Hawks-Sharks regular season goalscoring breakdown
by gmh on May 15, 2010 8:18 PM CDT
Email
PrintIn light of the interesting scatter of goals in the Chicago-Vancouver series, there were a couple of comments wondering what the plot looked like for the regular season series between Chicago and San Jose, so I whipped up a quick graphic:
(open in new window for full size)
Goal descriptions
Quick observations:
- Total goals scored: Chicago 17, San Jose 11.
- Average distance of goals scored: Chicago 26.5 ft, San Jose 27.3 ft. Not much separating the two.
- NO POWER PLAY GOALS BY THE HAWKS. Did score those three shorties, though, as well as one 5-on-4 with the goalie pulled late in the game. San Jose had three PP goals (Boyle one-timer, Demers snap shot, Pavelski slapper from the blue line) and one SHG. Even strength goals: Chicago 13, San Jose 7. This is encouraging.
- Goals off odd-man rushes (not counting short-handed breakaways): Chicago 2, San Jose 4. This is not encouraging.
- Goals off rebounds: Chicago 4, San Jose 1.
- All the goals scored from above the top of the circles (40ft or so) by both teams had screens in front of the goalie. It'll probably be a big factor on the Sharks PP, since the Hawks haven't been shooting as much from the point with the man advantage, at least in the first two rounds. It'll be interesting to see if Q changes this tactic against the Sharks. Who the hell knows, with our PP.
- The wild card of this matchup has got to be Niemi. Would he have let in some of the close-range goals that the Sharks scored on Huet? Can he deal with maybe multiple screens in front on San Jose power plays? How long until the Sharks start elevating their shots?
- Take a look at the list of goalscorers from the goal descriptions. Who do you think will make a huge impact on the series, for either team? My bets are on Hossa and Seabrook for the Hawks, and Pavelski for the Sharks. Dude has been bank all season long.
70 comments
|
2 recs |
Do you like this story?
Comments
Thanks go to
our resident stat guru.
"If you've only got one day to live, come see the Toronto Maple Leafs. It'll seem like forever." - Pat Foley
Section 324: We're on the list.
We BE thankful
for our resident stat guru.
"If you've only got one day to live, come see the Toronto Maple Leafs. It'll seem like forever." - Pat Foley
Section 324: We're on the list.
Thankful am I
for the guru of statistics that resides on this site.
/Note unceasing sarcastic laughter in background.
by burpchelischili on May 16, 2010 10:51 AM CDT up reply actions
Thankful we are
for the statistical guru residing at this site
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 16, 2010 10:54 AM CDT up reply actions
Wow, that's a really interesting distribution
Anybody else notice that except for a couple of goals in the slot, practically ALL of the Sharks goals were from the left side? (or right, or blocker side though we don’t know where in the net it went.. .what is the convention for this?)
Also great point about the PP… did anyone say clown shoes? Until I’m proven wrong, I’m going to go ahead and say that our PP success was due to nothing other than Vancouver’s awful awful PK and Buff’s ass.
(btw, I’ll try to get going on that program of mine so that we can get pretty pictures like this that you can sort by player, opponent, etc. — but props for doing all this hard work!!)
sample size blah blah blah disclaimer etc.
but you’re right, I did notice that as I was charting the goals, but couldn’t really find a clear explanation for that, so I didn’t mention it.
Aahhh, yes, that program would be a godsend!
btw, where did you get that rink pic?
I’m going to use the NHL one first (if I can find a way to grab that too) since it’ll fit whatever scale the shot coordinates are in, but it’ll be a lot easier to see stuff on a bigger one…
I 100% agree with your Clown Shoes thought.
The Hawks PP was Clown Shoes vs. Nashville. Vancouver’s PK was terrible throughout the playoff <70%, it was not just the Hawks, LAK lit them up, but that was about all they did. It also does help the PP% look better when the Hawks score 4 PP goals while the Canucks were dancing to “The age of Aquarius”.
So unless the Hawks come out and go 1/4 or 2/5 on the PP regularly against the Fishies, they are still Clown Shoes! But hey the PK is still going strong as the Hawks PP!
