|
different # of games so it's apples and oranges, therefore a bit arbitrary which way it gets done. sounds like it's what they do in real life from above post?
you could argue both ways with some undeniable merits. a higher win pct. with fewer wins... can't assume a win or a loss in any proportion with such a limited difference in games played. whereas a higher win percentage is simply a higher win percentage, nonetheless.
i think the current way makes the most sense. it favors what has been done and locked in as opposed to what ifs and hasn't been yet. .5(w1-w2)+.5(L1-L2) = gb
once the typical variances in games played between any two teams is not a large enough proportion of total games played, it won't happen after that point. eyeballing, too tired to think about it.
|