Would like to run a ranking for every single stock factor, and then sort the results by the annualized return for the top bucket. That way, I could find the factors which matter most to performance. Currently, it is a crap shoot to pick the best factors to rank. For example, # of Analysts probably has very little to do with a company's performance, while the P/E has a lot to do with performance. I would like to find all the factors that matter.

Bonus: Sorting the results by how closely they fit a straight line from low to high buckets. That is, do low buckets correlate with a low performance while high buckets correlate with a high performance? Do the buckets' performance rise from low to high? Does a low rank mean a low annualized gain, while a high rank mean a high annualized gain, or are the results completely random and there is no correlation between rank and performance? Hint: You would probably use the mean of the distances between the data point of the bucket and a hypothetical line with slope of 1 going up at 45 degrees. Maybe throw in a standard deviation trick or something.