Analysis of Martingale

I have to admit that I haven't got to this. My feeling is you may well be right.

Back in post #11 you said that "If a system has a win/loose percentage of 65% and I double up every time I loose the odds are in my favour."

Why not double up every time you win, also, since the 65% win rate is an 'overall' measure? Unless that you can prove to yourself that, with your strategy, the probability of a win is higher immediately following a losing trade, than it is following a winning one.

This is what most Martingalers fail to realize: the bet (position) size should somehow reflect the probability currently being offered by the market, not the trader's personal goal of recovering previous losses to date. If the probability of an upcoming trade's success can't be calculated (relative to prior trades, e.g. an 'A-grade setup' offers a better probability than a 'B-grade setup'), then there's no justification for increasing position size. As to recovering personal losses, it goes without saying that the market doesn't care about an individual trader's account balance. For example, after doubling up 5 times (....2....4....8....16....32), the market would now need to offer something like 32x the probability of a successful trade over the original loss, to justify a 32x increase. In a zero sum game involving willing buyer/willing seller, I can't see how one setup can ever be 32x 'better' than another.

The other point i would like to make clear is the above. It is the poorest risk to reward situation as you get deeper and I am sorry to say but very few of you would be able to face the reality when the drawdown hits.

If you were trading with infinite money then you would not need to trade...

If you had deep pockets I VERY mch DOUBT you would be willing to double a million dollar loss, just to make a few dollars (the orginal gain at x=1). This is the flaw, not in the probability that the market will change but in the psychological aspects of getting to a point where the risk is so huge and the reward is so low.

If your starting goal was to gain $100 out of a trade then in the situation where you had 15 losses then you would be looking at 1,1,2,4,8,16,32,64,128....... do the maths and work at where you would be at in order to get that original $100.

Yes there are times when you would get your original win back but is it really worth it.

Consider the alternative of using the martingale for a win. In the event that you had 15 winners in a row and you doubled up, would that not be a better experience....

If you consider the story of a pit boss who worked at a Las Vegas casino. He was working one night when a 'black swan' event happened on the roulette table. 'Gamblers' from all around could hear the commotion as the number black had come up 10 times. They would come, put there money on red commenting on the theoretically probabilities that were and sure enough black came up. No way, so they went on red again. 15 times black, 20 times black. The further the sequence went the more 'Gamblers' it attracted. It would seem the whole casino was here to put money on red, but the pit boss noticed NO one put there money on BLACK.

The sequence lasted some 40+ spins at which point there was no one there, they were empty, the casino had cleaned up on an event that NO ONE EVER ACKNOWLEDGES, Momentum and the unlkely. A patron commented to the pit boss at how unlikely it was. The pit boss commented ' Not as unlikely as you think, but they always come trying to be there when the change happens, unfortunatly for most it comes too late.

Razor

Hanover,

I Read the info you provided. Pure martingaling where you double down after every loss with the hope of winning it all back on one spin of the wheel is not what I was talking about.

Averaging in systematically when price goes against my position is what I do. The difficult part with forex is determining the range where price can move. It's alot easier in other markets.

While Commercials (banksters) don't martingale,they do add to losing positions without leverage. They have the funds to do it. The latest DXY COT shows their record short position they have been adding to into dollar strength. It also shows the massive long position of the Non Commercials (leveraged speculators) accumulated by chasing price. The Banksters will win. They always do.

I apologize if my earlier post (#20) came across as being conceited, or harsh.

Believe me, I'm not a naysayer by choice. I've bashed my head against a wall trying to find a way of using MM to shortcut the path to success. If somebody was able to show me a watertight mathematical proof that some form of progressive (or otherwise innovative) sizing offered an edge in itself – apart from simply increasing size relative to the potential quality of a setup – then I'd be eternally grateful, LOL. I'm just as keen to find a shortcut toward making a forex million as anybody else.

