Here is some information and a brief video: http://ninjatrader.com/support/helpGuides/nt7/?backtest_a_strategy.htm
Forward testing is slow...one day per day. Backtesting is usually much much much faster. I backtest. Then take the best model as determined by backtesting; then forward test. There is nothing "ungrippy" about backtesting. I suspect you simply don't understand what everyone else is doing, and why, and how. And as a result, you've assumed and summarily concluded that we all need to get a grip somehow. Funny.
http://www.investopedia.com/articles/trading/05/030205.asp I'm not here to convince anyone to backtest. I'd rather you put your money into the pot, with untested strategies. But I did provide the above link for you. I doubt it tells you something you don't already know. As you are the one making the extraordinary claims, the onus is really upon you to show why forward testing ALONE is better than doing both backtesting and forward testing.
If you backtest you can test for example 1000 trades. If you want to do that in forward testing it will take you months. And maybe to find out then that your plan sucks. Developping a tradingsystem takes a lot of time and a lot of creating, analyzing, changing parameters, throwing away some parts or replacing them..... In forward testing you can do each day one test, in backtesting you can go thru the same data with different setups over and over again. If it is not obvious that backtesting is better for a start, it is maybe not a good idea to start trading, as you need certain qualifications, of which you are missing already some as it appears.
Dudes, your psycho is not in play when it comes to back testing...meaningless. God forbade it takes time...doh, believe we need to wait for good set ups. Also, yesterday's market is no longer today's. Don't you folks get this??? In forex, the sentiment changes as soon as Europe begins to release data and changes if the U.S. beats them later in the day. How do you play this information...please, remove your stop losses when data comes out...this is how central banks make their money.
Psycho is only in play if the rules are not hard defined. If the system uses mathematical rules it is clear what to do, no matter what your psycho is. If you have no fixed rules every rerun will give different results, making testing worthless, because each time entries and exits will be different and the outcome too as a result of ever changing "rules".