Chicago - San Jose HtH Scoring Chances, Games 1 & 2
by Derek Zona on May 21, 2010 2:00 PM MDT
in Analysis
Email
Print
Commenter Passive Voice questioned my analysis in the Game 2 Scoring Chances post as I was basing my thoughts on head-to-head time on ice rather than head-to-head scoring chances. This forced to recalibrate and count the chances broken out by matchup. The results are extremely interesting and my first pass over this tells me that Todd McClellan isn't doing anything all that bad and that my original analysis wasn't too far off. Thanks to Passive Voice for making me go the extra mile, it yielded some cool stuff.
Thanks also to commenter Verstig for reminding me of angled cells because the chances matrix looked terrible until I changed the look and feel to what he's doing.
Note: Each matrix is compiled from San Jose's perspective
First off, we'll take a look at the the raw chances data, broken down into a head-to-head matrix. I'm only using the top three lines in each case because the ice time being given to the fourth lines in this series is insignificant.
A commenter in the game two thread took issue with me saying that Dave Bolland was getting drilled at even strength. He thought that there was some nuance about gamesmanship and so forth going on that made things better for Bolland. There isn't. Bolland isn't outchancing any of the top nine forwards from the Sharks. I don't think he necessarily has to do so, as I said in the last thread, he's providing cover for the Jonathan Toews line. The head-to-head chances show that in reality, Bolland is...getting drilled, but not only by Joe Thornton, he's getting drilled by Joe Pavelski as well.
What a straight analysis against time on ice didn't reveal is the extent to which Patrick Sharp and his linemates have beaten up on Thornton and company. These samples are insanely small, but I wonder if it makes more sense for Joel Quenneville to match Sharp against Thornton, Bolland against Pavelski and Toews against Manny Malhotra when possible.
These numbers are just raw numbers, so we need to bring a bit of context to them. Below is the head-to-head ice time matrix. This is the total even strength time for each team's top nine forwards against each other.
McClellan let Quenneville use Bolland against Thornton for the most part and Pavelski was forced to play Toews. Sharp's minutes have been well-spread against San Jose's top nine, as has Manny Malhotra's time against Chicago's top nine.
Now we can bring some additional context to the numbers by calculating the scoring chances data per fifteen minutes of head-to-head even strength time:
The biggest advantage in the series is by Pavelski on Bolland, but that's also the smallest sample of time amongst any match up. Next up is Sharp's advantage on Thornton and then Thornton's advantage on Bolland. Looking at the data this way, it shows that Dave Bolland is...getting drilled. It also shows that Pavelski is fairing rather well against Toews and Sharp, in addition to drilling Bolland. Chicago's only real advantage has been Sharp against Thornton.
11 comments
|
1 recs |
Do you like this story?
Comments
Interesting stuff, Derek. I guess it goes without saying that + and – figures are from San Jose’s perspective, but maybe you should say it anyway?
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
Typo?
A commenter in the game two thread took issue with me saying that Dave Bolland was getting drilled at even strength. He thought thatthere was some nuance about gamesmanship and so forth going on that made things better for Bolland.there might be elements of a hockey game that couldn’t be reduced to data to feed into my model. There isn’t.
All better.
Gentlemen! I have invented...this thing!
The “model is too simple” argument is basically the same as “I’m too wrapped up in the box score to put it down and too much of a homer to admit weaknesses in my own team”.
Which is fucking lazy.
If there are elements to the game you think matter, go out there and show that they matter. Just you saying something is true isn’t going to make us believe it is, in fact if anything it will make us believe the opposite.
I can’t believe that any serious hockey fan would actually be opposed to the idea of the scoring chance, it’s actually fundamental to hockey. To debase it as a “model” like those awful VUKOTA numbers or whatever… it really betrays a serious and irreversible misunderstanding of the game.
To be fair to cliffkoroll, in the last thread he was pointing out some (real) limitations in the way we track chances and simultaneously saying that he thinks that they have value. Saying that “statistics have limitations!” isn’t exactly new insight, but it is helpful to be reminded of it now and again.
by Scott Reynolds on May 22, 2010 9:05 AM MDT up reply actions
Thank you
I actually continue to be impressed with what Derek is doing here.
