What I'm about to say is pretty basic stuff, but it's always good to remind oneself about it.

The only information available to most of us about a project/company is the information that is on the internet. Based on the information that the team released in different forms (product documentation, team bio, interaction with the community through media, ...) you can draw some simple conclusions and distinguish the "complete trash ICOs" from "there might be something behind this".

Now the next step is to evaluate the quality of given information and to form some criteria which will help you decide if it's something "you want to bet on" or it's a pass

Quality of information and criteria examples:
Is the information about the team available and viable? Where is it available (there are some "industry standards", i.e. LinkedIn)? Does the team experience cover areas necessary for success?
Do I understand their whitepaper and are their goals clear? If not, can I find more valuable info about it? What are the obstacles, what are other people saying? And so on.

Sometimes you have to go through a lot of crap information, but with time and experience people learn to separate what's important and what's a distraction.

Making a plus minus list with your criteria (or google someone elses criteria) can be a simple and organized way to make a decision.
Hope I helped you a little.