The problem is at identifying all variables that drive the market. In your example you are including a few variables, but those are not all available variables. If you make an effort to identify the drivers of that market you might be in a fair position. I am sure that you will find many more variables to include in your entry criteria. One variable that you are not including as a first attempt is trade size. Look at this video on how trade size will change your success rate. Another driver is set by market makers. How do they set the pace of the market to achieve accumulation and distribution. Another one is the correlation to other markets, every market is correlated one way or another with other markets. So you'll have to include those correlations as well. You can see how you will easily extend your entry criteria with little research.
Interesting discussion, I think some of you give up too easily. It's a subject where the complexity can be as large as you want to go, you can spend weeks on figuring out the safe way to create and test a system without overfitting (that I wouldn't want to fill up this post with). Just one little thing I think I can mention to help your confidence - the trading Universe. The larger your universe, and the more data included in that which has not been used in strategy development or optimisation, the less amount of overfitting you have done. Eg let's say you have developed a system using observations on the behaviour of stocks in the Nasdaq 100. Then you test it on the Canadian stock market and find it still works, test it on the Russell stocks and it still works, test it on dates that you never used in development and it still works. Change the lookback and it still works. Then there's a good chance it will be better than not having a system. A better chance than trading a system that did NOT do well in these tests. I developed a short selling system using the US market, I tried it on Canadian stocks it didn't work. I tried it on ASX stocks it broke even. Then ran it forward on US stocks and it started losing money. It was the only system I didn't immediately throw out for failing other markets, but yes I should have thrown it out like the others. The long mean reversion and trend following systems that I run on 3 stocks markets did work though. Even though some markets are better than others for a particular system, it's robust if it works on all 3 and allows the markets to change and the lagging performance on one to pickup where another is down. When I say it works on other markets - it doesn't have to be a superstar, just has to make money and not be hopeless. I'm fairly green at systematic trading myself, but my limited time has shown me there's great opportunity. Not to get rich overnight, but an opportunity to create diversified systems that complement each other and to have a smooth consistent ride through the trading journey instead of the massive swings in equity that discretionary traders experience.