Does martingale really works?

[steven.G999]

i think martingaling leads to bust eventually. the house edge will eventually kick in

yes, even if you play with a capital 1-10 BTC, martingale still will not run  

their is no profit in martingale even at 10 btc .. anyone know what is main reason why martingale note run

I do not really know why martingale not always run smoothly, now I better play single bet, I'm really comfortable with it. and not immediately lose all my balance in a heartbeat (like martingale)

Right play with the method of martingale it is better to use a single play than by playing auto play because with the single mode we can change the roll from high or low.

so at least I did not lose in an instant with a single play.  

I also enough extreme to play a single bet using a multiplier x5 - x9

[djibrilantiuang]

Martingale in dice game is about balance endurance, if u have'nt u will burned down.
just dont do martingale with low or a little balance

[jmurjeff]

Martingale does not work because the games have betting limits. If there were no betting limits and you had a big enough bankroll, it is a surefire way to win.

[sed]

Maybe  the best way to put it is this: martingale works when it works and it fails when it fails.  It's all random afterall (unless you're being cheated!).  That means that sometimes martingale is gonna pay well and sometimes it won't.

[dooglus]

Here's an absurd example.

Roulette, betting 100 on Red and 50 on red and 50 on black have the same house edge, but one bet is obviously a lot worse.

So maybe the important point is that while we can't change the house edge by adjusting bet size, we CAN change the variance, and as gamblers we want the variance to be as high as possible.

To make your roulette example just a little more stupid, imagine betting $1 on all 37 numbers. Every spin of the wheel costs you $37 and pays you $36, so you lose $1, or 1/37th of your stake every spin.

Alternatively, bet the the whole $37 on zero. 36 times out of 37 you lose $37. 1 time out of 37 you win $37*35, for a net expected profit of:
(37*35 - 36*37)/37 = (35-36)*37/37 = -$1, or 1/37th of your stake every spin.

So the risked amount is the same, the expected profit is -$1 each spin, but in the first case the variance is zero, and in the 2nd case it's higher.

What roulette bet has the highest variance? Is that the one that's "best" to play?

[TriggerX]

I strongly advise you NOT to martingale or only martingale for a few rolls only. I remember I martingaled on 66% at primdice with a balance of 0.4 and within the first few minutes I lost it all due to martingaling. Stay away from dice sites and only bet what you can afford to lose.

[dooglus]

I strongly advise you NOT to martingale or only martingale for a few rolls only. I remember I martingaled on 66% at primdice with a balance of 0.4 and within the first few minutes I lost it all due to martingaling. Stay away from dice sites and only bet what you can afford to lose.

I had bad luck with it too:

[Phildo]

Here's an absurd example.

Roulette, betting 100 on Red and 50 on red and 50 on black have the same house edge, but one bet is obviously a lot worse.

So maybe the important point is that while we can't change the house edge by adjusting bet size, we CAN change the variance, and as gamblers we want the variance to be as high as possible.

To make your roulette example just a little more stupid, imagine betting $1 on all 37 numbers. Every spin of the wheel costs you $37 and pays you $36, so you lose $1, or 1/37th of your stake every spin.

Alternatively, bet the the whole $37 on zero. 36 times out of 37 you lose $37. 1 time out of 37 you win $37*35, for a net expected profit of:
(37*35 - 36*37)/37 = (35-36)*37/37 = -$1, or 1/37th of your stake every spin.

So the risked amount is the same, the expected profit is -$1 each spin, but in the first case the variance is zero, and in the 2nd case it's higher.

What roulette bet has the highest variance? Is that the one that's "best" to play?

Depends on what you are after. a bet on a group of 18 gives you the best chance of winning, but has the smallest payout. 1 bet on 1 number has the best payout, but the hardest way to win. Some combinations can make it hard to impossible to extra hard to make a profit, so you have to pay attention. If you put 9 corner bets down you will only make a profit if you win more than 1 of the corner bets, so it doesn't really make sense to put the other one's down (unless you enjoy seeing the dealer move chips around.)

The important point in this is that you can't beat the house edge. You can assist the casino to make it harder for you to win by doing stupid shit like betting on every number in roulette, or martingaling, but you can't beat math.

[b!z]

not on dices.
No matter what you do the house edge will be the same. No way to change it

Or any gambling game with a house edge, for that matter.

[dooglus]

as gamblers we want the variance to be as high as possible?

Depends on what you are after. a bet on a group of 18 gives you the best chance of winning, but has the smallest payout.

Betting on every number, including 0, has the best chance of winning - you 'win' every spin. It's just that you only win 36 units but pay 37 units.

Doing that gets you the expected result of losing one 37th of your stake, every spin.

As a gambler, you know that in the long run you're going to lose 1/37th of everything you bet. Your hope is to win big early on, and quit. That's your only chance of coming out ahead.

Betting on every number makes "the long run" very short. It has zero variance, and so you have no chance of "beating the odds".

I calculated the variance for each of the roulette bets: covering a single number, a pair of numbers, 3 numbers, etc.

