EA Tester's result - page 4

New comment
 
Period1 Minute (M1) 2013.10.21 00:00 - 2013.10.25 22:59 (2013.10.21 - 2013.10.26)
ModelEvery tick (the most precise method based on all available least timeframes)
You are opening mid bar on the M1. Ticks generated by the tester are variable on the M1 (open to high to low to close or open to low to high to close.) So yes you will get variable results. Just like you would live, on the M1
 
WHRoeder:
You are opening mid bar on the M1. Ticks generated by the tester are variable on the M1 (open to high to low to close or open to low to high to close.) So yes you will get variable results. Just like you would live, on the M1

Sorry, I don't understand. When you say "opening mid bar" are you implying an order is being opened. If so, is it not the case the criteria to open an order was met at this juncture ie EMA fan in BUY or SELL position, RSI within range etc...
 
properbearkhunt:

It makes the difference. The test takes a matter of seconds. Run it over 25 times. Nothing is erroneous. What is the rationale in this? Could be relevant to a lot of people testing whilst connected? Is humble pie on the Christmas menu :)
Your Broker sends new information about the Symbols on your platform, look at MarketInfo() all this data gets updated, it is changing all the time, for example if you wanted to calculate a position size based on a percentage of account size you would need TickValue (or a derivative of it) and for a pair other than XXXAccountCurrency the value would be changing from tick to tick . . . when you disconnect from your Broker this data does not change so you are in a fixed environment and each test you perform is under identical conditions.
 
RaptorUK:
Probably curve fitted . . . repeat the test without Optimization on 3 or 4 different Symbols.


when I change spread 10 rather from 5, the result becomes bad; I hope give bigger stop loss to resolve this, but it is useless.

How shall I modify it? or this strategy is bad in the fact?

1234
New comment