It is very important to only take a position when your system tells you to do so. If you are to late, just skip the trade and wait for the next one. There will still be plenty of occasions to trade. Having fear to miss a trade is a very bad attitude as it will have as result that you will not follow your system and that's the first rule you should never break: always follow your system. It is already difficult to make money following your system, so jumping in too late at a bad price will only make things worse. Better miss that trade and get angry if you see what profit you missed by not obeying your system. That will help you to follow your system for 100% the next time.
Back-testing is a good thing to do as it enhances your confidence about your strategy. You have to make yourself more and more confident with the passage of time as it is requirement for profitable trading. I am making great profits with back-testing technique.
I think back testing is a wise idea and there is no harm in doing that. even most experienced traders do that before applying their strategies in their live account.
I trade the S&P emini futures and I backtest every potential trade for 25 trading days prior to using live. Usually the backtest will show that the potential trade requires too much criteria to be satisfied before you can enter the trade and its not effective. However, I estimate that 10% of my backtesting has produced a very good trading pattern that is profitable.
I trade multiple systems that designed some time ago and results are just fine. Never used walk forward testing. To survive into the future you need sound logic and can't rely on curve-fitted garbage. Those are guaranteed to fail. One of the systems I trade every day was designed about 5 years ago, and the equity curve has never changed. Looks just as good now as it did years ago. So yes, it is possible but not easy. Like finding a needle in a haystack
I completely agree with you: Never used walk forward testing you need sound logic can't rely on curve-fitted garbage designed about 5 years ago, and the equity curve has never changed it is possible but not easy I think sound logic is the most difficult one. I saw many people that had no clue what logic to start with. And if the logic is good, the "edge" will survive for a very long time. The basic logic that I use still survives after +20 years.
Medallion Fund from Renaissance Technologis has made tens of billions trading automated systems that are based in good part on back testing.