We skeptics love to quote “Occam’s Razor” (i.e., that entities should not be multiplied unnecessarily, or the simplist solution is usually the correct one).
Occam’s Razor means that the strange lights in the sky are NOT likely to be interstellar craft piloted by little green men, since that explanation requires many additional assumptions and there are other much easier explanations.
Similarly, Occam’s Razor means that crop circles are likely NOT caused by “plasma vortices” or the aformentioned little green men, since, once again, those explanations require many additional assumptions whereas the theory that crop circles are made by pranksters with ropes and boards doesn’t.
My question is whether anybody has applied the scientific method to Occam’s Razor itself to test whether it is actually a valid principle or not? Did William of Occam come up with this principle after careful observation, or was it just something he came up with off the top of his head? Has anybody studied large sets of data to determine whether the simplest explanation actually tends to be the correct one more often than not? And, if so, how often does the simplest explanation turn out to be the correct one (in other words, does “more often than not” mean 55% of the time, or close to 100%, or what)?
In short, how do we know that Occam’s Razor is a valid principle, and to what degree are there exceptions to the general rule?