News Article: "I'm Not Fahrenheit"-A look at Corsi, the Man

[garret9]

Sheesh. Defensive much? I didn't say "death to all stats" though it seems you may have read it that way. Anyway, it is not so ridiculous as you may think. Scouts are willing to discuss it.

http://news.arseblog.com/2014/12/grimandi-questions-role-of-stats-in-scouting/

"... these numbers act as a safety guard for decision-makers who lack courage. That, yes, that annoys me. The other problem, is that the players are aware of this trend. .... We see more and more players fleeing the duel and I think that there is a risk that this continues until statistics count this refusal of a player to engage himself."

I am not sure why comments like these ... sometimes I make them ... polarize opinion so much. Without perspective, stats are useless, and without stats, perspective is useless. Players and agents are not stupid. There's a lot of money at stake. More and more players will play the stats game in favour of risk. Coaches too. I was just wondering out loud what this means over the long run for the game in terms of its entertainment value.

Not defensive... it was a terrible argument that I just found ridiculous.

Your evidence has nothing to do with your initial point, that stats will bring upon boring hockey and overly promotes player safety.

It measures player performance. If a player performs better, they get better stats. Stats padding (at least in hockey) does not work at this juncture.

For example, players with high shot volume do not necessarily have a better Corsi. It's a common misconception that a player will pad their Corsi by shooting low percentage shots. If it is a low percentage shot and you have no support for rebound, then you are losing possession and giving opportunity for the other team to sustain possession and chances against you. The relationship between iCorsiFor/60 and Corsi% is quite weak, especially when you take account more zone time means more opportunities and better players on average tend to be better in most areas.

In terms of hockey, this is a terribly weak argument. Physicality and puck battles are a large part of the territorial battle and out chancing your opponent which is a major part of what hockey statistics are measuring. If you are doing something less that is important part of out playing the other team, it's not going to pad your stats that measure out playing the other team. If it doesn't help you in out playing the other team, maybe you should reevaluate your weighting of importance on it.

It's no coincidence that players like Andrew Ladd and Adam Lowry who are not overly skilled players but smart and physical exude solid Corsi statistics.

 

[garret9]

And cue in the cavalry to save Mr. Garret.

Garret was wrong in his dismissive and rude response to Grand..... and it isn't the first time.

Glad Corsi was trained as an engineer and not a stats or comp sci major.

Gives him a little more insight into messy problems and the limits of the tools currently being advocated by people who really don't understand the dangers or the limits of the tools they use.

The argument was not whether or not Corsi limited. Everyone knows it is limited, including myself (example one, two, three). That's a Straw Man argument.
It's not even used as a WAR stat. It's used to represent a large part of what drives wins, approximately two/thirds of non-variance and luck driven success. But this is all besides the point since it was never the initial argument.

The original point being brought upon was Corsi bringing upon more concern for player safety and a more boring game. I have yet to see a shred of evidence to this, and until I do I will laugh at whatever I want finding ridiculous.

Currently the only major changes Corsi has suggested is: more skilled players on bottom six than grinders, many teams dump and chase or chip out too often, a defensive defenseman *may* have a worse defensive impact than a puck moving defenseman, and face offs are more about how a team reacts to a face off over the next 10s than whether a team wins the face off or not.
I don't see how that verifies the initial comment on player safety and boring game.
In fact, it's the opposite.

It has nothing to do with "how sure I am of myself" on the value of statistics so your reply is far off too.

Also, ironically, Corsi was brought in as player evaluation tool by two engineers, Tim Barnes and Gabriel Desjardins.

 

[garret9]

The comment was that Corsi is driving away from risk and exciting play.

I thought that it was ridiculous because so far in hockey stats have done exactly the opposite.

That was the argument.