Greetings quick question on the accuracy of the strategy tester on MT4. I am testing a strategy at the moment but the testing time takes too long 1:30Hrs for each pair. I decided instead to store the data and prices in an array and run through the data “offline” to speed up the process.
The testing now takes 2 minutes for the same period but I am getting slightly different results to the strategy tester. I have looked the EA over and again and can’t find the mistake. How reliable is the strategy tester - I have had some feedback in the past that it was error prone.
Hi Bob, cheers for the reply. Yes I am using TickData Suite. The EA is based on each tick and the count win or lose is based on the tick price movement to the TP & SL.
What’s odd is that is where I would expect there to be a mistake but going offline suddenly creates 7 more trades than on the Strategy Tester
I received some advice once about a bot slowing down strategy tester. If your program runs off each tick it will bog it right down. The bots I was testing initially did a pass in 10 mins but each subsequent pass increased by 10+%. So it didn’t take long for passes to take over an hour. This was due to the on tick function. It was suggested that I re-engineer the bot and remove on-tick functions where it wasn’t necessary an instead use an on new bar command. So that’s what I did. My tests then passed in under 2min each and every time.
My suggestion is have a look at your program and see if you can re-engineer it. Apart from that maybe head to mql5 forum and ask there. I believe you’ll get a better response from people far more skilled than myself.
