(a) 5v5 shot attempts for and against
(b) PP shot attempts for
(c) PK shot attempts against.
Shot attempts have been shown to be the best predictors of future performance for each of these situations.
MSM uses wins and losses in assessing the strength of teams generally, as well as focusing on goals in evaluating PP and PK strength. Hence, you will see that my power ranking will deviate significantly from someone like Pierre LeBrun of ESPN. He has Montreal ranked 1st! That's just wrong. It's like ranking Colorado first (or any where near first) last season.
My weighted values are something I derived using multiple regression analysis (too complicated to explain here). I have not seen this kind of analysis anywhere, so my weighting is probably unique. There is no intuitive judgement on this ranking, or noting whether teams are streaking (for better or for worse), or anything subjective. The ranking is purely statistical.
If you notice that a team is ranked higher or lower than they are in the standings (e.g., Oilers are ranked 14th in the Power Rankings, but 28th in the standings), you may think the ranking system is flawed. But if you look at the team's PDO, you'll see why a team is underachieving or overachieving.
In the Oilers case, you'll see that their PDO is way below 100 (i.e., 97.41). In other words, "bad luck" is affecting the team. Specifically, we have a very poor even strength save% of 90.13, which is 2nd worst in the league. My belief is that if we had an average save% (92%), we would be contending for a playoff spot. This doesn't mean I believe we have bad goal tending, but rather, that our goaltenders have had bad luck. I hope to see this save% improve by game 41.
The Top 10 in my Power Rankings
4. NY Islanders
7. San Jose
9. St. Louis
10. Tampa Bay
Full Rankings are here or you can find the link on the right hand side labelled "Walt's NHL Power Rankings 2014/15".