If you found a system that works on daily time frame and it's not a fluke then it should work on 4 hourly time frame.
I'm paper trading before committing any real $$. I am not a tech-savvy person so I will forward-test the strategy to see if it works. I'm the typical retail trader everyone talks about. I don't want to assume that just because I make money with it on the daily time-frame, that I will also succeed in the 4hr (more noise, more fees).
Got me a special little column in my handy dandy error log Titled just for those special little "Made Up Trades" that try to sprout like freakin' weeds all over the money patch. Little Bastards, lol.
Are you using a Purely Systematic method to determine your entry stop target exit prices and position sizes? eg. "written" instructions that are executed for each and every occurance of requirements for a "Trade". They can be written for digital execution if you can, or they can be written with a pencil on paper, either way, can your system be reduced to writing?
I'm glad some of the first responses talk about back and forward testing. It's absolutely critical. I would only add that you should spend the painstaking hours verifying the back testing results trade by trade. It sucks, and it takes a lot of time, but you need every bit of confidence you can find to bid or offer without any hesitation. Trade Ideas is notorious for overstating back test results. Consider yourself an unbiased tight assed auditor and verify them. A 90% win rate on TI is more like 60%. 99% is more like 75%. I've said it before, but there is an empty bottle of Cabernet on my desk as a reminder of how many times my friends were out partying, and I was at my desk verifying test results on a Saturday night. Good trading, my friend.
50 trades is not a large enough sample. I once, backtested an extremely profitable strategy, I though, I went thru 118 trades. Thought it must work somehow. With the huge slippage in prices in actual trading, that was enough to render the trading strategy, a huge failure. You probably, need atleast, 500 trades.