This really doesn't deserve a whole thread, but anyway...

[Mata Nui]

Hotch, It probably does deserve its own thread as the arguments around it highlight the sheer amount of BS and misinformation that you get on most trading forums. There is no sensible discussion just prople throwing out the same arguments they read elsewhere.

I dont use martingale in trading but i have used it on a few european roulette tables and very rarely come away with less than i started with.

 

[andycon]

Well it's quite simple, if that is "how probability works", then dependency doesn't exist.

That wasn't what I was saying. The coin toss probability is 50/50, yet you can still get a string of either heads or tails of 8, 10, 15 which short term seems to go against the 50/50 (the coin has no memory), however on the larger sample size i.e. 100 or 1000 the odds revert back to 50/50 (the coin must have a memory). The Martingale system fails because if trading at 1% of account you have insufficient funds to double up after 7 losses, which on a 50/50 system are highly probable. Even at 0.1% you have insufficient funds to double up after 9 losing trades which is quite probable.

 

[Hotch]

The same circular arguments seem to be coming out on both sides.

I'm not arguing, just stating facts, others are rattling off their opinions as if they think that they're important.

 

[Hotch]

That wasn't what I was saying. The coin toss probability is 50/50, yet you can still get a string of either heads or tails of 8, 10, 15 which short term seems to go against the 50/50 (the coin has no memory), however on the larger sample size i.e. 100 or 1000 the odds revert back to 50/50 (the coin must have a memory). The Martingale system fails because if trading at 1% of account you have insufficient funds to double up after 7 losses, which on a 50/50 system are highly probable. Even at 0.1% you have insufficient funds to double up after 9 losing trades which is quite probable.

What you said was:

"The Martingale system also assumes that the probability of a win increases after a string of losses, but this isn't how probability works."

EDIT, I'm being nice:

How about I make it simple for you.

We bet on whether a runner will win a 100m race, he races first against bob, immediately after racing bob, he races fred, who happens to bob's identical twin brother, who is, or all intents and purposes, exactly the same, bar responding to fred instead of bob.

Now, wouldn't you agree, that just maybe, that our runner is less likely to win the 2nd race?

because you've implied that it's exactly the same.

The last bit implies that everything is independent. Which I berated you for.

 

[wackypete2]

I'm not arguing, just stating facts, others are rattling off their opinions as if they think that they're important.

If other's people's opinions mean nothing then why start the thread. What were you expecting to get? a thread full of textbook facts?

Peter

 

[andycon]

What you said was:

"The Martingale system also assumes that the probability of a win increases after a string of losses, but this isn't how probability works."

EDIT, I'm being nice:

How about I make it simple for you.

We bet on whether a runner will win a 100m race, he races first against bob, immediately after racing bob, he races fred, who happens to bob's identical twin brother, who is, or all intents and purposes, exactly the same, bar responding to fred instead of bob.

Now, wouldn't you agree, that just maybe, that our runner is less likely to win the 2nd race?

because you've implied that it's exactly the same.

The last bit implies that everything is independent. Which I berated you for.

This scenario is not really relevant to the conversation, but no I wouldn't agree. It could be that the runners ability is so superior to both Bob and Fred that he has the ability to run two races back to back and beat them both. I was talking about sample size and the fact that most people who believe in Martingale use an insufficient sample size. Now if our runner had to run 10 races back to back could he still win them all, possibly depending on the ability of the competition. But could he win 100 races back to back regardless of the ability of the competition?

 

[barjon]

The trouble with mathematical theory applied to trading is that it totally ignores emotional human beings . This thread is the flip side of all those who were going to compound their way from £100 to £1m in no time at all.

So far as Martingdale is concerned you may well enjoy success for some time until you hit that inevitable "uncle" point because you either run out of money or simply can't bring yourself to risk house and furniture to win £10. BBMac came up with the thumbnail some time ago which is interesting from the theoretical perspective.

jon

 

[Hotch]

This scenario is not really relevant to the conversation, but no I wouldn't agree. It could be that the runners ability is so superior to both Bob and Fred that he has the ability to run two races back to back and beat them both. I was talking about sample size and the fact that most people who believe in Martingale use an insufficient sample size. Now if our runner had to run 10 races back to back could he still win them all, possibly depending on the ability of the competition. But could he win 100 races back to back regardless of the ability of the competition?

I said nothing on whether he would win both, just that the chance would be less. Glad to see you're sticking to your "everything is independent" stance.

 

[Hotch]

If other's people's opinions mean nothing then why start the thread. What were you expecting to get? a thread full of textbook facts?

Peter

I was trying to help educate, improve the quality of the forum etc. I didn't want a discussion on martingale, it's largely irrelevant to the OP. I was just hoping that people would actually think before they post stuff which is obviously false.

 

[andycon]

I said nothing on whether he would win both, just that the chance would be less. Glad to see you're sticking to your "everything is independent" stance.

With regard to life I believe the opposite, with regard to a coin flip provided the coin is true and the flip is true then yes each flip is independent and the odds on each flip will remain at 50%.

 

[andycon]

The trouble with mathematical theory applied to trading is that it totally ignores emotional human beings . This thread is the flip side of all those who were going to compound their way from £100 to £1m in no time at all.

So far as Martingdale is concerned you may well enjoy success for some time until you hit that inevitable "uncle" point because you either run out of money or simply can't bring yourself to risk house and furniture to win £10. BBMac came up with the thumbnail some time ago which is interesting from the theoretical perspective.

jon

The person who believes they have an 85% winning system may look at BBMac's thumbnail and employ a martingale strategy thinking they have such a low probability of getting the long losing run. The problem is they are likely to have used an insufficient sample size in determining there 85% win rate. Long Term Capital Management being a classic example.

