Just another note regarding points vs analytics, if that's what the contention is for some.

It's a false dichotomy as analytics are simply studying game events that contribute to the generation of points averaged out on a more consistent, normalized manner.

The Points Pipeline:

Goals <-- Shots on Net <-- Shot Attempts <-- Puck Possession in OZ <-- Puck Possession in NZ <-- DZ exits

At this point in the game, there aren't any publicly available League generated stats beyond shot attempts, so that is the best proxy for all the events leading up to it.

As anyone who has ever counted coin flips before can tell you, it takes a good number of flips to get that 50/50 split between heads and tails, so it takes a larger sample size of events to get stable numbers (true performance). In the points pipeline, the sample size of goals is minuscule per game compared to everything that comes before it (2 goals off 28 shots on net from 78 shot attempts for example), so it is not a stable measure until you get a quarter to a full seasons worth of games. If you're measuring just an individual player, it's a season+ because while a team may be on the ice for a full game, no individual player is.

And if you're a good coach, or manager, and you understand the above, then you know the focus for each individual player has to be on maximizing what he himself can control, and eventually the results will emerge consistently.