Double up and get rich???

Maybe I didn't understand what you meant but the probability of 10 losses in a row with a system which produces only 5% losses is 0.05 ^ 10 and not 5%.

Actually is that amount multiplied by 100, if you consider the result in percent.

That would only be the case if the previous loss effects the outcome of the next trade.

As in flicking coins... the probability of flicking 1million heads in a row is 50%. Because flicking a coin is a event which has no effect on future events. (Obvious considering perfect 0 bias flicking here )
So only trading systems where the previous trades effect the current expectation of trades would produce 0.05 ^10 *100 (for %).

My premise is that we can not measure the winning % of a trade before taken. At best we can use previous trades to predict the %. But even this is a fallacy. So to assume that 10 winning trades have a effect on each other and use that fact to predict max. consecutive loses etc... is a mute point.

The only way to create a model of possible max. losses is to first prove a interdependency between losing trades. Then we can look at the probabilities of losing trades following each other.

LL

Let's cut the crap and play a game. We toss a coin. For every 4 heads in a row I give you 100$. For every tail you give me 50$.
Your chance being 50%, you will become rich. Deal?

In that case the return is dependant on a sequence. The probability doesn't change but the return does. And you giving me less than the odds the coin gives me.

All trades do not have the same probability... thus it is extremely difficult to
calculate the % of a particular sequence. I.e 10 losing trades in a row.

LL

ok... lets play.

http://en.t45ol.com/play/3512/coin-flip.html

After 20 heads I'd accuse you of having a loaded coin if not double sided.

I do not understand that this topic get so much attention. This has been proven over and over again that you need a will of steel to survive never mind the discipline to carry on and yet the robot will fail.

I asked for some honest help in another topic and had 1 reply in the last month yet this over discussed topic never stop generating opinions.

Funny thing. I will sign off from this forum now as advertised free help is not even that.

Thank you

4xtrend(Pierre)

Your calculations are incorrect.

As in flicking coins... the probability of flicking 1million heads in a row is 50%. That is NOT correct. The probability of flicking EACH of the million heads in a row is 50%.
The probability of flicking 1 million heads in a row is 0.05 ^10

I must disagree....

What's the probability of flicking 1 head ? 50%
The next head ? 50%
The next head ? 50%

A coin has no memory. Every coin flick is a independent event. The events are unrelated...

I stand by what I said. The probability of flicking EACH of the million heads in a row is 50%. The probability of flicking 1 million heads in a row is 0.05 ^10
The probability of flicking 1million heads in a row is NOT 50%
Set it up on a bell curve and and then you will see.

The probability of flipping one head is 0.5

The probability of flipping 1 million heads, having already just flipped 999,999 of them, is 0.5

The probability of flipping 1 million consecutive heads from this point onward is 0.5 ^ 1,000,000 = a VERY tiny number.

and the probability that everyone will actually come to an agreement over any of this is 0.

about the same probability that i will become profitable.

It is true averaged over a big sample.
I give you the, "law of large numbers"

Nice coin Sir! Did you get it at the carnival?

edit: Is this thread a big joke?

This kind of thinking is REALLY interesting... There will be people in this world who have made fortunes and lived the most wonderful and interesting lives PURELY by chance, one fortunate event leading on to another, they just so happen to have been lucky their whole lives. These people WILL exist, somewhere.

I'm sure T&E speaks about this at some point.

^^ I was discussing this very thing yesterday. The sad thing is a human life isn't long enough for the "luck" to even out.
I guess it's no coincidence that some peole seem to live "charmed lives".

site is useful to make bets and draw hats , but the reason probabilities could be "FOR" you and "AGAINST" you. which is why you should only risk 1% of your capital, so you can handle 10 losses in a row.

and about the same probability i'll will the powerball lotto this week...

Like it or not, this thread has evolved into new topics...they always do. Anyways, here it goes...

1. I agree with mr.marketz and LittleLuca: Statistics are overused in trading, and it's probably just a personal comfort thing. But I guess, hey, whatever makes you happy.

The main problem with a lot of the stats people do is that the market is always changing. You can apply all kinds of statistics methods to coin flipping because that experiment perfectly meets the binomial distribution criteria. The odds of flipping heads is always 50% for every single trial...and all the trials are independent. There are other situations in nature like this where we have excellent statistical models and we can calculate all kinds of information about them.

In contrast, the odds of many phenomena in the market are not constant. Indeed, you can take most trading systems, test them on some price data and you'll get a certain % winning trades, average profit per trade, etc. However, if you test it on some older price history, or newer price history, the results will probably (no pun intended) change quite a bit.

Most of the statistics that people try to use assume that the data is "stationary" (for those who don't know the lingo, that basically means that the averages and volatility stay the same over time). But as we know, price certainly isn't stationary. Most of the things people try to analyze with stats in the markets are non-stationary.

Now there are more sophisticated mathematical methods that people can use...but to me, using all that complex mathematics to try and trade is like trying to kill an ant with a machine gun. (And this is coming from a guy who loves mathematics!) What can I say? I like my trusty trend line. It's simple and reliable.

So since the markets are always changing, and most of the statistics methods don't really apply, I think the key to success for a lot of good traders is flexibility.

2. Personally, I am against increasing the trade size after a string of losses to try and make back your money quicker. I think that's got trouble written all over it. I use a fixed fraction MM strategy. I usually risk 1% per trade. So after a loss, that naturally leads to a smaller trade size.

3. As billflet said, a lot of folks would consider a string of 10 losses an anomaly. It's funny how those same folks would probably consider a string of 10 wins to be solid evidence that they have a great trading system

4. Regarding Nassim Taleb: I agree that there is a good deal of randomness involved in trading. When I place a trade, I have no idea if it will be a winner or a loser. I suppose that means that luck is part of it. However, if you consider a trader who has been consistently profitable for many, many years, I think it's ridiculous to consider his success to be entirely luck. Such a trader obviously has a skill, and managing quasi-randomness is precisely that skill.

Of course some people do get rich entirely by chance:
a) I won the lottery! Yay!
b) My uncle suddenly died and he left me $1-million. Yay! err...I mean, poor uncle.

5.


Ted1983: What happens when you take a larger sample and the average changes? Maybe your sample is 100 trades and the average profit is +12 pips/trade. Then you take a sample of 1000 trades and the average profit is +1 pips/trade. Then you take a sample of 2000 trades and it's +20 pips/trade. Then a sample of 5000 trades gives -40 pips/trade. Stuff like that happens all the time with trading because the averages are non-stationary.

Now the Law of Large numbers works when your experiment fits nicely into a known probability distribution and you know the expected value for that distribution. But what distribution do prices follow?