My first EA, interpreting results

[kwanann]

Sushi,

Erm, why not just forward test it for a few months and see how it goes?

[aSushiRoll]

Sushi,

Erm, why not just forward test it for a few months and see how it goes?

yes ofcourse I'm rewriting it, to forward test. It will take quite some time to forward test.

The thing is; in order to reliably assess an EA, in my opinion, you need hundreds of trades to really see how it performs in different circumstances. While a forward test on lower TF is something that i can reallisticaly do, a forward test on higher TF's will take SOO much time and it would open just a few trades, there's no point in doing that (D1/W1)

[kwanann]

Sushi,

Maybe u can use the likes of dukascopy which gives tick data to see if that is a better test, i know Empty4 backtest is as good as useless unless you do not care about the candle data (e.g you only open/close trades on candle open)

That said, conditions have always been changing, so even if the backtest is successful, it doesnt mean that it will work in forward test. How the market was behaving in 2010 is different from 2011, 2012 and 2013. That is what makes most EA fails.