Post-game analysis discrepancy - GR rating
 

In the post-game Analysis, it says C Nobert didn't stay in his assigned role but his GR IS High.

Isn't that wrong?

Shouldn't the post-game analysis say he did stay in his assigned role?

Here are more examples - Torkki and Farj.

The Post game analysis says they didn't stay in their assigned role but the opposite is true.

No, this is correct, it's still possible (albeit less likely) for the player to have a good game even if he wasn't doing what the coach wanted.

How could Karjalhti be said to not stay in his assigned role when his GRs are in the 70s and 80s (76, 71, 81)?

It isn't about having a good game or not, it's about staying in his assigned role.

The point is that the post-game analysis says Karjalhti didn't stay in his assigned role but his GRs shows he did.

He did not stay in his role but played well in the role he switched to.

So for an (extrem) example: you set "Enforcer" as role for Patrick Kane. Most likely he will not stay in his role but instead play "Playmaker" (or something compareable). The Game Ratings are then based on that role and if he has a good game based on that role, the Game Ratings will be good.