Why roulette might be beatable and what the method should look like

[reddwarf]

Hi all,

I'm reddwarf. I like roulette, although I lost money overall. I also like to test and think about it, like so many of the forum members.

I posted this, because I'm very interested in what other fourm members think of my idea: Roulette might be beatable.

To explain why I think this, we first must think about what binds roulette and statistcal theory (maths as you call it):

I. Roulette and Statistics
Unless to what some people on this forum claim: statistics is proven, all phenomena in roulette can be described by statistics, just by assuming that the spins are independent and the numbers have equal probability of falling.

So the streaks, the "law of the third", the formation of patterns you name it: all can be described by statistics and yes, on the long run, when we play roulette the measured statistics more and more converges to the "theoretical" statistics.

According to statistical theory: roulette can not be beaten. Does this imply that the game of roulette is unbeatable? Yes and No!

more to follow (just wanna make sure that all I typed is not lost - again)

[reddwarf]

So why does statistics prove and does not prove that roulette is beatable?

Statistics is a field of science where, independend of underlying processes we can say something meaningfull about the outcome of the processes, by using a few simple and testable assumptions (independant probabilities etc).

For example: based on this assumption, we can show that after 24 spins, the probability of more than 1 repeat is extremely close to 1 (hence "law of the third" - it does not equal 1, so it is not a true law!). Or 31 blacks in a row must occur now and than! and 5 times a repeat must occur now and than also!

This basically means that if we use a mode of play what I call random play, the law of large numbers will govern our result. So what is random play?

Random play examples:
1. waiting for an event that will make you win
2. trying to predict what comes next
3. using a systems that uses: "what goes up must come down"

in other words: all systems based on gamblers fallacy are actually random play!

Now, lets assume that there is 1,just 1 method that is not random play. Can this beat roulette? All we can say is that it is undecided at best!

Sounds strange? Think about this example:

I'm an alien from Mars, and I'm really interested in the particals that use a wormshaped vehicle to move (=humans commuters in a train).There are so many, that I'm not going to count each and every particle. After studying the behaviour of these so called humans that start at A en travel to B or C for 1 year, I came up with the following statistical law:

1. of all humans that enter the wormshaped vehicle, 99.00000% leave the vehicle at B
2. the remaining humans, 1.00000% leave at C.
3. at C the wormshaped vehiclereturns to A again

The alien is also an avid gambler and a clever one also. On any give day he projects numbers on the heads of the humans, and his fellow aliens can bet on what human leaves where, or what human stays on the train. How would you bet? Oh yes, it is a Dutch bet: you will always lose in the long run.

Any thoughts?

Think about this: the train returns to A again
The numbers are based on a zillion observations.....

Bet on the driver to stay on the train!!!

Maybe a strange example, but it illustrates that statistics does describe the process, but does not equal the process (in this case)

to be continued

[reddwarf]

Hmm that's a good question:

I think that most people wil say yes because 99.9999% of al methods, systems and strategies in every forum are random play. Yes sometimes you win,but averaged out over all players and time,it's a loser. And no, when you are a winner for a year: it has nothing to do with special gifts, it's just plain luck (can also be calculated how many people emerge as a winner after 1 year using a specific system!).

We can put it in a different light altogether and see is the answer is undecided or not:
Lets assume that a roulette session is either won or lost. What makes up a session has te be seen yet, but for now lets assume that we know when a roulette sessions starts or ends.

We can now write down all spin combinations that might occur in that session. Of course there are a tremendous amount of combinations (depending on the session length).

The claim is actually that when we play random, there is always at least 1 combination that will cause us to lose.

But can we prove that there is No or more than 1 method that will cause us always to win? My claim is: this is undecided at best. (simply because I never found a proof that such a system does not existin math literature)

to be continued

[reddwarf]

2. What must a winning system look like?


Before we can answer this, we have to understand what makes any system method or strategy loose. As mentioned before: every random play loses in the long run.

The reason why random play loses is that not all permutations in a session are covered. An example to illustrate:

say we have a session of 36 spins. we bet on last EC to repeat, using a martingale, we can now show that overall we will lose because the probability of no repeaters is larger than 0. So 1111111111111 etc or 22222222222222 are permutations that will cause this method to lose in the long run

By the way: this is also a nice method to debug any system: just find a permutation that will cause a loss, can you find one? and RNG or roulette ball will definatly find it example: the Dz/Cl method of black pearl: a sleeping dozen or column (please note the probability of a dozen or column sleeping for more than 9 spins is ~ 2.6% It will happen say every 40 sessions!!)

method feature 1:
the method must be such that all permutations are covered OR the impact of permatuations that will cause a loss are minimized. Because roulette is a Dutch bet, we are not able to cover all permutations. So we must minimize the impact of "negative"permutations.

This minimization must follow the rules of a winning system: we must be 100% that the loss due to a negative permutation is less than the loss without covering for the negative permutation

By the way: such bets DO exist SOMETIMES  

to be continued

[reddwarf]

Apart from feature 1:
feature 1b: sometimes all fields will be covered
feature 2: inside bets MUST be used at some point in the sessions
feature 3: the decision to start a session can be equal to the decision to end a session (independant spins), so no waiting is required


To illustrate what I mean: lets create an example roulette game and try to meet all above requirements. please note this is just an example,not a winning strategy!! It just illustrates how we can use the features to build a method + some bonus info on which I will never elaborate!!

