Go with the latter theory - definitely. I do not consider robustness to be a strategy's ability to work equally well on both sides of the trade. What I do consider robust is taking the ratio of the average winner over the average loser and seeing numbers like 3:1 or 4:1