Can roulette be beat by volatility extraction?

I need to be clear here…

I do not mean in an EV sense, I do not expect the wheel to change. EV is based on total wagered…kinda a shit metric and stupid premise in general. More important for the casino’s bottom line than mine.

But, after tons of monte carlos, 100k round sims etc

it does appear that certain strategies are stable enough to bounce up and down enough times

that you can extract more than your initial starting balance on average.

Roulette can’t be beaten in a static odds sense, but I am quite confident various bet types and progressions can create enough fluctuations that they total FAR GREATER than your starting balance.

and any gambler probably knows this or has witnessed this sort of thing. You have $100 and play and get the notification on ur phone “you have wagered $603 in the past 30min”

meanwhile ur balance is $45 lol

lets use $200 bankroll, 3 streets as an example



$1 a pop. Double every 4 losses, reset on win.

• 4 losses → $12
• 8 losses → $36
• 12 losses → $84
• 16 losses → $180
• 17th loss → +$24 = $204

You bust on the 17th straight loss.

Bust Probability

q^{17} = 0.7632^{17}

≈ 0.98%

you have a 99+% chance at profiting any given cycle… as it gets deeper the profit increases.

avg cycle profit ~ $15

regardless… on average how many cycles could you do this til bust?

well… the MEDIAN is 78, the mean is ~ 113

so lets super lowball here…

if you withdraw $15 each successfully completed cycle

and survive 30 cycles

is that more than $200..??

sounds closer to $450

and so lets look at that, what’s the chance of hitting just 30 cycles before bust?

is exceptionally below the avg expectation but It makes an important point.

~ 76%

TLDR you can more than double your money at a rate of 70+%

but the vacuum math nerds (likely casino hired) look at everything per spin and talk static odds instead of probability windows.

they lie to your face claiming “the game can’t be beat”

actually it definitely can

what you ACTUALLY mean is…

…you cannot alter the static odds per spin.

yeah uhh.. everyone already knew that

shocker!

even if my math isn’t perfect, I know for a fact this along with many other strategies, create enough fluctuations in the bankroll… that you can extract 2x your starting balance

more times than not

P({survive } { cycles}) = q^n
• Survive 10 cycles: 0.9912^{10} \approx 0.914 → 91.4%
• Survive 20 cycles: 0.9912^{20} \approx 0.835 → 83.5%
• Survive 30 cycles: 0.9912^{30} \approx 0.768 → 76.8%
• Survive 50 cycles: 0.9912^{50} \approx 0.642 → 64.2%
• Survive 100 cycles: 0.9912^{100} \approx 0.412 → 41.2%
• Survive 200 cycles: 0.9912^{200} \approx 0.170 → 17.0%

even just 20 cycles gets u over $200 profit.

please someone tell me im wrong.

@wizard hook me up on that “nobody can beat roulette” money 😂

actually im willing to believe it was a farce proposition and yall expect people to beat the wheel itself which everyone knows isnt possible

Quote: cheddachaser

$1 a pop. Double every 4 losses, reset on win.
link to original post

Thread moved to Betting Systems.

Quote:

please someone tell me im wrong.

All betting systems are equally worthless.

Quote: Dieter

...
All betting systems are equally worthless.
link to original post

If no progression can work at all, neither can card counting.

Prove me wrong!

Quote: AutomaticMonkey

Quote: Dieter

...
All betting systems are equally worthless.
link to original post

If no progression can work at all, neither can card counting.

Prove me wrong!
link to original post

Monkey,
The wheel gets shuffled every hand.

Quote: Dieter

Quote: AutomaticMonkey

Quote: Dieter

...
All betting systems are equally worthless.
link to original post

If no progression can work at all, neither can card counting.

Prove me wrong!
link to original post

Monkey,
The wheel gets shuffled every hand.
link to original post

Aw shucks, I can't come up with some pedantic possibility where a progression might give you something on a wheel too.

I'm at a loss for words!

