A.A large number of people think there is some good reason to give "1" ratings to everyone, and therefore a straight average simply isn't appropriate. The method we use for nullifying the effects of these waves of "1"s is making the overall rating base for each profile on a root mean square algorithm. To put it simply, we compute the square root of the averages of the squares of all your combined individual ratings. A simple example would go something like this: Somebody gives you a 2 rating, and somebody else gives you a 10 rating. A 6 ranking would be the result and by using our method, instead of using a straight overall average, you get a 7.2 rating by squaring 2 and squaring 10 (4 and 100, respectively), and then averaging the results. It would give a final result of 52, which upon extracting the square root, gives you a final rating of 7.2, which for obvious reasons is weighted more closely to the intent of the more reasonable voter.

SmokeyBear's take on that is, "If you can't give the individual a decent rating, then just hit the "Skip this image" link, and rate someone you can. On the wrong morning you might be a 1, but you wouldn't want to be told that!