I present the following argument that I found recently, for your debate and discussion. [Got that, did you? It's not my article]

Feel free to tear it apart if needs be.

   
• If there is any internal truth we need to find it.
• If it's a flawed argument we need to know that too.
• If the consensus is that it's valid and sound I'll reveal the source. Otherwise it can simply wither on the vine.

Over to you...




Many roulette players tend to succumb to the idea that their chance of winning will increase immensely if they're betting on more numbers.

However, this is a wrong move if you want to win in the long term.

Yes, your chances increase in the short-term when you're playing multiple numbers, but short-term winnings can just as easily and quickly be lost again if you continue to pursue that betting strategy using multiple numbers.

You need to concentrate on long-term winnings if you want to strike it rich.


I will prove to you below through mathematics why that is so:



Let's assume you play 37 spins and each number on the wheel lands just once during those 37 spins. You are betting $1 on number 0 throughout all spins.

When the ball lands on 0 you'll win $35 ($36 - $1 = $35). But you'll lose $1 for each of the other 36 numbers/spins.

That means that after those 37 spins you've ended up with a total net loss of $1 ($35 won - $36 lost = -1$).

Now let's play that example through again. This time you're betting on 2 numbers during each of those 37 spins.

Let's say you're betting $1 on number 0 and $1 on number 1. When the ball lands on 0 or 1 you'll win $34 ($36 - $2 = $34).

But but you'll lose the $2 total bet for all other 35 numbers.

At the end of those 37 spins you therefore have incurred a total net loss of $2 ($34 x 2 - $35 x 2 = -2$).

We're going to raise our stake now, shall we? Okay, now let's say you'll bet on 10 numbers during each spin of those 37 spins.

You place a $1 each on number 0, 1, 2, 3, 4, 5, 6, 7, 8 and 9.

When the ball lands on either of these numbers you'll win $26 ($36 - $10 = $26).

But you'll lose the $10 total bet for all other 27 numbers.

After 37 spins you've now suffered a total net loss of $10 ($26 x 10 - $10 x 27= -10$).



See how your losses mount the more numbers you play?

It's a mathematical certainty and the idea that the more numbers you bet on the higher your winnings will be is simply a misconception that doesn't work.

The unshakable fact is that you losses become exponentially higher the more numbers you play.

If you bet $1 on 20 numbers, you'll have lost $20 after 37 spins. There is nothing you can do about it.

It's been proven through thousands of records that the above is true in the long term and of course under the assumption that each number really lands on average once every 37 spins.

Betting on multiple numbers at the same time is a fundamentally wrong decision for long-term winning.


The perfect solution, however, is betting on only one single number and stick with it.

You are right. Cover the table is not the best. Every time you win on one chance you 100% lose on other. My testing with one straight up shows good result.
Everybody should know 2/3 bet is loser bet.


The odds of winning are the same according to pay out on all inside bets. Playing one number you do not "waste" any chips which can impossible win.


In short numbers of spins you may have better result playing more numbers, but if playing too many the winnings are more unsure.

...Interesting discussion dear Max, thanks for bringing it to the fore.

To me one number brings the opportunity for abrupt deviations from the mean, with wild swings.

It might be a blessing or a curse, depending on who you ask. But it does provide the "wildest" positive runs since the payout accompanies on the evening-out trams.

Remember roulette has "game duties", among which is to balance outcomes. Since one number is (physically) a hard target to hit, the game can miss a pocket during several cycles; but it must deliver as per its "duties" and then a single cycle MUST get more hits than expected when considering 37 individual spins in order to balance the ratios.

Here you "milk" it.

It's true, playing one number will maximize your chances of achieving a target, but because the variance is high you also run the risk of losing more than you would had you been playing more numbers. That's the beauty of roulette, you can choose according to your risk/reward preferences.

The deviation, while aginst your play, will in worse cases be impossible to match. that's more the reason than the HE a session will be lost.


I have played 1 number, had a 400+ sleeper and ended up with a win. The number in question got better hit rate later.


I used 5000 units to stand a long drawdown, got a long streak of winnings of total  13670 before the first bust, which happen yesterday.


