The post at Skulls in the Stars about how Michael Faraday attempted to experimentally find out if electricity and gravity were related is fascinating in its own right. But it's also an example of how a good scientist bows down to the evidence rather than trying to fit the evidence to his own beliefs. In fact, Faraday found flaws in his own experiments that had otherwise supported his ideas. Contrast this to most pseudoscientists, who latch on to the flimsiest of evidence as if it were incontrovertible proof for their beliefs, often even ignoring strong evidence that their beliefs are simply wrong.
Simply having a kooky belief does not make a kook (although in Faraday's case, believing electromagnetism and gravity are somehow a unified force wasn't a kooky belief, just an unsupported one). Loudly proclaiming the kooky belief must be absolutely true despite lack of evidence or even contradictory evidence is what makes someone a kook.