:1payout.
1red marble, and 1 black marble in a box, I take 1 out, and put it back.
(follow casino rules.)
we may still expect to lose, in many seasons, why?
NEGATIVE VARIANCE.

[SWEET]

until you found a bet selection, that produce, stable result, always hit, within math expectation...
you should not bet in casino, because negative variance, will hit you very hard, bleed you dry...
you are not rushing to LOSE MONEY!

[falkor2k15]

SWEET, what you are describing is when the variance is unbiased (by default) and progressions will break the bank (by default).

But if the variance is *biased* then it follows - based on my extreme examples - that we can gain an edge through virtual wins/losses and that even progressions would start to win within the table limits, i.e. the gap between wins or losses will start to become smaller as variance is manipulated further.

Therefore, if true, then it means all the advice we were told about a system must win flat-betting and progressions being a no-no was not only incorrect but also a false dichotomy because virtual wins/losses were never considered due to equally-like outcomes when Non-Random can make outcomes unequal or temporarily locked out; independence vs. dependence never even comes into it, hypothetically speaking.

Of course there could be a fallacy to this whole line of reasoning that I'm missing. It's not my usual way of thinking.

[SWEET]

(English not my first language, I struggle to understand, even need to look up for meaning of "dichotomy". hahaha)
I think we should ask...
WHY WE LOSE?!!
why we lose!!?
why we lose!!?
the answer simple...
The flatbet or progression cannot "closed", because, the winning hit, keep not coming, till our br bust!
Thus, our flatbet, or progression, must have a betselection that produce enough hit, to close within our br capital, or cutloss target.

[SWEET]

So, if ,
(oversimplified example)
next 100 has only 30hit, (there never less than 30/100, as virtual limit of variance),
then LABBY cant closed, because labby neefs 33.34% to closed, then to avoid possible 30hit/100 negative variance that bust your br.
Wait for, say 10losses in row, (or long losses with only few hits), then,
next 90 has atleast 30hit, thus 33.33%..that labby has chance to close before 100th spin.

[ati]

why we lose!!?
the answer simple...

It is simple, but maybe not why you think.
We lose, because each spin and therefore each result is independent. Every individual bet has negative expected value, so the law of large numbers guarantees that you will lose in the long run.
You cannot look at the imbalance in variance in the past 20, 50 or 100 spins and bet against it, because all results will be independent. If you truly understand this, then you can also understand why a progression is not a long term solution.

It is also important to understand that mixing the spins from 10 different wheels, or ignoring every other spin, or mixing up the numbers on the table in any way, will not change anything in randomness.

Some people claim that this mathematical fact can be broken, but that requires a radically different way of thinking 

[Joe]

I started a google blog on this topic a little while ago. I aim to find out whether any bet selections make any difference to variance, even if they don't affect the long term edge.

link:s://roulettesystemanalysis.blogspot.com/

I tend to agree with ati that no bet selection makes a difference, although many players swear it does.

[Joe]

Here's an interesting thought for what it's worth (or what little time we have left)

Remind me when doomsday is again? 

Perhaps you should add it to this list -

link:s://en.wikipedia.org/wiki/List_of_dates_predicted_for_apocalyptic_events

[jay]

That’s happening soon in a theatre near you, when they release 5G.
We are all going to die.
Joe, ......... again mate

[SWEET]

think like this...
we cant predict future.

Now, if we bet only RED, for next 100spins, what will happen?
1.RED will hit around 50%.

2.RED, will hit, at worst...only 30hit.

3.RED, may, at best hit, 70hit.

THAT have nothing to do with edge, in short 100spins, but probability of random, at best, and at worst RECORD.
Go and look at your real bm, any EC data.
You cant find , 100spins, that has less than 30 hit.
seek, before you disbelieve...

[winforus]

think like this...
we cant predict future.

Now, if we bet only RED, for next 100spins, what will happen?
1.RED will hit around 50%.

2.RED, will hit, at worst...only 30hit.

3.RED, may, at best hit, 70hit.

THAT have nothing to do with edge, in short 100spins, but probability of random, at best, and at worst RECORD.
Go and look at your real bm, any EC data.
You cant find , 100spins, that has less than 30 hit.
seek, before you disbelieve...

Then I have this proposal for you: We will flip a coin 100 times, and for every tail, I will get $10.50 and for every tail you will get $10. Deal?

[jay]

Sweet
It doesn’t really stop when you have a 30/70 ratio. You might think that it will even itself out in the next 100 spins.
The real problem is, sometimes the trend extends to a much larger number of spins like 150/400.
Extreme variance.
One side can’t catch up, it takes 1000s of spins.