It seems to me that the less numbers you cover, the higher the variance, and so the more chance you have of being ahead at some point.

cover    payout   chance   expect variance
----- --------- -------- -------- --------
    1 36.00000x   2.703% -0.02703 34.08035
    2 18.00000x   5.405% -0.02703 16.56684
    3 12.00000x   8.108% -0.02703 10.72900
    4  9.00000x  10.811% -0.02703  7.81008
    6  6.00000x  16.216% -0.02703  4.89116
   12  3.00000x  32.432% -0.02703  1.97224
   18  2.00000x  48.649% -0.02703  0.99927
   36  1.00000x  97.297% -0.02703  0.02630
   37  0.97297x 100.000% -0.02703  0.00000

Here's how I came up with that table:

Code:

#!/usr/bin/env python

targets = [1, 2, 3, 4, 6, 12, 18, 36, 37];

print '%5s %9s %8s %8s %8s' % ('cover', 'payout', 'chance', 'expect', 'variance')
print '%5s %9s %8s %8s %8s' % ('-----', '---------', '--------', '--------', '--------')
for target in targets:
    payout = 36.0 / target
    chance = target / 37.0
    win_profit = payout - 1
    lose_profit = -1
    average_profit = chance * win_profit + (1-chance) * lose_profit
    variance = (   chance  * ( win_profit - average_profit)**2 +
                (1-chance) * (lose_profit - average_profit)**2)
    print '%5d %8.5fx %7.3f%% %8.5f %8.5f' % (target, payout, chance*100, average_profit, variance)

[sakira]

occasional use martingale, if already get enough profit, STOP.

and then I use another method

[dooglus]

No matter what you do the house edge will be the same. No way to change it

Or any gambling game with a house edge, for that matter.

You can change the edge on some games. You can play blackjack really badly, hitting until you bust every chance you get. The house edge for such a player is massive. You can play blackjack really well, taking advantage of your knowledge of the composition of what is left in the shoe to adjust your bet size according to your chance of winning the next hand, and to adjust your playing decisions within a hand. If you know the remaining cards are rich in low cards and low in 10's, you'll want to hit your 16 against a dealer 10. For example.

Even just by refusing insurance every time you're getting a lower edge than the guy next to you who takes it.

[el kaka22]

ahh martingale reeeeally work on sport betting with single play dont parlay but still doesn't work in dice
i use martingale in slots as well

[djibrilantiuang]

just dont do the martingale, its sucks! i have lost all i have with that suck method.
so i recommended dont do martingale.

[ranlo]

No matter what you do the house edge will be the same. No way to change it

Or any gambling game with a house edge, for that matter.

You can change the edge on some games. You can play blackjack really badly, hitting until you bust every chance you get. The house edge for such a player is massive. You can play blackjack really well, taking advantage of your knowledge of the composition of what is left in the shoe to adjust your bet size according to your chance of winning the next hand, and to adjust your playing decisions within a hand. If you know the remaining cards are rich in low cards and low in 10's, you'll want to hit your 16 against a dealer 10. For example.

Even just by refusing insurance every time you're getting a lower edge than the guy next to you who takes it.

Edge on some games is a tricky thing as well... a new one I like (High Card Flush) has an "average" house edge, and then separate ones from there based on whether you're a good or a bad player (like your BJ scenario). I think what people don't realize, though, is that at the end of the day, the house is always still going to be positive. It can be diminished somewhat with some games, but you won't negate it completely.

[cramved]

If you have a big enough bank roll and small enough bets the chance of losing is very small although it is very small there is still a chance

[dooglus]

Edge on some games is a tricky thing as well... a new one I like (High Card Flush) has an "average" house edge, and then separate ones from there based on whether you're a good or a bad player (like your BJ scenario). I think what people don't realize, though, is that at the end of the day, the house is always still going to be positive. It can be diminished somewhat with some games, but you won't negate it completely.

And when enough people get good enough at the game to get close to negating the house edge, they'll change the rules on you... continuous shuffling machines to stop card counting, blackjack pays 6:5 instead of 3:2, etc.

[Hfleer]

ahh martingale reeeeally work on sport betting with single play dont parlay but still doesn't work in dice
i use martingale in slots as well

it doesn't matter where you use it. it is always negative EV.

[XinXan]

ahh martingale reeeeally work on sport betting with single play dont parlay but still doesn't work in dice
i use martingale in slots as well

it doesn't matter where you use it. it is always negative EV.

Well you cant really calculate the EV on sports because it changes. And there is people that win a lot of money on sport bets so its not always negative

[dooglus]

it doesn't matter where you use it. it is always negative EV.

Not strictly true. I've seen people martingale their investment on Just-Dice:

Invest 10 units for an hour. If it makes a profit, leave it, if it makes a loss, double the investment and wait an hour. Repeat.

Since investing is +EV, the net effect of the martingale is also +EV.

A better statement is that martingale doesn't change your EV. If flat betting is -EV then so is martingale betting.