$1 a pop. Double every 4 losses, reset on win.

• 4 losses → $12
• 8 losses → $36
• 12 losses → $84
• 16 losses → $180
• 17th loss → +$24 = $204

You bust on the 17th straight loss.
$1 per street, 3 streets. Streets pay 11:1 or 12 for 1.

So you bet $3 four times and wind up at -$12. A win would put you back at even or above.
2nd set, you bet $6 four times and wind up at -$36. A win on the last of those four spins would get you back $24, you'll still be -$12 short. You could just start over those 4 spins at 3X $2 again.
3rd set, you bet $12 four times and wind up at -$84. A win on the last spin could put you back up $48, you'll still be $36 short. You could just start over with those 4 spins at 3X $4 again.
4th set, you bet $24 four times and wind up at -$180. A win on the last spin could put you back up $96, you'll still be $84 short. You could just start over with those 4 spins at 3X $8 again.
5th set, you bet $24 one time and wind up at -$204. A win on the last spin could put you back up $96, you'll still be $108 short. You could just start over with the last 4 spins at 4X $8 again.

For a $3 roulette machine, you need to multiply all values by $3. For a $5 machine, multiply all values by $5. For a $10 table, multiply all values by $10. For a $15 table, multiply all values by $15.
$3 machine buy-in = $612
$5 machine buy-in = $1,020
$10 table buy-in = $2,040
$15 table buy-in = $3,060
$20 table buy-in = $4,080

If you were starting at $5 machines and doubling your bets every time you exceeded the next doubled buy-in, you'd need 13.6 wins ahead to double up your buy-in. (16X $75 = $1,200) (16X $150 = $2,400), (16X $300 = $4,800; and winning $180 extra + $360 extra + $720 extra equals a $1,260 extra for a restart or a session up. So at 48 cycles, you'd be at around a 2 in 3 chance of making it if you reset the way your numbers did. If you double up your bets on 16 wins, you could cash out $1.2K + $2.4K+ $4.8K = $8.4K + $1K buy-in = $9.4K and still be under a CTR.

I'd have to see more about how your actual betting turns out when you reset your bets on a win. It's vague to me as it relates to profits. But be sure, the automatic plunger on these roulette machines can evade your betting choices to the extreme. Get to a real table after you graduate to the table minimums.

I personally prefer to bet two dozens bets and climb up a half spin bet per win.

It's important to find a high limit roulette wheel with only one zero. With your street bets counting as inside bets, a $100 table minimum could have 3X $35 or $40 bets to start. With $40 street bets, you'd need $8,160 buy-ins. You could double up your street bets on 16 wins to $80 bets, and $160 bets and still be under the $500 table maximum.
You could win (16X $600 = $9.6K, 16X $1.2K = $19.2K, $16X $2.4K =$38.4K) or a total of $67.2K for 48 wins.

But where I live, the double 0 wheel has a limit of $2,500 per spin. I could eventually get my street bets to $200, then $400, then $800 on a $40.8K buy-in and cash out at $376,800.
Really, all it takes is 3 winning sessions in a row to get from $1K to $400K+. After that, you'll have 400 buy-ins to try it all over again.

Quote: Dieter

All betting systems are equally worthless.
link to original post

Untrue. There's a long history of successful betting systems in various forms of sports betting which yielded a significant positive expectation. You need to add a qualifier "in a game of independent trials".

For roulette specifically the answer is not technically accurate either. On a wheel with any physical bias creating serial correlation, progression systems will have a second-order effect on expected value — positive for anti-martingale-type systems, negative for martingale-type systems — proportional to the degree of bias. Most wheels do in fact have some kind of bias though not generally < p=1/34. Games like blackjack and baccarat are similar: there are second-order effects due to card removal which have a very minor effect on expectation.

As a practical matter at roulette though you would be looking to exploit the source of any bias directly rather than use a progression.

