On Sun, Apr 15, 2012 at 8:01 AM, Tim McNamara <paperl...@timmcnamara.co.nz> wrote: > One of the good reasons for this format (although I have no idea if it > is why it's why it was chosen) is that there are some good statistics > behind pairwise comparison.
The main problem with that format is comparing apples with oranges. features should be compared at the same level of usage and eventually complexity not just because B come after A or a random has been applied. - benoît