I will not know then next will come, no way to say it will last until I reach 5000 units or about 45 sessions won.


I have used methods wich the inventor use all kinds of math and statistical research, and lost. I have used pure woodooo and won, it is a game of chance,
but still I think in the longer run you do better (lose less) using some methods.

Without a robot, who can wait out 400 losses for a win?  In a casino, no less?  Bathroom, anyone?

I'm not a math person, but this doesn't really strike me as being totally, 100% accurate.

Betting one number-------->on the way to a win---------->there can be no other wins.
Betting 24 numbers---------->lots of wins waiting for that loss he speaks of.

Not voting either way; I think all types of betting have their place.

Sam

A casino visit may be 400 spins a visit. A BM casino use to have higher min stake  than a casino on internet. The play and the way we do it must be different on line or at a casino there 400 spins an hour is possible. The whole nigth at a BM casino  must have methods which can possible end,  The play will differ.

The theoretical models are  not practical or even effective for a shorter time as a visit on a casino or an hour playing on line.

The variance is so broad, it is not possible to know very much before the play is over.

Winning 2/3 of  the spins is not the same we are winning. Winning 1 of 35 can also make a winning. The odds are the same on most of the bets. In some casinos in EU the outside bets has lower disadvantage.   Playing one number 37 times is about 2/3 chanse to hit once. But  this is not to be mixed of how long it takes when we actually try.

The strategy :

Select the hottest number.

In 300 more spins you need 9 wins to break even.
Hot numbers hit from 12 to 24 times per 300 spins.
A hot number can cool off at any time.
"Know when to hold them, know when to fold them."

One number is statistically the worst, to play. It can offer you the worst odds possible and the biggest variance too along with lowest expectancy to "appear and clump". Not a very practical idea in a real casino.

Quote from: albalaha on January 04, 2013, 06:47:26 AM
One number is statistically the worst, to play. It can offer you the worst odds possible and the biggest variance too along with lowest expectancy to "appear and clump". Not a very practical idea in a real casino.

In a BM casino it goes too slow to run a longer session on over 2000 spins. Otherwise I do not see a single should be harder to win playing. They cluster and sleep as well. The variance even out, but takes longer runs. The pay out  is high, so you need to win just a fraction of the spins.


I have done well using straight up, and most of the time it is one or a few.


I think I have done 10000 spins in two weeks playing one number only.

Ralph,
          that applies to all bets but waiting for a single number to clump can be as disastrous as #3 of zumma. May be playable with bot only.

Quote from: albalaha on January 05, 2013, 02:31:10 PM
Ralph,
          that applies to all bets but waiting for a single number to clump can be as disastrous as #3 of zumma. May be playable with bot only.

Easier with a bot, I have done 1000 of spins manual as well. 4000 in an evning manual is possible.


TCS had a run of 10000 spins and never come out using a split.


I have had a 2300 spins until 2 units plus, the funny was it was never down much.


A RFH you can get with any method.

Nobody seems to have disagreed with the mathematical arguments in the article.

Do we assume from that everyone agrees with the writer's figures? 

Or do we assume nobody read it right through?

I have found the flaw in the math.  When I get time and am not so sleepy, I'll write it up.

Thanks, Esoito....

Sam

Betting one number offers the best payout available in the game but as Bayes points out the variance is high compared to the low probability of winning. The examples are OK but the author assumes you're winning your very first bet and achieving maximum return every time, which clearly isn't possible and kind of undermines the whole argument. What if you hit on the tenth spin? You have previously already lost 9 units. What if the number doesn't hit at all? You're down 37 units.

To lower variance play more numbers, but you lose more and potentially more quickly. In the end, though, there is no actual difference.  The EV for a single zero game is the same for all bets. But that's not really the problem, the EV just really says you can't win flat betting by gaining a real edge. Unofficially then, you don't need a real edge to win. What if you could consistently guess at least 1 in every 4 even chances? You could win all day by doubling up on only a 25% hit rate. There's no maths that says you can't play a bet selection in a regular state of low variance. And that applies to all payouts.