I'm still trying to get my mind around the expected goals stuff. How's this for a paradox? NYR played 3 games against CAR this season. In the 1st and 3rd, they were well underwater for 5v5 xGF% (40.22 and 41.70) and in the 2nd they cleared 50% (50.39) according to NST.
Except they won the 1st and 3rd games 2-1 and 1-0 respectively and lost the 2nd 6-1 (which was during the Rangers' January slump). So, when they had the best xGF%, they got completely blown out, while at a significant xGF% deficit they won 2 gutty defensive slogs (as I remember them) where the two teams were roughly even in high danger chances at 5v5 (8-8 and 10-11).
In the 1st game (11/2), NYR and CAR each had a PPG and Cuylle scored the only other goal at 5v5. Total SOG were 27-26 CAR. In the 3rd (3/12), Fox scored the only goal of the game at 5v5 and the Rangers were outshot 28-24. So it's not like Igor had to stand on his head and make 45 saves in either game.
The Rangers were 5-2 vs COL, CAR and DAL, three of the top five NHL teams in xGF%. That's roughly the same winning% vs that group as the rest of the league. Their 2 best games for 5v5 xGF% - by a fairly wide margin - were the two they lost, by 6-3 and 6-1 margins.
Go figure.
There is something about what makes the Rangers good at winning hockey games that is not captured by these numbers - whether or not xGF% has correlated to success. We'll see if it plays out differently in the Playoffs.