Basement for Stategy | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Hedge Stategy tester software 13 replies

News Trading Stategy 4 replies

Big Brother Stategy!!! 1 reply

stategy tester - reliable? 4 replies

Stategy Builder FX - Metatrade Down? 3 replies

Basement for Stategy

  • #1
  • First Post: Jan 6, 2011 6:12am Jan 6, 2011 6:12am
Hi, I have been reading forum this morning and got interesting idea in my mind. There are so many strategies and EA, some of them are good, some off them are not very good.

The idea is simple and I can bet that somebody already had it but...

There are four strategies with 50% win rate. We need strategies that work on the same time period and use different instruments to make a forecast. Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%

Now what will we have if we take ten, fifteen systems. Yes we will have only couple trader per month but chance to win will be very high.
  • #2
  • Edited 6:30am Jan 6, 2011 6:26am | Edited 6:30am
50% win rate of 2 or more systems does not increase the odds.


Consider flipping a coin, the chance is 50% that it will turn up heads. If you flip two coins, the chance is always 50% for each.

A loss will cancel out a win...

You cannot increase your win rate by flipping more coins simultaneously, because you are evenly increasing your loss rate.
  • #3
  • Jan 6, 2011 10:25am Jan 6, 2011 10:25am
Quoting LocoGuy
...
There are four strategies with 50% win rate. We need strategies that work on the same time period and use different instruments to make a forecast. Chance that all this systems will fail together is 0.5*0.5*0.5*0.5 = 0,0625 it means that we will win in 100%-6,25%= 93.75%

Now what will we have if we take ten, fifteen systems. Yes we will have only couple trader per month but chance to win will be very high.
Here's the fatal flaw in your argument: the opposite of "all 4 systems fail in a given period" is not quite "we win." Rather, it's "at least 1 system wins." You've correctly calculated the probability of that outcome as 93.75%, as long as all the systems are truly mutually independent (which in practice is often not the case).

However, "at least 1 system wins" is a necessary but not sufficient condition for "we win / we make money" (assuming all the systems use equivalent risk and reward). Now you see why that is, right?

Simultaneously trading more than one positive expectancy strategy, though, can be a very good thing, in so many ways. Especially if they are based on different 1) underlying concepts, 2) market dynamics (trend, retracement, reversal, range), 3) time frames, 4) times of day or 5) currency pairs, to name a few of the ideal key differences.
  • #4
  • Jan 12, 2011 7:32am Jan 12, 2011 7:32am
Quote
Here's the fatal flaw in your argument: the opposite of "all 4 systems fail in a given period" is not quite "we win." Rather, it's "at least 1 system wins." You've correctly calculated the probability of that outcome as 93.75%, as long as all the systems are truly mutually independent (which in practice is often not the case).

I'm not agree with you. Here I see interesting seeds. So we have two independent day time strategies. Both make decision to open a position after bar is close. One showed 50% chance on back tests and another one showed 80% win rate. They are running now on different accounts. Here is the task: what is the possibility that first and second account each will make positive trades?

It is: 0.5 x 0.8 х 100% = 40%

What is the chance that at least one account will have positive trade?
  • #5
  • Jan 13, 2011 1:21am Jan 13, 2011 1:21am
Quoting jm8985
50% win rate of 2 or more systems does not increase the odds.


Consider flipping a coin, the chance is 50% that it will turn up heads. If you flip two coins, the chance is always 50% for each.

A loss will cancel out a win...

You cannot increase your win rate by flipping more coins simultaneously, because you are evenly increasing your loss rate.
Yeap, you are right there is something to think about!
  • #6
  • Jan 13, 2011 1:39am Jan 13, 2011 1:39am
Quoting LocoGuy
Yeap, you are right there is something to think about!
Man, there is good idea in your post. Don't leave it.
  • #7
  • Last Post: Jan 13, 2011 4:09am Jan 13, 2011 4:09am
Quoting Avalong
I'm not agree with you. Here I see interesting seeds. So we have two independent day time strategies. Both make decision to open a position after bar is close. One showed 50% chance on back tests and another one showed 80% win rate. They are running now on different accounts. Here is the task: what is the possibility that first and second account each will make positive trades?

It is: 0.5 x 0.8 х 100% = 40%

What is the chance that at least one account will have positive trade?
I'm curious which part of that quote you selected you don't agree with. Your post doesn't tell us that. Here's the quote, once again:

Quoting nhct
Here's the fatal flaw in your argument: the opposite of "all 4 systems fail in a given period" is not quite "we win." Rather, it's "at least 1 system wins." You've correctly calculated the probability of that outcome as 93.75%, as long as all the systems are truly mutually independent (which in practice is often not the case).
You don't agree with that, in what way?

The answer to your question is 1 - ((1 - 0.5) x (1 - 0.8)) = 90%. And...?
0 traders viewing now