Tonight's game v. Penn State

[TimV]

I'm guilty of holding several empty seats.  We had tickets, as for all the MSG games this century, but I spent Saturday doped up in a hospital trying to flush out some kidney stones.

That would be a legitimate excuse.::scream::

No.  No it's not.  Man up, you pussy.  Those IV stands have rollers on 'em.

Oh...wait... I forgot - MSG makes you discard your bottles on the way in.  Never mind.::whistle::

[adamw]

Ryan Lambert wrote an interesting article on Penn State and their shooting strategy in the context of shots-for analysis (it follows his discussion of Merrimack): http://www.collegehockeynews.com/news/2014/11/11_second_thoughts.php

For the record, I disagree, somewhat with Lambert - and told him so before I published his article

If shots were the end all be all, Penn State would win every game.

But he basically says that Penn State's strategy is a good strategy for Penn State; they take a lot of bad shots because, given their talent level, the alternative would be "very few shots" instead of "better shots". He basically credits Gadowsky for optimizing the performance of a weak team. It isn't strictly in praise of a raw number. Anyway, thanks for hitting publish on it.

Right. Where I disagreed with him is that, Gadowsky has always played that way, no matter how much or little talent he has.

[adamw]

IMO sports analytics proponents have a tendency to draw conclusions that are too strong given the available data.

That is exactly the issue I have with hit. Not that it isn't useful - but that some people - particularly the secondary proponents/disciples of the original statisticians - are too absolute in their conclusions.

[Tom Lento]

IMO sports analytics proponents have a tendency to draw conclusions that are too strong given the available data.

That is exactly the issue I have with hit. Not that it isn't useful - but that some people - particularly the secondary proponents/disciples of the original statisticians - are too absolute in their conclusions.

People who don't have a strong understanding of statistics tend to overstate their position. I find it equally irritating that so many people dismiss and denigrate statistical analysis from a position of ignorance.

Find people who draw reasonable conclusions and publish them. You can even debate them on the merits and they'll likely engage in a productive manner. Ignore the rest.

FWIW, I thought that article was well-reasoned, and I'm glad you published it. The fact that Gadowsky always employs that strategy doesn't really take away from the conclusion that his strategy might be beneficial to a team with Penn State's talent, and you can make the same general complaint about the Merrimack example (although I don't know anything about Merrimack's coach, the fact is that good teams also employ that strategy). My main takeaway there was that you shouldn't discard the metric over a couple of outliers, especially if there are reasonable explanations for why those outliers exist. The impact of these strategic choices on teams of given talent levels would be a subject for a whole different article.

My only quibble was I wanted a little more discussion over the distribution of the metric. Shots for percentage in college is correlated with success, but the scatter plot was pretty much a big cloud. I'm used to dealing with that, but as an outsider to hockey analytics I have no idea where that places the metric in terms of usefulness compared with other available stats. I mean, does the variance decrease a lot if we consider PP time, or time spent with a 2+ goal lead? Both of those should be at least straightforward to compute.

I recognize that I am actually a stats nerd, though, so I don't think this is a meaningful complaint for a publication catering to the average college hockey fan.

[adamw]

IMO sports analytics proponents have a tendency to draw conclusions that are too strong given the available data.

That is exactly the issue I have with hit. Not that it isn't useful - but that some people - particularly the secondary proponents/disciples of the original statisticians - are too absolute in their conclusions.

People who don't have a strong understanding of statistics tend to overstate their position. I find it equally irritating that so many people dismiss and denigrate statistical analysis from a position of ignorance.

Find people who draw reasonable conclusions and publish them. You can even debate them on the merits and they'll likely engage in a productive manner. Ignore the rest.

Sorry to self promote, but relevant ..... I don't know if you guys saw my article on analytics from the preseason, which has many coach comments and gives various perspectives:

http://www.collegehockeynews.com/news/2014/09/19_the_analytics_era.php

[upprdeck]

one thing you cant tell from the shot chart is how good a shot something is.. you can have 2 people take the same shot from 20ft away and get vastly different results. coach would rather see the film to decide is something was a good scoring chance.  you also dont usually see computed how many goals get created from taking bad shots and to whether ths goals created offsets the goals allowed.

[margolism]

All shots are not created equal.

All saves are not created equal.

Adfing a qualitative component to these statistics to make them more valuable and telling.

[marty]

All shots are not created equal.

All saves are not created equal.

Adfing a qualitative component to these statistics to make them more valuable and telling.

And all assists are not created equal.

[Jim Hyla]

All shots are not created equal.

All saves are not created equal.

Adfing a qualitative component to these statistics to make them more valuable and telling.

And all assists are not created equal.

For sure some assists are not equal, watch Saulnier (19) on CU's second goal. It starts around 1:10. Now that's an assist.

[George64]

Assists are almost always given to the last two players to handle the puck before the scorer.  I recall, however, that back in the late 60s or early 70s erstwhile referee Giles Threadgold awarded an assist to a Cornell player who never touched the puck, arguing that his screen and movement in front of the goal was integral to the score.  I think that Bardreau's effort on Hilbrich's first goal last night might so qualify as he clearly diverted the Denver goalie's attention and helped make the goal possible.  Moreover, he nicely positioned himself for the rebound that never happened.  At the very least, he deserves kudos from the coaching staff and fans.

[margolism]

I feel that an assist should be given to any player that played an instrumental role in the score.  This should NOT be limited to players that touched the puck.

[George64]

Assists are almost always given to the last two players to handle the puck before the scorer.  I recall, however, that back in the late 60s or early 70s erstwhile referee Giles Threadgold awarded an assist to a Cornell player who never touched the puck, arguing that his screen and movement in front of the goal was integral to the score.  I think that Bardreau's effort on Hilbrich's first goal last night might so qualify as he clearly diverted the Denver goalie's attention and helped make the goal possible.  Moreover, he nicely positioned himself for the rebound that never happened.  At the very least, he deserves kudos from the coaching staff and fans.

Although most referees today are rather bland, Giles Threadgold was an exception. For those of you too young to remember him, here's an interesting article from the March 17, 1970 edition of the Sun.

[upprdeck]

has anyone read the hockey stat guide? is it even possible to award an assist by rule to someone who doesnt touch the puck?  in hockey can you award more than 2 or someone other than the last 2 ?  in basketball you can but its hardly ever done.

[TimV]

Although most referees today are rather bland, Giles Threadgold was an exception.

Emphatically seconded.  Giles would often hop up and stand on the dasher ledge when a scrum occurred at the boards and balance there with his arms folded, looking down at the battle beneath him like a Greek god amused at the antics of the mere humans below.  I loved him.::rock::

[Bahnstorm]

Although most referees today are rather bland, Giles Threadgold was an exception. For those of you too young to remember him, here's an interesting article from the March 17, 1970 edition of the Sun.

My father was just talking about Giles last month and commented about how the crowd would cheer him. Timely post.