I haven't said anything about your ability to use them. You did say something about mine and so I merely pointed out that this was silly of you to say because I did not even try to use them. Nowhere in there does that constitute an inability to either think critically, use them or understand arguments being framed around them. Without naming names it was frustrating and time wasting sifting through all the useless posts cherry picking stats b/c people desperately had points to prove. Others would use them correctly which resulted in time not being wasted and good discussion. But those who did know that they did
Advanced Stats
- Thread starter Thirty One
- Start date
-
Xenforo Cloud will be upgrading us to version 2.3.5 on March 3rd at 12 AM GMT. This version has increased stability and fixes several bugs. We expect downtime for the duration of the update. The admin team will continue to work on existing issues, templates and upgrade all necessary available addons to minimize impact of this new version. Click Here for Updates
I haven't said anything about your ability to use them. You did say something about mine and so I merely pointed out that this was silly of you to say because I did not even try to use them. Nowhere in there does that constitute an inability to either think critically, use them or understand arguments being framed around them. Without naming names it was frustrating and time wasting sifting through all the useless posts cherry picking stats b/c people desperately had points to prove. Others would use them correctly which resulted in time not being wasted and good discussion. But those who did know that they did
That's because its being worked on at this very moment. Shot correlation is still very good.
Looks like it to me:
I see a 30% spread at minimum. That doesn't fly anywhere in the world.
BlueshirtBlitz
Foolish Samurai
On a different note:
When I go on behindthenet, which I assume (rightfully, hopefully) is the NHL equivalent of fangraphs, there's so many Corsi stats I don't know what to look at. There's relcorsi, but then there's Corsi QOC, and then there's Corsirel QOC. I assume the one that takes all three into account is the most complete stat, but I don't know because I don't know maths.
When I go on behindthenet, which I assume (rightfully, hopefully) is the NHL equivalent of fangraphs, there's so many Corsi stats I don't know what to look at. There's relcorsi, but then there's Corsi QOC, and then there's Corsirel QOC. I assume the one that takes all three into account is the most complete stat, but I don't know because I don't know maths.
Thirty One
Safe is safe.
- Dec 28, 2003
- 28,981
- 24,357
CorsiRel qoc measures the competition a player faces weighted on their CorsiRel
that's unfair. There are many posts that are just rants; but there are also many posts that discuss why a player is not good in one's opinion and those shouldn't be degraded since this is an opinion board.
His +- will tell you if he's helping to keep the puck out of his own net. I would also submit that if Brian Boyle has the puck, that means aside from our opponent not having the puck that one of his linemates doesn't have the puck which would be better than BB holding it along the boards. And this goes for alot of players in the league where them having the puck is only marginally better than the other team having it.
Maybe what these stats are missing are measures of variability? That is, if player A has the same Corsi every game and player B has Corsi's that fluctuate dramatically, can the players still end up with the same rating? If so, the within-player variance in Corsi rating could not only tell you about a player's consistency, but also whether the difference between Corsi ratings is meaningful.
It always comes down to stats: A win occurs when and only when GF > GA. If you don't score (or facilitate scoring scoring), or prevent scoring, you're useless. It's no coincidence that the top scorers are generally the top earners. Physical and mental attributes are just means to an end.
MorrisWanchuk
Registered User
- Sep 23, 2008
- 2,930
- 0
The Gloaming
23
Ail
Based and Rangerspilled.
Here it is, finally the dagger in the heart of advanced statistics. We knew this day would come. Lock 'er up boys.
Thirty One
Safe is safe.
- Dec 28, 2003
- 28,981
- 24,357
Chaels Arms
Formerly Lias Andersson
Puck possession is useful and it's good that people are developing a means of tracking it. This league is long overdue for a moneyball-esque overhaul but the only way that is going to happen is if the advanced stats can provide a GM or a coach with a legitimate edge over traditional scouting and stats.
Look at the leaders in corsi on with at least 30 games played. Kopitar, Williams, Muzzin. What important (and accurate) information can I really take away from that other than concluding that Los Angeles is a really dominant possession team? If I'm a GM, does that mean I should offer my elite "traditional" stat player for Jake Muzzin?
What is the predictive power of corsi on? If I take any one of those players away from the Kings and put them on the Florida Panthers are they going to replicate that corsi performance? If not, than what conclusion can I really draw about any of those players individually?
Possession can't be gauged in a vacuum, it just doesn't work.
The other problem is that all the top corsi on players are generally recognized as being really good players by traditional stats guys and scouts as well. Does a GM really need to look at the advanced stats to know that Anze Kopitar is a dominant player? Marty McSorley a couple of weeks ago said he would take Kopitar over Ovechkin. I doubt he looked at behindthenet.ca to make that assessment.
So now let's look at some of the "surprise" players on the corsi chart. Tyler Toffoli is 10th in the league. But the problem with that is, again, there are 7 LA Kings in the top 15 in corsi. Which one of those guys is actually driving the possession and which guys are just along for the ride?
