Quote:
Originally Posted by mmbt0ne
So I'm not of the mind that I can have a single best lineup for any single week. I expect to end up with a handful of highly correlated lineups that I am not comfortable saying "this one is clearly the best." Maybe that's just me, but that's the starting point.
That's a fatally-flawed starting point. One will be best even if you can't distinguish it with your eye test or whatever you are using (or maybe > 1 lineups have equal theoretical medians, in which case they would draw as n --> ∞).