First of all, crushes can get to the state where it is really hard to think rationally about them (Love and Limerence by Dorthy Tennov includes several example of irrationality). So I recommend trying to think rationally, but always consider the possibility that you are completely or partially wrong.
Second, realize that your own estimate of their attraction to you may very well be in the statistical noise range. From "Benefit or burden? Attraction in cross-sex friendship" by Bleske-Rechek et al. your estimates of their attraction is much more correlated with how much you are attracted to them then they are attracted to you. "Friends’ reports of attraction to each other were weakly, but not significantly, correlated, r(87) 1⁄4 .19, p 1⁄4 .09. However, men’s attraction to their female friend was strongly associated with their estimate of how attracted their friend was to them, r(87) 1⁄4 .61, p < .001, and women’s attraction to their male friend was strongly associated with their estimate of how attracted their friend was to them, r(87) 1⁄4 .49, p < .001."
Third, remember Conservation of Expected Evidence. For anything that you consider a signal, consider if you would consider the opposite happening to be an opposite signal.
Fourth, I expect that if you were rational about it, you would end up with something like P(they have a crush on me) = .2 or something like that. Eventually, if you want to increase your certainty, you have to actually do something where their response actually gives you good evidence. Unfortunately, these do have the chance of rejection. But I recommend leaving a way out when you do this. I have had good friendships result from romantic rejection.
Lastly, be a friend to them. Most long term romantic relationships are also friendships. It is a lot more fun to think about what interesting thing you want to tell them, or waiting to see what they bring up in conversation, than obsessing over a few glances.
If you have a crush on someone you usually want to find out if they have one on you too. In my opinion outright asking them is often not a good solution, because if they don't have a crush on you yet it decreases the chance of this ever happening if they know you have one. This believe is based on what I read about love psychology. Hovever I don't really want to discuss the option of outright asking them in this thread, therefore I have not elaborated further how I got to this believe.
The alternative to asking them is trying to interpret signals that they might give you. However to know how many signals you need before you should believe that they are in love with you, you would need the prior. I have not been able to find anything about the prior of someone being in love with you. Therefore my Idea is to do a survey in order to find out how likely it is that a person you know has a crush on you. The plan is to ask the person taking the survey how many people they know well enough to possibly have a crush on them and how many people they actually have a crush on.
I have created a Survey for this and would be really happy if you would participate.
The next stepp would be to discuss how certain signals a person can give you raise the probability of them having a crush on you. That part is quite difficult. I think probably the best way would be to check how your friends react to certain situations and what body language they show you and then, if you find out someone has a crush on you, to look up what he did differently from people who are merely your friends. I am currently not in a good position to do this experiment but if someone wants to try or has results about this to share please do so. However I think this part is less important than finding the prior, because most people have at least a general idea about what certain signals mean from personal experience while at least I have no idea at all what the prior might be.