Guys, thank you for your positive or negative replays. Then starting this thread, I never planned to develop a real professional tool (nor I did). My goal was to underline by statistical 'evidence' the contradictions we have then evaluating our D players (only Defs).
My logic was:
1. I am watching a game. Player A is good.
2. I am listening to the coach. Player A did well.
3. Player A has a lot of credit from the stuff. Playing top minutes vs top opposition. Etc.
4. All this is subjective or difficult to measure.
5. I came here to discuss our players performances. And I read "A is bad because he cannot perform a quadruple axel (exaggeration from my part), B is good because he can skate on one leg (exaggeration again)".
6. I think: "OK, I may be wrong on A. I should find out!". And I look to the stats carefully to see 'Is A good or bad?'.
7. More I work with stats more I am convinced: A is an excellent player. I cannot find anything (almost) he is really bad at.
8. I came here again and I open this tread to illustrate my point of view and (hope so) discuss it.
You know the rest of the story.
It is not trolling as someone pointed out. It is not easy to talk stats with passioned (=sometime irrelevant) fans. First formula was not OK for me. I developed The second formula which is (in reality) a better personnel variant for Corsi stat. We count offensive and defensive actions for each player. The sum for all players will result in a 'similar' team Corsi stat. This is why I knew it will be correlating with Team points. Actually, if one does it correctly (with good weights) this stat should work much better than Corsi or +/-.
The constructive point here is that stats (even the awesomeness 2.0) is based on greater amount of relevant information compared to one’s opinion after watching some (many) games. If it does not change your mind, it should? at least make you think "am I correct with my opinions on players?". This thread is about it. Let's imagine we can be wrong! Let's think outside the box.