For full disclosure, I am not as well versed in advanced stats for hockey as I am for baseball and basketball, so if I omit something or explain something either incorrectly or confusingly, let me know.
There are many ways to judge hockey teams and players. The most obvious way to judge teams is by wins. It’s fine to say that wins is the most important stat, and it’s the truth, but it is even more important to figure out what makes teams win. People have started using goal differential as a way to explain this, but goals in and of themselves are pretty luck induced events. Everytime you see a puck take a weird bounce off the end boards and go right into the slot, or see an own goal, think about that. Similarly, goal difference places a huge amount of importance on how a goalie plays. Having 1/6th of the players on the ice influencing a large portion of whether or not a goal is scored, and then assigning some sort of true value to that, is a very inefficient way of doing things.
A much better way is to look at shot totals. And specifically, shot totals at even strength when the score is tied. Why even strength is obvious: over the course of the game, the team with more PP time will likely get off more shots. Also the team with the better power play/weaker penalty kill will get off more shots/surrender more shots. And as for score tied, score effects lead to the team that is behind getting off more shots, either due to the leading team being in a bit of a shell or the trailing team taking more chances.
The same works for players. Goals and assists are nice baselines to measure guys, but they need some context: who are they playing with? Who are they playing against? etc. Here then is a quick rundown of things you need to know, hopefully in a semi logical order. (note: all of these stats can be found at the awesome site Behind The Net):
5v5 QoC (or QualComp): Basically a measure of how good a players’ opponents are. League average last year was around -.001, so I’ll just use 0 as the average. The higher the positive number, the better/tougher the competition.
5v5 QoT: The inverse, it measures how good a player’s linemates are. League average last year was -.017, so again I’ll probably be lazy and use 0 for an average. The higher the number, the better the linemates.
5v5 G/60, 5v4 G/60, 5v5 A/60, 5v4 A/60: Pretty self explanatory, these are the scoring and assists per 60 minutes either at even strength or on the power play. Note: the assists are only primary assists.
5v5 CrsOn, 5v5 Crs Rel: This is tough to follow, but bear with me: 5v5 Corsi On has an intimidating name, but all it does is is take the difference in shots for and shots against when Player X is on the ice. 5v5 Corsi Rel. takes the difference in shots for and shots against when Player X is off the ice and then subtracts that from 5v5 Corsi On. Basically this measures a player’s influence on shot totals, and is a pretty important stat to know.
5v5 TOI/60, 5v4 TOI/60, 4v5 TOI/60: Contextual stats that not only provide context for stats like goals per 60 minutes (if a guy scores 1 goal per 60 minutes but barely plays, it’s not that impressive, etc.). It also provides context for who gets the most time (and relatedly who the coach trusts) on the power play and penalty kill.
5v5 Ozone Start%: Measures how often a player starts in the offensive zone. The more often a player begins his shift in the offensive zone, the higher his point totals should be. Cushy zone start percentages are 54% and higher, whereas guys who play a lot of tough minutes will be 47% and lower.
5v5 BZS: Conversely, this is where a player finishes his shift. Why it is BZS I have never been too sure, but nonetheless. I probably won’t worry as much about this stat as the others, so don’t worry about it for now.
5v5 PIM Dif.: Simply the difference in penalties taken and penalties drawn, usually expressed in a per 60 minutes form. Positive means a player draws more penalties than they commit. The high last year was 1.9.