What I mean by that, is generally speaking if it works well on the short side, should it work well on the long side? For a certain period of time I find out what the market averages per x amount of bars. For example i find out from 2001 to 2005 if you hold for 21 bars (60 min) you average -.2% (if you buy)... So you enter on the long side after a trigger and you notice that your average trade is .3%. You're like wow!!! not too bad, the mkt lost and I made .3% for a difference of .5%. So you are thinking well let's see what happens if I go the other way. When I mean go the other way is if originally you went for example when mov avg 1 crossed over mov avg 2, the other way would be defined as mov avg 1 crossed under mov avg 2. So, when you do go the other way and now are shorting the stock you are thinking if you did .5% better than the market going long then you could assume .5% better than the market going short for a total of .2% plus .5% or .7%, however what happens is you end up losing -.3%. and underperformed the market by .5% Recap 2001 to 2005 Mkt - .2% (if held long), Strategy long .3% Strategy short -.3% I tested from 2005 to 2007 and I basically get the same exact results for everything. It was a down market, and the long performed better than the short. Conclusions: In the past I have only traded strategies that work better on both the long and the short side, like if the market did .5% going long, the strategy going long would have to do better than .5% and the short would have to do better than -.5%. If both sides didn't do better than the market I would consider the strategy not robust and scrap it. so what I'm really getting at here, is could you really develop something and consider it robust if the long side does GREAT and the short side does terrible even in both bear and bull markets? and just trade the long side?