Hockey is full of statistics that are basically worthless when you examine them closely.
"Hits" are meaningless because the stat is so weakly defined and so open to interpretation. Teams frequently have vastly different hitting totals at home vs. away, which is a strong indication of biased or overenthusiastic hit-trackers at home. Without any sort of consistency the "hit" stat has no value at all.
One-goal wins are meaningless because they are no better than a coin-flip. Winning a game by one goal is essentially the same as losing that game.
A team's powerplay percentage is meaningless because it has little correlation with team success. The 2nd-worst team in the league (Edmonton) has the third-best powerplay, while one of the NHL's top teams (St. Louis) has the 26th-ranked powerplay.
Don't even get me started on the penalty kill.
+/- is meaningless because it is affected by so many contradictory and irrelevant factors. If you make an awful mistake but your goalie bails you out you don't get a minus. If you hop on the ice on a line-change one second before your team scores you get a plus. It's madness.
Assists are meaningless because they are so random and arbitrary. Secondary assists are among the most worthless things ever invented by man.
"Goalie wins" are meaningless because an individual goalie usually has so little effect on the final score of a game. Once you add in the strange things that can happen when goalies get pulled mid-game the whole statistic becomes laughably illogical.
Goals Against Average is meaningless because if a goalie plays behind a bad team they are going to see a great deal more shots per night than a goalie on a good team, and this will inevitably lead to a higher GAA over the long-run, so GAA tells you nothing about the quality of a goalie.
Shots are meaningless because the only thing that really matters is goals.
Goals are meaningless too.
Everything is nothing, including the consciousness of nothing.
Prediction: Sharks win 3-1, but it doesn't matter.