A few comments from my experience trading forex (not advice, just observations).
1. Have a plan. Know before you check the markets for the week/day/hour/minute (depending on your trading frequency) what would make you to go long, what would make you go short, what would make you close a position, what would make you stay out altogether.
2. Your plan needs to include some means of setting stops. Guessing or "she'll be right" isn't really a good idea IMO.
3. You need to know that the method is actually profitable unless you really do like sitting up at all hours worrying about your account being wiped out. TEST IT!
4. You need to know the maximum drawdown that the method is expected to produce. Position sizing (amount of leverage) needs to be CALCULATED based on this. Don't guess when it comes to leverage.
5. If you are going to limit yourself to a reasonable (say, 20%) maximum drawdown then I would be very surprised if you are going to be able to use leverage greater than 5:1 and most probably the leverage will need to be less than this. Want to use higher leverage? Back test to see what happens when you have a bad run and you'll probably decide that very high leverage isn't such a good idea after all... There are trading systems that work with high leverage though, you just need to be SURE before you start leveraging your trades.
6. 100:1 leverage WILL blow your account sooner or later. Probably just after you think it's starting to go well...
7. Keep away from dodgy brokers. This is far more important with forex than with stocks. Be particularly wary of overseas-based market makers that rely on traders' losses, rather than simply collecting the spread, to be profitable. I probably shouldn't mention names here but there's a real problem when your broker WANTS you too lose and isn't regulated by anyone.
8. The cheapest broker is not necessarily the best.
9. Don't forget to include buy/sell spread when designing and testing trading systems.
10. Just because a system works really well long or short DOES NOT mean that it works in reverse. Logic might suggest that it ought to, but that's not necessarily the case in practice. You can lose a lot of $ if you assume that it works both ways without testing it.
Just a few thoughts. It's certainly not a comprehensive list though.