In fact you may test two expert models:
Tick oriented expert does some major operations every tick. So because we have
no real tick data the same interpolation method using to recover them. The quality
of testing depend on an accuracy of interpolation.
Bar oriented expert ignores all ticks exept one,opening a new bar. The testing
based on real existing data. So it more reliable.
static datetime baropentime;
if(baropentime==Time) // Does new bar jast open ?
return(0); // No, Wait the following tick
baropentime=Time; // Here You are at last
GoAhead(); // Go