Xela, i have a question please, on the number of samples, how do u gage it in terms of duration,,, for example,,, lets take the macd on a daily chart lets say,, i used to go back 20 years daily chart and back test for all of the signals and look for optimal outcome from various sources, then i found 20 years is too far back because its not very applicable to whats happening today for that instruments, then i switched to back testing still on daily chart but for only 9 months back this time, of course there are less samples,, so the question is how do approach the above mentioned when ur doing a longer period back which will have more samples but more irrelvant ones verus a shorter period which will have less but more relevant ones thanks