It certainly is not easy to build an EA but I´m not sure that it has to be ´rocket science´ either. I purchased an EA writing program and was writing EAs the next day. Not that those were money-making EAs but I used them to test trading ideas and to become familiar with the pitfalls of EA writing and execution and individual broker platforms. While doing this you will also become more familiar with how the forex markets behave and I don´t think that you can really be a successful EA writer without having that familiarity.
After months of testing I have an EA system that I am happy with. I run the same EA twice against each currency pair, one in default mode (where a buy/sell signal results in a buy/sell order) and the other in reverse mode (selling on a buy signal, buying on a sell signal). Each EA is optimized separately so it´s not a true hedging system but it is one which I hope will prove robust enough to survive changing market conditions. A kind of ´cover your butt´ system. I mention this not to promote this particular ´solution´ but just to encourage folks to try different approaches, other than trying to write the perfect, good-for-all-markets, directional EA - because that is going to be a tough task!