Strategy Tester that unreliable?

 

I have been messing around with strategy tester the last few days, and have found that it has lots of errors. Mainly trades showing gains of 30 pips over the course of 1 minute. So is it really that unreliable?

 
einvestor:
I have been messing around with strategy tester the last few days, and have found that it has lots of errors. Mainly trades showing gains of 30 pips over the course of 1 minute. So is it really that unreliable?

yes, its really sucks. I gain over 12 million $ within 2 years test. When I looked at the results, It seems that it opened a trade and close it the same minute with the T/P that I asked.... time after time.

I am writing my own test to make sure that my strategy is working. I cannot trust MetaTrader for that

 

tick modeled

does this happen to tick modeled back test trades as well?

cuz i'm familiar with the $12 million type of profit from backtesting.

but once i turn the modeling to tick modeled, it is a never cease to loose

 
don perry:
does this happen to tick modeled back test trades as well?

cuz i'm familiar with the $12 million type of profit from backtesting.

but once i turn the modeling to tick modeled, it is a never cease to loose

Yes, it even happens when you model every tick.

 

Rassclawt,

so then, all my work i put into my EA's are basically worthless then?

 

first ea experience

ive test my first ea on test of mt4

i watch my ea loss

i configure for win

and ive test in real and its total wrong

and now i am go back to my first configuration

and now i win

the "strategy demo" test is bullshit on mt4?

sorry for my poor english

 
pillso:
ive test my first ea on test of mt4

i watch my ea loss

i configure for win

and ive test in real and its total wrong

and now i am go back to my first configuration

and now i win

the "strategy demo" test is bullshit on mt4?

sorry for my poor english

There is some possibility that your EA is well coded just to name one possible problem.

 

hi

pillso:
ive test my first ea on test of mt4

i watch my ea loss

i configure for win

and ive test in real and its total wrong

and now i am go back to my first configuration

and now i win

the "strategy demo" test is bullshit on mt4?

sorry for my poor english

try with different EA Parameter ...

===================

Forex Indicators Collection

 
elihayun:
yes, its really sucks. I gain over 12 million $ within 2 years test. When I looked at the results, It seems that it opened a trade and close it the same minute with the T/P that I asked.... time after time. I am writing my own test to make sure that my strategy is working. I cannot trust MetaTrader for that

Are you sure what you describe is not a result of the shitty data that Metaquotes supplies??? This garbage pricedata is full of instances where the close/open of adjacent bars can differ by as much as 100 pips, which would trigger trades exactly as you describe...

I have been using the backtester for a few years and I don't recall coming across a single instance of where the backtester is at fault! MQL has bugs however, which is not the same thing... Until someone can show me a verifiable instance of where the backtester fcuks up the backtest results seriously, I will continue to believe it is either the data or the programming...

PS. - Just decided to throw together this EA to show how 'suspect' this pricedata really is. Just backtest it and it will print out the number of price gaps between the Close & Open of adjacent bars (the pips can be set by the user). Remember, this number is the number of potential false trades that your EA may generate, not because the backtester is at fault, but because of the pricedata! It should also illustrate why scalper backtest results are a complete waste of time...

Files:
 
einvestor:
I have been messing around with strategy tester the last few days, and have found that it has lots of errors. Mainly trades showing gains of 30 pips over the course of 1 minute. So is it really that unreliable?

Backtesting is not reliable. Forward testing is really the only way to go. I use backtesting on EA's for one reason. To make sure it sells and buys. Also to see if TP and SL are working. Its very hard to achieve more than 90% modeling quality in strategy tester. Look at it this way. I'm giving you directions to get to my house. You must turn down 10 different streets to get to my house. I will give you 9 streets out of 10. Do you think you could find my house given what I told you?

 

I was using the Strategy Tester feature of MT4 platform to test the results of an EA.

I tried to save the Results so that I can print later to study the results, but have not been successful.

Can anyone help? Thanks