My only disagreeement is with the strength of the conclusions he draws from this analysis.
Gentlemen! I have invented...this thing!
by cliffkoroll on May 22, 2010 10:31 AM MDT up reply actions
Of course there are limitations
It’s only capturing one aspect of the play on the ice — pucks going off of sticks and toward the net. There’s a lot of other stuff that contributes to the flow of the game.
I’m starting to feel like there’s a dead horse being beat here on this, but this is where the so-called “eye test” comes in, try to to pick things apart further to figure out what else might answer the question people have been asking, which is “why is the Sharks top line struggling so much” conditional on “Dave Bolland’s line is matched up against that line, for the most part.”
Rather than arguing the merits of what’s been done (which I really applaud the effort on, and find it’s useful because there’s nothing else like it), I think the debate should center on what else could shed some light on this.
This is great
Definitely gives a clearer picture of what’s going on. Not to be a homer, but I have to wonder if checking lines will almost always be outchanced (interesting how Corsi goes the other way — kind of makes sense that their lesser scoring ability and mentality would lead them to take much lower-quality shots)… and in that case, how one would account for that. It’s almost like I have to try to figure out how to apply Hawerchuk’s (the BehindTheNet guy) QUALCOMP numbers or other normalization.
That said, I totally agree that the media blew this way out of proportion, but then again, they’ll make stories out of anything these days.
(also did you use Excel 2007 for these? they look really good, I feel like I have to upgrade now!)
Checking lines are almost always going to be outchanced because they are playing tough comp and starting in their own end. But that’s not a reason to make up stories about what’s happening.
(also did you use Excel 2007 for these? they look really good, I feel like I have to upgrade now!)
I did, but it was a complete pain. None of the menus are where they are supposed to be.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Menus? What are these menus you speak of?
Yeah that’s why I haven’t gotten it, even though I can get a big discount through my company. Having more than 5 non-ugly colors to choose from is nice, though.
Also, re: “making up stories” — I agree that people have really blown it up, but I do see the role of analysts as trying to “tell a story,” so to speak — to try to reveal something that may not be so obvious at first. And while Marleau is starting to throw some cold water on these theories, I do think people were struggling to answer the question as to why their line was being shut down and this was the best thing they could come up with (I think it’s just hard to pick stories out from the ‘Hawks play this postseason, so the writers just cling to stuff like this, or Byfuglien’s impact, and so on).
My reaction to this is more, oh look, the checking line is finally clicking (which it hadn’t been), and they’re kind of limiting (as opposed to dominating) their chances a bit. Personally I think whatever it is that’s gotten McLellan to ice them for ridiculous minutes per game has to be a big part of the explanation, too.
McClellan just went PvP for two series and didn’t have to make any strategic changes. Though I’m not sure if he feels he needs to. His goaltending doomed him from the start.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
Northwest Standings
Oilers Schedule
Next 5 Games
@
Ottawa
|
Sat 02/11 | 12:00 PM MST |
|
Toronto
|
Wed 02/15 | 8:00 PM MST |
|
Colorado
|
Fri 02/17 | 7:30 PM MST |
|
Vancouver
|
Sun 02/19 | 7:00 PM MST |
|
@
Calgary
|
Tue 02/21 | 7:00 PM MST |
Last 5 Games
|
@
Detroit
|
Wed 02/08 | L 2 - 4 |
|
@
Toronto
|
Mon 02/06 | L 3 - 6 |
|
Detroit
|
Sat 02/04 | SO - W 5 - 4 |
|
Chicago
|
Thu 02/02 | W 8 - 4 |
|
Colorado
|
Tue 01/31 | W 3 - 2 |
In Association With
Latest Puck Daddy - NHL from Yahoo! Sports
- Thursday’s Three Stars: Max Pac hat trick; Chris Phillips’ memorable 1,000th game
- It’s over: Montreal’s Scott Gomez finally ends goalless streak (Video)
- Scott Hartnell’s goal celebration now includes fighting Dion Phaneuf (Video)
- Puck Previews: Leafs meet Flyers; Steve Yzerman not playing in Winter Classic Alumni Game?
- Three theories on why Hurricanes might pull Tuomo Ruutu off trading block
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
