Back testing warning

 

A lot of programmer are wasting time and effort setting up strategies with back testing, often these tests give different results time by time. The must important thing is that the history chart have to be error free and the quality model must not to be "n/a" but almost "90%". I suggest to test the chart quickly compiling an ea that do absolutely nothing, a simply template, this allow a rapid history test that is better than wait several minutes, or sometime hours, testing the own ea and then discover that the model is not correct. It is important that the tester is able to use M1 timeframe. Once tested the chart in few second you can start to test your ea and relax.