Measuring reliability of an entry method

I’m currently programming and backtesting some basic mechanical entry systems.

I would like to get your thoughts on how to accurately measure the reliability of different entry methods.

My initial thought was to only evaluate the winning % if TP and SL are equal e.g. how often will price move 20 pips in the forecasted direction rather than against it. The problem is what number of pips to work with?

Another option is to see if price has moved in the predicted direction after x number of bars.

Any advice on the above will be appreciated.

If you have a 100% mechanical entry system you would like me to program and test please have a look at this thread: http://www.forexfactory.com/showthread.php?t=275628
So far it seems that achieving 55% winning trades with equal SL and TP is not easy at all, especially for systems that trade frequently.

The above link might give you some ideas for a starting point.

This is the actual link:

http://www.forexfactory.com/showthread.php?t=223071

The edge ratio looks like simply MFE / MAE over a preset period of time, where MFE is maximum favorable excursion and MAE is maximum adverse excursion.

John Sweeney's two books explore this subject in depth: Campaign Trading: Tactics and Strategies to Exploit the Markets (1996) and Maximum Adverse Excursion: Analyzing Price Fluctuations for Trading Management (1997). I own and recommend the first one.

Thomas Stridsman also devotes considerable time to these concepts throughout his Trading Systems that Work: Building and Evaluating Effective Trading Systems (2000), essential reading for any mechanical system developer or trader.