Does martingale really works?

[GannickusX]

He said about using 1 satoshi as base bet, so I was thinking that he meant martingale (as the headline suggests). With a negative expectation he would never get 0.01 btc in profits without successively increasing his bets starting at 1 satoshi (doubling or however else)...

I understand, but he is talking about having to bet 0.01 BTC in a single bet, which you would have to do on around your 21st bet in a losing streak if you were doubling each time. (2^20 ~= 1e6)

That is in 20 steps as I said in my post. I just don't know how one would have to play around 2 million bets to get 0.01 btc if he started at 1 satoshi...

To get 0.01 profit not to bet 0.01 ive never said anything about betting 0.01

[ranlo]

That math actually shows that you have a 2x greater chance to bust than to double up, as well (which does coincide with my own findings). Amazing how big that is! Basically, you'll bust 2x for each time you double. I think that really helps put it into perspective, rather than just looking at the 1%, 0.5%, etc. house edge.

It's even worse than that.

To double up you need to win $50 177 times.

The chance of winning 177 times, when each individual win is a 99% chance is:

0.99 ** 177 = 16.88%

So your chance of busting is 83.12%

ie. the chance of busting is 4.9 times bigger than the chance of doubling up (if you do it this way, trying to win $50 each time)

If your goal is to double up, there are more effective ways of doing it. But the chance of doubling up will always be smaller than the chance of busting if the house edge is positive.

Wow. So basically... the best chance to double up would likely be to throw it in at 50% and hope you hit. At least then you have almost the same chance to double as you do to lose it all (just slightly off). Not to mention you save the time of constantly betting.

[deisik]

He said about using 1 satoshi as base bet, so I was thinking that he meant martingale (as the headline suggests). With a negative expectation he would never get 0.01 btc in profits without successively increasing his bets starting at 1 satoshi (doubling or however else)...

I understand, but he is talking about having to bet 0.01 BTC in a single bet, which you would have to do on around your 21st bet in a losing streak if you were doubling each time. (2^20 ~= 1e6)

That is in 20 steps as I said in my post. I just don't know how one would have to play around 2 million bets to get 0.01 btc if he started at 1 satoshi...

To get 0.01 profit not to bet 0.01 ive never said anything about betting 0.01

Okay, but how on earth are you going to win this amount if you start at just 1 satoshi? How did you get that number, 2 million bets? I guess you won't be able to do that with any number of bets since at around two million bets you will be deep in loss...

[GannickusX]

He said about using 1 satoshi as base bet, so I was thinking that he meant martingale (as the headline suggests). With a negative expectation he would never get 0.01 btc in profits without successively increasing his bets starting at 1 satoshi (doubling or however else)...

I understand, but he is talking about having to bet 0.01 BTC in a single bet, which you would have to do on around your 21st bet in a losing streak if you were doubling each time. (2^20 ~= 1e6)

That is in 20 steps as I said in my post. I just don't know how one would have to play around 2 million bets to get 0.01 btc if he started at 1 satoshi...

To get 0.01 profit not to bet 0.01 ive never said anything about betting 0.01

Okay, but how on earth are you going to win this amount if you start at just 1 satoshi? How did you get that number, 2 million bets? I guess you won't be able to do that with any number of bets since at around two million bets you will be deep in loss...

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

[deisik]

That is in 20 steps as I said in my post. I just don't know how one would have to play around 2 million bets to get 0.01 btc if he started at 1 satoshi...

To get 0.01 profit not to bet 0.01 ive never said anything about betting 0.01

Okay, but how on earth are you going to win this amount if you start at just 1 satoshi? How did you get that number, 2 million bets? I guess you won't be able to do that with any number of bets since at around two million bets you will be deep in loss...

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

[dooglus]

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you probably won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and probably so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

[Berau]

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

Maybe one could have a bot which allows you to use multiple accounts to bet at once, maybe even using virtualmachine. But nonetheless the profit is extremely small.

[jemsmuhamad]

sometimes it works on me, but sometimes martingale killing me.
so i think martingale method its just wasting your money.

indeed.. martingale maybe only work once in a while, not in a long time ...
i only use a martingale if i am sure the results 2-3 times roll over after lost i will get win

yeah as my experience using martingale is not working everytime, just sometimes.
so if just once or twice in a day to try using martingale its just okay. but dont do it continously.

[dooglus]

Maybe one could have a bot which allows you to use multiple accounts to bet at once, maybe even using virtualmachine. But nonetheless the profit is extremely small.

One could, and people do.

More important than "the profit is extremely small" is "the expectation is negative". And so by speeding up the process by using bots, multiple accounts, etc. you are only accelerating your own demise.

[anshar]

Primedice logic:

Slower bets = more time to lose

So the next you complain about slow rolls, thank PD for not eating your BTC so quickly!

[kpitti]

Yes, it works, very well. You win small fractions and you are happy. Your bankroll is slowly increasing and you are more and more confident, that this is a great idea ... you will be rich. What you need is a time only, but you have time. You play and play .... no once happen series of losses four in row, five ... seven, !eight! ... may be more. You still confident so you increase your bet and hoping for return.

No it does not happen and you lost all and time you invest in the martingale
Bad luck? No just math and odds.

Good luck!

