I'd like to add to this because I teach statistical courses, and we have to keep in mind the difference between conditional probabilities. Let's say that a particular symptom occurs in cancer. If 90% of patients with this cancer have that symptom, that is NOT the same thing as saying 90% of people with this symptom have cancer. That's what makes Googling so dangerous. They tell you what the symptoms might be IF you have that disease, but is far from saying vice versa. What the stats don't tell you is that say 20% of people without cancer also have that symptom, and that particular cancer is rare.

For example let's say that 90% of Writhing Bogey patients have sore pinky fingers. However, only 1% of the population has Writhing Bogey, and 20% of the population has a sore pinky for some reason or another. So 2,000 out of 10,000 people have sore pinky, but because only 100 have the illness, and 90 of those have sore pinkies, then inputting all this data, what percent of people with a sore pinky have Writhing Bogey? Turns out 90 out of 2000, or 4.5%!! Even if 99% of WB patients have sore pinkys, the probability swells to 99/2000, still less than 5%.