My central tenet for justified belief is basically this: that the evidence we have (e.g. sensory or memory evidence) is a reasonable basis for belief.

This isn't because I think we can argue that the evidence is true, but because we don't have an alternate basis for logically interacting with the world. Our evidence is singular, and we can either accept it with some degree of doubt or we have no basis whatsoever. If we were to accept a skeptical hypothesis instead, then we would have to logically conclude that we have no evidence of the external world and no means of logically interacting with it. I don't know that my evidence is true or accurate (and in fact have good reason to think that at least some of it is untrue), but it's more reasonable for me to accept it ceteris paribus than it is to reject it.

For definitions of knowledge, I would recommend looking up knowledge in a philosophical source, such as the Stanford Encyclopedia of Philosophy. JTB is far from the only definition of knowledge, but it's the core of externalist conceptions of knowledge, which are generally more popular than internalist ones (which have their own issues, such as lack of grounding to reality or the possibility of false knowledge). I stuck with JTB because it's the simplest version and I didn't want to devote 50 pages to different forms of knowledge in this paper.

The clock problem itself comes from Russell's "Human Knowledge" and has been discussed fairly extensively as a proto-Gettier problem, largely as a criticism of JTB.