[GannickusX]

Primedice logic:

Slower bets = more time to lose

So the next you complain about slow rolls, thank PD for not eating your BTC so quickly!

Yes slow bets are actually better, real roulettes are slow, you have to wait each bet a good amount of time and in between bets you have time to think so you dont act foolish and bet all your money on red, its still happens but is better slow betting. I used to play on 999dice and you could bet so fast it was easy to make mistakes and easier to loose

[deisik]

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

He is not talking about martingale or any other ways of increasing his bets. As I understood, he spoke about betting the same amount (that is 1 satoshi) over and over again ("you win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis which is 0.01")...

[shanem]

Martingale don't work.
The house edge no matter how small will show up in your bets eventually.
The house has a tiny edge in every bet you make which will cause you to lose all your btc in the long run.
For those who claims that martingale can work is just lucky at the moment.

[GannickusX]

You win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis wich is 0.01 but you wont be able to win 1 million bets straight you would have to bet a little more around 2 million bets in total, its not an exact number, and you wont be in deep loss necesarly if you have a big bankroll and you are not extremely unlucky

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

He is not talking about martingale or any other ways of increasing his bets. As I understood, he spoke about betting the same amount (that is 1 satoshi) over and over again ("you win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis which is 0.01")...

Nope dooglus explained it in detail, that was exactly what i meant, with 1 million "won" bets, because when you use martingale you gonna lose some bets thats why i said you would need to bet a lot to even get 0.01 i was just trying to say that even if you can bet extremely low amounts of money, martingale would still not work and would actually be even worse

[deisik]

I hope dooglus can explain it better than me why this will never work out (you won't listen to me all the same). In any case, you can check your theory through automatic betting, and you don't need to roll 2 million times. I think 2-3 thousand rolls will suffice to persuade you...

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

He is not talking about martingale or any other ways of increasing his bets. As I understood, he spoke about betting the same amount (that is 1 satoshi) over and over again ("you win 1 satoshi with each win bet, with 1 million won bets you get 1 million satoshis which is 0.01")...

Nope dooglus explained it in detail, that was exactly what i meant, with 1 million "won" bets, because when you use martingale you gonna lose some bets thats why i said you would need to bet a lot to even get 0.01 i was just trying to say that even if you can bet extremely low amounts of money, martingale would still not work and would actually be even worse

I don't see much difference if you make just one bet at 1 satoshi incessantly, or 2 bets, first bet at 1, second at 2 satoshi, and then return to the base. In any case, as I said before, you can try automatic betting to see if it actually works...

[dukeneptun]

Martingale don't work.
The house edge no matter how small will show up in your bets eventually.
The house has a tiny edge in every bet you make which will cause you to lose all your btc in the long run.
For those who claims that martingale can work is just lucky at the moment.

But it works only longtime. I do believe it can make at least 1% return.

[arallmuus]

Martingale don't work.
The house edge no matter how small will show up in your bets eventually.
The house has a tiny edge in every bet you make which will cause you to lose all your btc in the long run.
For those who claims that martingale can work is just lucky at the moment.

But it works only longtime. I do believe it can make at least 1% return.

you dont believe anything like this, it is indeed a math where you can count the posibility , here read this explanation by dooglus who is an expert in gambling house edge calculation, and this is just a few post above yours

quoting the below as to hope that everyone can read instead of keep repeating the same question and thoughts over and over again

He's right, actually. Every winning bet earns you a single satoshi in net profit. After a million winning bets you'll have 0.01 BTC profit. The average martingale sequence is 2 bets long, and so it takes 2 million bets to be 0.01 BTC up.

In 2 million bets you're unlikely to see a losing streak longer than length 25, so as long as you have a big enough bankroll (and the house has a big enough maximum bet) to let you double 25 times, you won't bust. Of course, doubling 25 times will cause to you be reaching around 1 BTC bets, but that's OK.

In 3000 rolls you probably won't see a losing streak of length 15, and so won't be betting more than 0.00065536 in a single bet.

Again, this isn't a good idea, but it would probably work if you're happy to risk 10 BTC to earn 0.01 BTC and are wiling to place 2 million (mostly) tiny bets.

i think this is the best explanation from this whole page

[badjacks99]

i think martingale can work in the short run, taking a set amount in to play and setting a goal and stick to it. once you hit the goal, leave for the day. its never foolproof and eventually its gonna fail. GL

[dooglus]

I don't see much difference if you make just one bet at 1 satoshi incessantly, or 2 bets, first bet at 1, second at 2 satoshi, and then return to the base. In any case, as I said before, you can try automatic betting to see if it actually works...

He's talking about using Martingale betting, where you double your stake until you win, then reset to the base bet size.

The two strategies (flat betting vs. Martingale) give vastly different probability distribution functions:

* flat betting gives you a high probability of losing a little and lower probabilities of doing slightly better or worse than that; it looks like a bell curve centred on a loss of 1% of the amount you stake

* martingale betting gives you a high probability of winning a little, and a low probability of losing a lot; there are two separate non-zero areas on the curve

Here's an image showing what I'm talking about:



The blue area is centred around a small negative number, with its tail crossing into positive numbers.
The red area shows a high chance of winning a little and a low chance of losing a lot.