This is not, however, a rant about the horror of amateur family historians or an attempt to discourage you from starting your family tree. Quite the opposite. I remember the thrill of clicking and adding relatives when I first started. There were no shaking leaves then but still plenty of information that was easy to find. If it had been more difficult, I might have gotten frustrated and given up. I have seen my students experience that same thrill. I admit that I am not a purist. I jump around from branch to branch and work on whatever part of my tree captures my attention in that moment. I do not work methodically on a single line until it is done.
However, I have learned to temper my enthusiasm with a healthy dose of skepticism about what I find online. I use trees of other members as starting points for research and look for records to back up what is recorded. I have made my trees private after being ripped into several times by those who believe no information should be recorded until it is completely verified.
Finding ancestors is like being a private detective. You have to follow the clues and collect enough evidence to make your case solid. You will receive lots of tips and most of them turn out to be irrelevant and misleading but with persistance, you can find reliable information and solve the mystery. Sometimes it's a preponderance of circumstantial evidence and that's the best you can do.
Assembling ancestors into a family tree is like putting together a jigsaw puzzle. Sometimes a piece seems to fit but later you notice the subtle clues that show it isn't quite right and you begin again to find the piece that goes in that spot. Each piece you correctly identify and fit into place goes into making the larger family picture come to life.
I do make errors but even when I have to lop off a branch from the tree because I've made a mistake, I always learn something and often find interesting stories of people and history. So get in there and start looking. Use the easy hints to get some traction but go back and evaluate with more care. I suggest using two trees - one for leads and possibilities and one that is as carefully documented as you can make it.