The series win probabilities have been posted on our projections page now, but it's worth explaining where they come from. The first thing to remember is that away goals are now the first tie-breaker after each home-and-home series, followed by the usual Extra Time, Coin Flips, and Most Chest Hair tie-breakers if needed. That plays a role in the probabilities!
Playing at home with a 4 - 2 lead, New England is in the best position of any team. New England just needs to lose by no more than one goal. In fact, even if the Revs lose by two, they would win most plausible tie-breakers as Columbus would have to score at least four goals.
According to our model, New York is in the next-best position due to two primary factors. D.C. United has turned in very unimpressive xGoals tallies this season, and, of course, the Red Bulls hold a two-goal advantage.
SEA: 68.5%, LA: 64.5%
Though both top seeds out West tied on the road, Seattle gets a probabilistic edge over LA in that Seattle scored a goal in Frisco, while the Galaxy failed to score in Rio Tinto. Seattle would advance with a 0 - 0 draw, and play extra time with a 1 - 1 draw in regulation. But if LA draws, it must be a 0 - 0 draw to even get a chance to win in extra time or penalties.