I've talked about this concept in the past, but came across this post on Danny's blog here today. Larry Hite also talked about it in his Market Wizards interview:
"There are really four types of trades or bets: good bets, bad bets, winning bets and losing bets. Most people think that a losing trade was a bad bet. That is absolutely wrong. You can lose money even on a good bet. If the odds on a bet are 50/50 and the payoff is $2 versus a $1 risk, that is a good bet even if you lose. The important point is that if you do enough of those trades or bets, eventually you have to come out ahead".
What this basically means is that a trader can only control the quality of the trades he places. If your method or system for selecting trades has a positive expectancy, you will never know on entering a new trade whether it will be a winning bet or a losing bet, but in theory all bets will be good bets (i.e. you don't place trades that contradict your system rules). This applies to ALL trading styles and timeframes.
This is particularly relevant when you encounter a string of losses - when reviewed, can you say whether those trades taken were 'good bets'? Given the performance of my recent trades, I am happy that yes they did. That's all I can control.
By using sound money/risk management levels, and by eliminating the bad trades, you will be well on the way to being a successful trader. Remember that, by following the above principles, you do not even need a high percentage success rate to make a lot of money. Some of the most successful traders in history have had a success rate of 40% or even less.