Traders who are successful over the long-term have clearly defined their 'edge' in the market or markets they trade. To me, an edge is basically a method that statistically generates a positive expectancy over a large sample of trades.
In order to capitalise on that edge, the trader needs to develop the ability to use that edge. This basically means that a trading plan which is constructed to exploit that edge is followed as closely as possible.
Any difference between the theoretical returns and the actual results generated I term as being an 'expectancy gap'. This normally arises where traders have a tendency to override or second guess their chosen rules.
Your rules should be designed to capitalise on your edge, whatever that may be. So if you don't follow your rules, by definition you are not trading to capitalise on your edge.
Where's the sense in that?
The bigger that expectancy gap, the higher the possibility that a trader will turn a winning method into a losing one.
Your edge is particular to you. What may give you an edge may be anathema to someone else. This may well come back to the beliefs that you may have about trading and the markets, which may directly conflict with another trader's beliefs and their own successful method.
For example, because of my own beliefs, I find it impossible to buy at support or go short at resistance, or even buy on pullbacks or retracements.
That is not me saying that approaches or methods designed to do those things cannot work - because they clearly do, in the right hands.
However, they do not work for me because of the conflicts those types of approach have with my own beliefs.
Finally, remember that having an edge does not just not mean focusing on entries and exits. An integral part of any winning method is having strong risk control.
You can have a method of identifying entries and exists which potentially has a positive expectancy, but having poor risk control could mean being unable to survive a run of losses, which would fall into the normal distribution of trades. So again, a winning method could be turned into a losing one - not by overriding the signals, but simply by taking excessive risk.