 

[ffsear]

trade 1 risk £100
trade 2 risk £200
trade 3 risk £400
trade 4 risk £800
trade 5 risk £1600
trade 6 risk £3200
trade 7 risk £6400
trade 8 risk £12800
trade 9 risk £25,600
trade 10 risk £51,200

All this to chase you origional £100 loss???

No thanks!

 

[Hotch]

With regard to life I believe the opposite, with regard to a coin flip provided the coin is true and the flip is true then yes each flip is independent and the odds on each flip will remain at 50%.

Trading isn't a coin flip. Glad we finally got there.

I'm not saying whether there are or aren't strategies where the results are independent, I would guess there are some of each, but I'm not stating either of them.

I was merely pointing out, that if your argument against martingale is that you could have a long run of losers wiping you out, that that same argument applies to all trading, and so you shouldn't be doing it.

Personally my reasons for not using martingale is that it doesn't make me as much money

 

[Shikamaroo]

I was trying to help educate, improve the quality of the forum etc. I didn't want a discussion on martingale, it's largely irrelevant to the OP. I was just hoping that people would actually think before they post stuff which is obviously false.

False? but u haven't disproved the independence of trading events. I personally dont see how two separate trades are dependent on each other which would be one requirement for the martingale to succeed. If i take a trade long and then another trade long and the first one failed, is there any reason to believe that the second long trade has a better chance of a profitable outcome?

I also dont understand your need to go into running to find examples of probability since the discussion is on trading mechanics not pure probability theory.

 

[safvan]

It can be used as a long term strategy for big interest holders who are looking for value in good companies they expect to outperform in the future.

Virtually all trading techniques work out well in a strong trending market. If things are going down or sideways...all strats would fail.

Trend is the primary indicator.

The research should be whether martingale works well in an uptrend/downtrend or sideways market and also what time frames. Trying to make it work on two min charts may not cut it when the weekly trend is sideways. You never know when that big reversal comes.

 

[wallstreetwarrior87]

trade 1 risk £100
trade 2 risk £200
trade 3 risk £400
trade 4 risk £800
trade 5 risk £1600
trade 6 risk £3200
trade 7 risk £6400
trade 8 risk £12800
trade 9 risk £25,600
trade 10 risk £51,200

All this to chase you origional £100 loss???

No thanks!

The main point that no one is really picking up on is that trading is not fixed odds. There for why would you only increase your "stakes" just to return £100??? Need to compare apples with apples. Trading is not the same as a coin flip or roulette spin. For example what if on trade 8 you get a runner, and profit to the tune of £25,000. Would not seem so bad then. Once again traders are trying too hard to control the "win" side, ie limiting your upside, why do this?

As one poster pointed out, dont think that major trend traders dont adopt this style, because they do. The question to ask is if this is the best use of your capital. Martingale staking leads to irrational participation in most, ie not participating at the right moments. So if you can not control yourself it wont matter what staking method you use, because you havent developed the basic skills required anyway.

Also remember once you have placed your bet on a roulette table and the ball is spinning you have to stay with it and hope for the best. In trading you can cut when ever you want. This is why talking about trading and fixed odds in the same sentence is flawed. In fact it is the opposite, in trading you loss is limited (by your risk) but your profit is unlimited. This can not be said for coin tosses and roulette.

Fixed stops, fixed targets, many threads on here concerned with this, let the stabilisers come off, and let things ride, there is no need to be obsessed with being in control.

 

[Hotch]

False? but u haven't disproved the independence of trading events. I personally dont see how two separate trades are dependent on each other which would be one requirement for the martingale to succeed. If i take a trade long and then another trade long and the first one failed, is there any reason to believe that the second long trade has a better chance of a profitable outcome?

I also dont understand your need to go into running to find examples of probability since the discussion is on trading mechanics not pure probability theory.

If you look at the OP, I don't say separate traders are dependent, I never set out to prove it. I said there are possible times when it is.

Once again, the OP merely states that you can have a run of losers and wipe out an account using 1% or whatever, just the same way you can have a run of losers and wipe out using martingale.

I could come up with reasons why trades might not be independent, but I really think it will just muddy this thread even more, if I can't get people to understand the OP then I have no hope explaining that.

 

[Joey25]

I have read once again that "it is a statistical certainty that martingale systems will blow up", and that this is why you shouldn't use them.

Well fine, but if it's a certainty that a martingale will blow up, it's also a certainty that any type of strategy will blow up.

I have been impressed with the sudden increase of quality in T2W of late, and am hoping that we can continue this improvement by thinking before we post the same old mantras. It's this sort of blind faith BS which stops people being profitable.

I'm not saying martingale is good, or I condone it or anything, having never used it. I just find it shocking that people are that retarded.

I eagerly await the replies of "durrrrr, martingale bad, hurrrrr" from those who once again have failed to read what I post.

Thinking about it, the best way to play these games is to exploit a big edge - the ability to walk away.

If the game is fair, so that the probability of a $1 win is the same as the probability of a $1 loss, then it can be shown that the probability of a gambler with a bank of x reaching y before losing x is:

Pr(Getting to Y before losing the bank X) = X / Y

So someone trading 1:1 trades where the outcome of each is equally probable has a 3/5 chance of getting to 5 before losing all 3. Similarly, a 1/5 chance of getting to 5 before losing a bank of 1.

The martingale strategy is just a variation on this - as long as the trader is willing to walk away, then he knows the probability of achieving this before going bust. It can make a roulette night span out quite nicely, but whether a trader could genuinely walk away after turning $100,000 into $150,000 I don't know.

 

[Masquerade]

Good job on the trolling, you managed to get about 5 pages on it. :clap: