The only way to win with a low winning percentage is to trade many, MANY non-correlated financial instruments at the same time. That's how the famous Turtles won millions in the Futures market. For example if you trade 20 different commodities, you will ALWAYS catch a mega trend in a couple of commodities (coffee and heating oil for example) in any given year, with a simple moving average or a breakout system. In other words the big winners will allow you to psychologically stay in the game.
Keep in mind that the risk/reward ratio is just an average. On a trade per trade basis this ratio will of course vary. I could risk $100 to make $100 on Monday, but risk $100 to make $450 the next day, for instance, depending on each trade setup.
R:R is another example that conventionally it should be Risk:Return, but I have used it as Return:Risk several times on ET. Following this kind of conventions/ frameworks/ boundaries/ free-advices-on-ET (like this post of mine) without critically analyzing them is Not an edge! http://www.elitetrader.com/vb/showthread.php?s=&threadid=279873
I agree, but even return:risk is of doubtful utility. As the trader has no idea what his reward/return is going to be and as the market couldn't care less how much he's willing to risk, I don't know why anyone would bother calculating something so insubstantial, particularly if doing it in advance as a component of a trade decision. The challenge for the trader is to take however much the market is willing to give him, preferably with gratitude, not grab the money out of its hand, slap it across the face, and slam the door. One should look for legitimate reasons to stay in the trade, not get out of it.
Rumor has it (and I have no idea how true this is, just commenting on the "conventional wisdom") that trend following strategies tend to have low winrates and counter-trend strategies tend to have high winrates. If true, perhaps this is an argument to stick to counter-trend strategies or to automate tf strategies as much as possible..