Originally Posted by Shinigami

I can tell you whats wrong.
I can show you a backtest of 1,5 years with initial depo of 10,000 dollars which has profit/loss of 1,500,000 dollars
Simple breakout strategy, money management, highest drawdown is 35%.
Can you tell me whats wrong with my backtest?
I can show you another backtest of yet another system of mine, which ends up with 51,000,000 dollars. But I don't use it. After the back-test ends, it starts losing - I got it over-optimized.
If your backtest isn't over-optimized, be happy and move to forward-test and finally to real account.
Backtest isn't your goal, its just one step towards the goal. If you are happy with it, move on to the next step. And do so till you get so rich you want to LOSE money on forex

Originally Posted by forexx

hi guys plzzzzzzzzzzzzzzzzzzzz help me
make the EA open .1 and .2 and .4 not .15 or .27 for example
i mean my trading company does not allowed microlots
and make it always start with .1 (no need to calclate 5%.........)
plz any one????

Originally Posted by Shinigami

I can show you another backtest of yet another system of mine, which ends up with 51,000,000 dollars. But I don't use it. After the back-test ends, it starts losing - I got it over-optimized.
If your backtest isn't over-optimized, be happy and move to forward-test and finally to real account.
Backtest isn't your goal, its just one step towards the goal. If you are happy with it, move on to the next step. And do so till you get so rich you want to LOSE money on forex

Originally Posted by Minnelli

Over optimized is when you use variables that work perfect for a certain time period, but stink in another time period. You might use Dlong=14 Dshort=11 TPLong=18 and TPShort=28 -- just as an example and maybe it turns a 1000 into a million in a month. The problem is the numbers were curve fit to give the best result. Chances are the future won't work with these numbers.

I've been testing v1v2 from 1/1/2007 until 3/31/2007 and get the results I like. Then I run the same variables from 4/1/2007 until today and see if they hold up. The best variables in Q1 don't always work in the next period. So the best is to find something that works for a longer time period. For me a decent number has been bands of 20 with TP of 30, but I will publish more results later.

Minnelli