Quote: Dieter

Quote: cheddachaser

$1 a pop. Double every 4 losses, reset on win.
link to original post

Thread moved to Betting Systems.

Quote:

please someone tell me im wrong.

All betting systems are equally worthless.
link to original post

Some betting systems are more worthless than others, which means some systems are better than others. They may all be worthless, but some are less worthless than others

Here is an example- After two losses, you cut your bet in half until you win. It does nothing to help you win money, but it helps you to lose less.

Quote: cheddachaser

please someone tell me im wrong.
link to original post

In the words of Otto from The Simpsons: Challenge Accepted!

Quote: cheddachaser

but the vacuum math nerds (likely casino hired) look at everything per spin and talk static odds instead of probability windows.

I have never worked for a casino.

Quote: cheddachaser

it does appear that certain strategies are stable enough to bounce up and down enough times that you can extract more than your initial starting balance on average.

Key word there.

Quote: cheddachaser

if you withdraw $15 each successfully completed cycle

There's your first problem. Losing your first 4 bets happens (29/38)^4 (on a double-zero wheel), or about 1/3, of the time.
2/3 of the time, you will win - but your profit in this cycle is only $9 if you win on the first spin, $6 on the second, $3 on the third, and zero on the fourth.
Kind of hard to withdraw $15.

Am I reading your strategy wrong?
I am reading it as: you are starting with a $1 bet on each of three streets (e.g. 1-3, 4-6, and 7-9).
Bet 1 per street until you win, and start a new cycle, or you lose 4 in a row, in which case you double the bet to 2 per street.
Suppose you lose your first 11 bets; that's 4 losses of 3, then 4 losses of 6, then 3 losses of 12, for a total loss of 72.
You are betting 4 per street on your 12th bet; if it wins, you profit 36 on the bet, but you are still behind 36 for this cycle.

The "gambler's fallacy" is that something will "eventually" happen.
You might as well throw in the Law of Large Numbers, which says that the results will "eventually" balance out to reach the expected probability.
While this is true, it applies only if you realize that "eventually" can be a very long time from now. I have done simulations with fair coin flips, and while the heads/tails numbers always came back to matching, there were times when it took trillions of flips to get there.

Time to break out my Surefire Guaranteed Cannot Miss System from my Ten Rules of Gambling, Three Of Which Are "Don't Play 6-5 Blackjack":
1. Determine the smallest number such that a particular bet cannot win that many times in a row (e.g. "Red cannot come up in roulette 25 times in a row").
2. Wait until that bet wins (that many times minus 1) times in a row.
3. Bet your house, your car, and your things that it loses.
4. Did it win again? Don't look at me; you're the one that claimed in Step 1 that this was impossible.

not claiming to beat the wheel, just making the assertion that

you can 2x ur money more than 50% of the time, with a simple delayed step prog

which isnt EV+, i dont make that claim nor do i care to… this actually all works within the bounds of a negative expectation game which is honestly pretty cool and appears profitable.

perhaps the whole EV debate needs reframing.

if i can double my money more often than not, I could care less if the bet is technically skewed against me and considered EV-

my thing is more this.

can we confirm that it doubles money at these rates?

because then in theory your freak tail event and oh no i bust $100

is quite negligible

you could find yourself in a place where losing 20 $100 buy-ins, in a row.

is negligible

the only real issue ive found is that when u ramp up the volume there is a point of diminishing returns and the -2.7% starts to eat everything.

could maybe be fixed via scaling up or down. am unsure.

Finally someone who understands. It’s the whole independent trial issue.

I think people who try to change the odds of the spin or deal itself are kinda nutjobs chasing some slight advantage

that appears to be less useful than even a simple progression paired with a probability window.

