Topic: Using Strategy Tester as a "sanity check"
I've created many nice EAs using FSB. After generating, analyzing and refining in FSB I will do a quick "sanity check" using MT4's Strategy Tester just to make sure the EA runs within the MT4 environment. Usually the stats look okay in Strategy Tester, but at that point I'm not so interested in Strategy Tester's results -- I just want to make sure the EA trades and doesn't crash.
In Strategy Tester we can run the back test using one of 3 models: Every tick..., Control points..., or Open prices...
I don't bother with "Every tick" because it takes forever and life is too short. I usually use "Open prices" and that works well.
From time-to-time I'll have an EA that shows excellent statistics in FSB and executes >100 trades over a 6 month period using FSB back tester. When I run that EA using MT4's Strategy Tester over the same period I get *zero* trades. The Journal shows no errors or crashes. Granted the Data Source is different -- but I don't think that is the problem. At that point I'm using Metaquotes Data Source in Strategy Tester and I have also tested that Data Source using FSB's "Multi Market" tool.
For those problematical EAs I will then try "Control points" for the Data Source -- which works, and I'll see a similar number of trades as FSB and everything looks good.
The fact that "Open prices" sometimes doesn't work whereas "Control points" does, seems to be related to the 'Opening Point of the Position' indicator that I'm using -- e.g. Ross Hook.
Do you have an explanation or any insight as to why I see this behavior?