awe. Expectancy is important as unless a system has a positive one it will lose money however you touch on an interesting point.
Potentially a system with 30% win rate can perform as well as a one with 80% hit rate as long as the first has much larger winners than losers and the latter can take smaller wins than loses.
Some peoples psychology prefer home runs and others prefer consistancy...
The following simulator will let you play with expectancy
http://www.hquotes.com/tradehard/simulator.html
Whilst I don't disagree that expectancy is of overriding importance, I also don't think it pays to get too hung up about it. All that trading with positive expectancy means is that, over time, you're going to make money. How it happens - well, that's another story.
Another aspect you need to consider is profit to max. drawdown as a ratio. Let's say you're targeting 50 pct return in a year, what kind of drawdown are you going to tolerate to achieve that? 50 pct in a year is a bloody good return, and most mechanical systems will have a ratio of 1, i.e. you're going to get a drawdown of 50 pct if you target annual returns of 50 pct.
Stop and consider for a minute what a 50 pct drawdown will feel like. Most (if not all) people overestimate their tolerance for drawdown, and halfway into a 50 pct drawdown, will start to either change their strategy or simply abandon trading.
Discretionary trading suffers from this aspect in that it's impossible to backtest - who knows what you WOULD have done. But whether you're a discretionary or mechanical trader, drawdown is unpleasant and herein lies the value of BELIEVING your system has positive expectancy - you're going to stick with it, during the sh*t, which is what you should do.
If you're mechanical/systematic, you can backtest to know if you have +ve expectancy.
If you're discretionary, you will only know that you have +ve expectancy from trading for many years and knowing that you CAN make money.
Either way, you need to have the belief and discipline to stick with what you know SHOULD work, whilst at the same time recognising that markets occasionally go through "paradigm shifts" and the style needs to be altered.