Anything is better than the sample you have. 15 minute candles, do you honestly think going into this, IYO, this is a significant enough sample size to base anything off of? I can almost 10000% guarantee you any edge you thought you had in the less volatile days of 2000 is long gone now. There are many fx programs I can think of others and mine that deteriorated over time (plus loosing much more), this is using 1 minute data (not tick). I could build you some excellent programs in GBP JPY that paid off over 50% + year in 2000-2004. I'll give that to you for free. Guess what? It is god awful now.. but why? Maybe if I listened to Talib or some other guys random opinion about the market. As if it mattered, everyone has an opinion. I'm telling you, you are better off not even thinking about this and just doing something else. I get the feeling you will find out for yourself. And quoting others won't help your cause. You are going to find out Mr. Talib doesn't know shit when it comes to FX strategies. Let him know I said that. Good luck. You used to be me. Don't end up as me, it isn't great. Seriously though, good luck. Link your strat to the page.
If you have the programming skills, why not code systems for other traders in your spare time to make some extra cash. Then trade with your earnings. You can easy make $50 per hour coding if you know enough about trading. There are probably customers reading this very thread.