Ive known some BJ counters, just because he had a fractional edge didnt mean he won.

and if ya told any AP, “hey ive got something that doubles your money even 55% of the time”

thats better than whatever bullshit they were doing before.

also the delayed step thing is just risk mitigation

u can choose the rate required for profit each hit or opt for more safety if willing to hit a potential dead zone… not really my point.

the delayed step part is just to ensure the bet size doesnt spiral out of control

you can adjust for your own risk tolerance

Quote: billryan

Quote: Dieter

All betting systems are equally worthless.
link to original post

Some betting systems are more worthless than others, which means some systems are better than others. They may all be worthless, but some are less worthless than others

Here is an example- After two losses, you cut your bet in half until you win. It does nothing to help you win money, but it helps you to lose less.
link to original post

TANSTAAFL. Betting less than table minimum can take forever to lose it all, and they may discontinue inducements.

You can ask for Tito's & soda, but don't be surprised when you get a glass of Sprite with a drip of Popov in the straw.

InInteresting thread. I've been working on a related problem —
minimizing loss frequency rather than chasing edge.

My system hits 99.22% win rate per cycle with a fixed capped budget,
verified on 100,000+ simulations.

Full breakdown with math here:

Happy to discuss the math openly.
— R. Sitdykov

Quote: AutomaticMonkey

Quote: Dieter

...
All betting systems are equally worthless.
link to original post

If no progression can work at all, neither can card counting.

Prove me wrong!
link to original post

Progressions work great if you have a consistently winning method but then you don't need a progression if that's what you have. Progressions are only used if you lose more than you win, if you win more than you lose why would you need it. I never use a progression it's pointless and if you're a consistent winner it draws extreme heat from the casino.

Quote: cheddachaser

not claiming to beat the wheel, just making the assertion that

you can 2x ur money more than 50% of the time, with a simple delayed step prog

which isnt EV+, i dont make that claim nor do i care to… this actually all works within the bounds of a negative expectation game which is honestly pretty cool and appears profitable.

perhaps the whole EV debate needs reframing.

if i can double my money more often than not, I could care less if the bet is technically skewed against me and considered EV-
link to original post

First of all, if you can double your money before losing all of it more than 50% of the time, then that certainly is positive EV, as you are, in effect, betting your entire bankroll on a single even-money bet.

Second, as I explained in my earlier post, either you are describing the system wrong, or one of your key assertions as to how much money you make on a particular pass is wrong.
Can you please answer these:
You lose seven consecutive times, and win on the eighth. How much did you lose in those first seven bets, and how much did you gain on the eighth?
If you are betting three streets at a time and betting 1, 1, 1, 1, 2, 2, 2, 2 (per street) on each bet, you lost a total of 30 on your first seven bets, and on the eighth, the street that won made a profit of 22 (bet of 2 paying 11-1), but you lost 4 on the two streets that did not win, so you only profited 18 on that spin, which means your total for that round is a loss of 12. Is that correct?

Quote: cheddachaser

not claiming to beat the wheel, just making the assertion that

you can 2x ur money more than 50% of the time, with a simple delayed step prog

which isnt EV+, i dont make that claim nor do i care to… this actually all works within the bounds of a negative expectation game which is honestly pretty cool and appears profitable.

perhaps the whole EV debate needs reframing.

if i can double my money more often than not, I could care less if the bet is technically skewed against me and considered EV-
link to original post

"perhaps the whole EV debate needs reframing."

Math is a "debate"? LOL

If you make one big bet, you have a better chance of doubling your money than by using a series of small bets.
Bet it all on the Big in Sic Bo, and you'll double your money almost 49% of the time. That doesn't make it a good bet.

Quote: billryan

If you make one big bet, you have a better chance of doubling your money than by using a series of small bets.
Bet it all on the Big in Sic Bo, and you'll double your money almost 49% of the time. That doesn't make it a good bet.
link to original post

Given the lack of modern social mobility I'm beginning to wonder if that is true any more.

Someone setting up a business now has average terrible odds of making it big. Maybe the casino is better. I'm exaggerating...but not by that much.

Quote: cheddachaser

I need to be clear here…

please someone tell me im wrong.
link to original post

You're wrong.