One final point: Cobra (post #13) talks about 'backtesting' a Martingale. How does one test for the likelihood of a 'death trade'? Any black swan event, is by definition, an unlikely and unforeseeable outcome. We could have 20 years worth of data without a death sequence, and yet one is just as likely to occur tomorrow as it is 50 years from now.

It's true that we don't have to double up to the point of wipeout. But let's say we set a limit of 5 losses. At that point we've lost 1+2+4+8+16 = 31x our original stake. A further double would allow us to recover it in a single trade. But by setting the limit, we presumably must now return to our original 1 unit position size. Now it's going to take 31 consecutive wins to return the account to breakeven. If we set our limit to 6 losses, it's now going to take 63 consecutive wins to return the account to breakeven. And so on. Sorry, but I can't see how that is a viable MM methodology.

No it doesn't. Martingale is a staking strategy, and staking strategies do not provide an edge, they simply change the distribution of wins and losses. You stated you rarely see 10 losses in a row - sure, but it is likely to happen at some point, and the times it happens will make up for the wins in between. Any system that uses martingale would have the same edge without it.

All that will do is up the winrate at the cost of occassional large losses. The profit won't be any greater if equivalent risk is maintained. If equivalent risk isn't maintained, the profits may be greater, but that would only be due to increased turnover.

If the further price moves away from a mean, the greater the probability of a reversal, then increasing position size as one averages down will increase winnings. That logic makes perfect sense to me. However, if I may be pedantic, that's made possible because the edge comes from the exploitation of the price behavior, rather than the MM itself. Comparing this to something like roulette, where the probabilities remain constant, makes the distinction more obvious.

As I attempted to point out in post #22, the goal should be to size one's positions so that they remain realistic, in terms of keeping with the probabilities offered by the market. Doubling up recklessly, or without limit, is tempting fate.

Nice explanation, Mr J, much simpler and more eloquent than mine.

Glad somebody understood what I was trying to say . The argument against martingale could be as simple as questioning why good no good traders use it. I also saw a mention of averaging/scaling into a position. These people aren't increasing their position because their previous trade lost, but because they think the price warrants another entry. Very different.

I also thought your explanation was very lucid.

I still remember as clear as yesterday,

I was a sophomore in college, thought i had the "holy grail", standing by the roulette table at Claridges as the ball was spinning in that particular "series" for the fourteenth time, and the host while writing me a coffeeshop coupon and obv taking all contact details, managed to put her hand on my head and gently nudge it away drom the wheel. She whispered "dont look, you what it's gonna be"!

See me and my best friend were total boneheads, we had a bankroll of 5K or so with which we used to come and pick nickels in front of a steamroller every friday at AC. We actually thought we were super safe, what with waiting for three reds in a row before our "Entry". Well $500 was our goal every visit and we were successful till our fourth, when i saw 10 deadly reds after three harmless ones roll by.

I guess this is off topic, so sorry, but martingale always stirs up memories for me, lol.

The ball has no memory, whatsoever.

Fourteenth was black obv.

^^^^ Owch lol!

The "safer" you are, the lower the return.

If you wanted to be "certain" that you wouldn't blow up, then you wouldn't be making any return worthwhile.

Maths still works, Martingale still doesn't, who da thunk it?

You say that martingale works by gambling on price retracements, but instead of betting on a. Retrace it would to be much more profitable to double up in the direction of the last candle/bar. After all, how often, especially on the lower timeframes, does the price action go green bar, red bar, green bar, red bar etc. You would be using martingale to support a trend, even a small one.

martingale trading is easy if know what to do:

- max level 9 (most 1-4)
- doubled entry steps, like 25, 50,100,200,....
- reduced profit on higher levels (closes mart. orders earlier)
- no stop loss
- well suited account
- trade 10 pairs with 0.01 then 1 pair with 0.10 lot size

personal i trade martingale, d'alembert and fibonacchi algorythm since 2004, it works - if you know how it works and you understand how it can help you.

attached a 01/2008 till 05/2011 martingale with dukascopy 99% tick data