Decoding the draft:
An excellent article on size and drafting
Stuart makes the ten worst list in this article
Decoding the draft:
An excellent article on size and drafting
Stuart makes the ten worst list in this article
http://www.sportsnet.ca/hockey/nhl/a-whole-new-way-to-look-at-nhl-defencemen/
Ya I'd be hesitant on that information for now...
Care to explain your skepticism?
Care to explain your skepticism?
The conclusion the article makes is interesting, but I think there is not enough data included to test out the authors theory.
Very cool. I look forward to reading it.I don't know about you guys but I'm super excited.
This can help break down why and where certain players have bad and good Corsi numbers.
A very powerful tool when combined with zone data and video work.
How good do you feel because your team has a new coach? I mean, really…it’s almost like a new-car smell. So many possibilities – This time, things will be different. With the exception of coaching changes due to disastrous, unexpected things, the typical hockey fan was ready for that moment, and were happy to see the coach go. But is that eagerness for change based on real results?
http://hockey-graphs.com/2014/09/23/remembering-dellow-forward-and-defensemen-goal-regression-corsi/Over the past year, I based a lot of research off of former work by Tyler Dellow. It is a bit funny because I actually never read any of Dellow’s work until well after I started writing about underlying metrics in hockey. I knew of him, but mostly was brought up on Gabriel Desjardins, Eric Tulsky, Ben Wendorff (yes, Hockey-Graphs’ own Wendorff), and a few others. It is also a bit difficult now because Dellow’s website has gone dark with his hiring, which removed the work I quoted or built upon.
One Dellow article that will be severely missed is Two Graphs and 480 words will convince you on Corsi.
Dellow presented analytical data in simple and effective ways. It made understanding of complex concepts -such as regression in goal differentials- easy.
Chris Kuc @ChrisKuc 1h Coach Joel Quenneville said #Blackhawks' analytics mostly focus on scoring chances.
What would they use then. Something they measure themselves?
Well, not sure scoring chances are one of them... yet (subject to change when sportsvu comes out for the reasons I list below will disappear).
Fenwick/Corsi only proxies scoring chances, but the relationship is close enough that you will be right almost all of the time. The analytical community used to track scoring chances like we do now with zone entries and exits, but it turned out that it was a lot of effort for very little new information.
The benefit though with underlying metrics is the fact that they are already tracked for you.
Now that sounds lazy, but there is an additional bit to it.
So you are tracking scoring chances and you notice Thorburn's is way lower than the top 9 forwards. How do you evaluate that in context? You know Thorburn is a 4th line player and less skilled than those above; you expect his results to be inferior. But, is it good or bad relative to Thorburn's usage or a fourth line player?
That's where Corsi/Fenwick surpass tracking scoring chances. Unless you track scoring chances for all teams and all players, you lose the ability to research things.
* How much does zone starts affect results
* How much does line matching affect results
* How much does linemates affect results
* How much does TOI affect results
* In what order do these take effect, which one dominates
* What is good and what is bad relative to a particular usage
* What is good and what is bad relative to a particular line
etc.
Two good articles on the topic:
http://nhlnumbers.com/2012/7/3/shot-quality-matters-but-how-much
http://nhlnumbers.com/2012/6/26/sho...ation-between-scoring-chances-and-shot-totals
Those two images show that scoring chances and Fenwick are very similar, and the players who tend to have differences in the two values tend to not sustain it.
Well, logically speaking, someone has to not suck.
The point isn't that scoring chances are bad. Given enough sample size, it will tell you the same thing. So one can't be considered bad and the other good. Both are pieces of evidence that help you make more good decisions and less bad.
I just mean that comparatively shot metrics have been superior thus far when coming to evaluating teams, players, and conducting research.
However, if I was in a statistical analyst role for a hockey team, I'd still track them. They are useful, especially if you can track the events that are leading up to them. It can help lead to the why's certain things happen, much like zone exits, zone entries, denials, and neutral zone score. But, I would still use Corsi/Fenwick as the superior overall impact measurer (with usage context and other statistics such as p/60).
I'm not sure how they would be using scoring chances, but they might be measuring and using them in ways that are different from the broader usage of them. I think that some teams now have the resources to do a better job of collecting input data, so they might look at scoring chances and what leads to them in more advanced ways than has been done by others.
I'm not sure how they would be using scoring chances, but they might be measuring and using them in ways that are different from the broader usage of them. I think that some teams now have the resources to do a better job of collecting input data, so they might look at scoring chances and what leads to them in more advanced ways than has been done by others.