majormajor
Registered User
- Jun 23, 2018
- 24,196
- 28,669
That's a great article by the way.
I think how a player performs in goals for % at 5v5 relative to his teammates is a super important stat. It is as close as a stat gets to being the results, and you can then use the various underlying stats to figure out if a guy is just getting lucky or whatnot.
+/- would be an excellent stat if PP, PK & empty net situations were simply totally omitted from the stat.
Now you can compare +/- stats only for players that have very similar TOI distribution of ES/PP/PK. And as always with team stats, direct comparisons for players on different teams are not very helpful. "Relative 5-on-5 goal differential" should be pretty good comparison, just like Corsi Rel. would be if the goal of the game was about just getting a lot of Corsi events for, which it isn't.
5v5 +/- is a stat I use all the time, I actually forget sometimes that a lot of lay people still use the traditional plus-minus. I think we'll see a switch to a 5v5 or ES +/- soon enough.
I use +/- in a general way. The further away from even in the minus direction will make me look at that player closer to see if it is his play that is the cause, or is it the lines and or situations they wind up in. Same for the plus side.
I have been trying to learn the advanced side of stats, but I admit, I have a hard time keeping things straight.
The intuition behind +/- is great, but a quick improvement you can make is to use 5v5 +/- or 5v5 GF% instead of the traditional plus-minus. You can compare the 5v5 GF% to the traditional plus minus and see how they differ.
For the Jackets this year the results aren't much different for most players. Domi is dead last in both. But you can also sometimes have cases where guys are on the ice for a bunch of empty net goals against, or shorthanded goals, and the next thing you know they're a minus player despite doing okay at 5v5.