Disliked{quote} If the trade count of each system is the same and of a statistically significant amount, then I am not so certain that a daily system that traded for 10 years is much more curve fitted than the system that traded for 2 months on M15. The thought is that that the system only needs to see a statistically relevant number of trades over a given test window, whatever that window may be (I don't know this to be true yet). All this keeps leading me to question how many trades are a statistically significant sample size.Ignored
I would put my confidence in a system that generates a net profit from 10 trades over 10 years than 1000 trades in 1 year of M15 data.