I think both have their value. A quantitative approach is inherently going to miss nuances, a vote is inherently going to be skewed by rider popularity. So they are complementary to an extent.Certainly interesting but in order to test this type of formula you need to have multiple sets of data. I'd be curious to see what this would yield if you did it for say, the last 25-30 years as it'a always dangerous to come up with a formula which yields what you feel to be the correct answer for only 1 set of data. While I always find this kind of thing interesting ,for something like this award I generally prefer a subjective vote. Any formula you come up with is going to have holes and you're going to get strange winners at times.