Session: we start at will, we stop when 3 unique dozens have appeared
negative permutations: repeats of previous dozens!, zero's or a combination

Lets define the betting unit U as 4*b, b=base bet, 1 dollar, 10 cents or whatever. So in our system 1 unit = 1U =4*basebet

bet1:
if previous dozen=1 or 3: bet 0.5 units on dozen, bet 0.25 on line that completess the half, bet 0.25 on the zero and bet 1.5 units on the oppisite half
if previous dozen=2: bet 0.25 on line, bet 0.5 units on dozen that completess the half, bet 0.25 on the zero and bet 1.5 units on the oppisite half

what can happen:

1. opposite half (another unique dozen): we win, technically the session is not ended, but we end it anyway
2. zero, we win, technically session is not ended, but we end it anyway
3. covered dozen -> we win 0.75 units on the negative permutation, so overall we lose 0.75 units (instead of the 1.5 units we would have lost without covering the negative permutations

spin2:
if zero falls, or the previous dozen repeats, we can repeat the previous bets
if the line falls we cover the previous 2 dozens + the zero (with 0.25 units) + the half that contains the dozen not yet fallen

to be continued

[reddwarf]

This example illustrates several things. One of them is how to create a system that might be ableto beat roulette.of course, this is not the one. This can be easily seen: it is still random play because not all negative permutations are minimized: A long streak of dozens will kill this system. And yes, it will happen.

reddwarf

[reddwarf]

So to finalize:

1. Statistics does not say that the game of roulette is unbeatable: it only proves that it is unbeatable when using random play
2. if a winning system exists than: we have to define what a session is, and we have to create a step-by-step approach to avoid random play.

Please note we did not say that all steps shouldn't be random play: ate least some of them during the sessions shouldn't be.

Is this useful? I hope so

grts reddwarf

[esoito]

My word -- you have worked hard to produce this very interesting and thought-provoking thread.

I'm about to read it all again -- slowly and carefully this time!

Meanwhile let me throw this thought into the pot:

Roulette is beatable in very short term play if you go in, and then come out as soon as in profit. (Yes, along the lines of hit'n'run.)

This avoids the casino's long term edge stripping you bare.  (After all, why else do they ply punters with food'n'drink?...To make 'em stay there until their edge cuts in, of course! It's a statistical certainty that keeps the casinos afloat...)

[reddwarf]

Hi Esoito,

I agree that a hit and run approach, combinedwith a systems that minimizes random play might be a long term winner. You stop the session when you are in the plus.

The second requirement is important: let's assume that I bet on the color to repeat, and I use a marti, or a labby. When I'm in the plus, i quit. The next day I use the same or another method. Still I claim that in the long run I will loose! The only thing is, It will take longer ( in days) before I notice that I loose.

reddwarf

[ScoobyDoo]

So will this thread remain theoretical or will there be some solid tactics to produce the method that you mentioned above?

Scooby Doo

[reddwarf]

Give me some time, I will come up with some suggestions

but I'm kind of busy building a starting BR ;-)

reddwarf

[Johnlegend]

Hi all,

I'm reddwarf. I like roulette, although I lost money overall. I also like to test and think about it, like so many of the forum members.

I posted this, because I'm very interested in what other fourm members think of my idea: Roulette might be beatable.

To explain why I think this, we first must think about what binds roulette and statistcal theory (maths as you call it):

I. Roulette and Statistics
Unless to what some people on this forum claim: statistics is proven, all phenomena in roulette can be described by statistics, just by assuming that the spins are independent and the numbers have equal probability of falling.

So the streaks, the "law of the third", the formation of patterns you name it: all can be described by statistics and yes, on the long run, when we play roulette the measured statistics more and more converges to the "theoretical" statistics.

According to statistical theory: roulette can not be beaten. Does this imply that the game of roulette is unbeatable? Yes and No!

more to follow (just wanna make sure that all I typed is not lost - again)

Roulette is beatable REDWARF, theres no maybe about it
                 

[MauiSunset]

Hi Esoito,

I agree that a hit and run approach, combinedwith a systems that minimizes random play might be a long term winner. You stop the session when you are in the plus.

The second requirement is important: let's assume that I bet on the color to repeat, and I use a marti, or a labby. When I'm in the plus, I quit. The next day I use the same or another method. Still I claim that in the long run I will lose! The only thing is, It will take longer ( in days) before I notice that I lose.

reddwarf

I like your statistics approach but how can you say that a "hit and run" system is better than a "Hit and stay" system?

Martingale must NOT be used with your money - you must use casino money on this very very risky way to prove that you are correct when the wheel keeps telling you that you are a loser.

I've seen this fascination with hit and run before and I can't find any statistics that back it up.  If you want to put a table maximum on a single number that's perfectly ok as long you you realize it's a totally insane way to risk your BR.

Eventually ALL winnings will be wiped out with a "hit and run"; statistics doesn't operate differently on "hit and run" verses longer term betting.

[esoito]

"I've seen this fascination with hit and run before and I can't find any statistics that back it up. "

Would you deny that the longer you play the greater the chance of the casino's statistical edge (goodness me -- there's that S-word again) draining your capital?

After all, the edge is their sole raison d'etre...they certainly don't exist solely for the players' benefit!

[MauiSunset]

"I've seen this fascination with hit and run before and I can't find any statistics that back it up. "

Would you deny that the longer you play the greater the chance of the casino's statistical edge (goodness me -- there's that S-word again) draining your capital?

After all, the edge is their sole raison d'etre...they certainly don't exist solely for the players' benefit!

The odds of Green 0 showing up is 1/37 in each spin.

Doesn't matter if it's 2 spins or 100 spins the odds are the same.

The odds of ANY Roulette bet are well known and work on 2 spins or 1,000.

The idea of "hit and run" is just as bogus as Gambler's Fallacy - it sounds good but no math can back up the claim...