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 15, 2010 9:13 PM CDT up reply actions
The one time I really miss Havlat
is during the PP. He was magic last year on the half-boards.
"If you've only got one day to live, come see the Toronto Maple Leafs. It'll seem like forever." - Pat Foley
Section 324: We're on the list.
Isn't that good news for us
Niemi’s go go gadget legs seem to like the right blocker better
I went to a fight the other night and a hockey game broke out.
- Rodney Dangerfield
by stacie7 on May 15, 2010 9:23 PM CDT up reply actions 1 recs
I was going to point that out
I don’t know if it me but it seems that he has a tendency to get beat more from the circle to huis left.
Maybe, just once, someone will call me 'Sir' without adding, 'You're making a scene."
rec'd
for “go go gadget legs” and associated insight.
Gentlemen! I have invented...this thing!
by cliffkoroll on May 16, 2010 10:37 AM CDT up reply actions
Where do you and VerStig get your stats?
You guys have got to have some massive break the shit down database. I know you watched the actual goals for this, but man.
I’m stuck with NHL.com, Blackhawks.com and ESPN.com (Go ahead and shoot me now for that one). I can get GF GA SF/G SA/G etc anywhere. But where do I find things like, Time of Possession, time on attack, blocked Shots, shots blocked, shots at net and so on, you know the stuff that tells more about plays and playing, not just scoring.
But once again awesome work here, have you quiet your day job yet? Might want to think about it.
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 15, 2010 9:04 PM CDT reply actions
my sources
the Event Summary pages (here’s the one for Game 6 of the Vancouver series) for each game provides the bulk of what I work with: S A/B MS HT GV TK BS FW FL F%, shots at net is just S + A/B + MS? Those, as well as a full Play-By-Play, Faceoff reports, TOI reports, and Shot breakdowns for each player, are all linked from the NHL and Blackhawks websites — if you go to the front page of the Hawks site and expand the “Last Game” section, you can find them all under the “Stats” button.
Those are just the most basic ones. Shift charts and Corsi/Fenwick can be found on the timeonice site. Hockey Numbers is also a good source, and BtN has more of the sabermetric stuff (sortable, yay).
There is no Time of Possession stat, unfortunately. I wish there were.
Hope that helps?
Also, if I quit my day job, I would no longer be able to afford hockey tickets, beer, or internet to watch hockey for free. I don’t have a day job, anyway, I have an evening job. :P
Thanks
That is just about what I was looking for.
I thought about using the Summary pages, but for two teams and a season that would be 160 games total for these two team, ughhh! My head is about to asplode just thinking of doing that one at a time!
I knew Time of Possession and on Attack were not “official” stats but was hoping someone had done it. I think you can pull out allot more from an entire teams character with those two added in the mix in projections, and even with individual players. For Example Player A has 20:00 TO; on ice for 2 goals 2 EV . Player B has 20:00 TOI 3 goals 3 EV. Well it looks like Player B is stronger. But if you knew this on top of it.
Player A, 8:00 minutes Team Possession, 5:00 on attack means 1 Goal per 2:30 on attack (The rest of Player A TOI is 6:00 PK, or 6:00 EV Defense)
Player B 14:00 Team Possession , 12:00 on attack 1 Goal per 4:00 minutes on attack (The rest is 8:00 EV Defense).
This is a simple and extreme example, but you can see what type of additional data you can pull out, from the player, line and team. Who is good a puck possession, who makes the most of time with the puck, who defends well, etc….
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 15, 2010 9:45 PM CDT up reply actions
the closest metric I can think of
to ToP would be zone starts for faceoffs, which BtN does make available, I believe? Obviously it doesn’t measure flow or whatever, but it does allow us to weigh some of the stats based on whether they occurred in the offensive or defensive zone.
yeah, or you could look at the time of the first play in a particular zone
(or read below… I feel silly being a half step behind you guys in this discussion)
Yeah that's what my program is for
If someone wants to volunteer to collect all the Game IDs for any team I can get totals of any of the Event Summary stats for the season. I only did the ’Hawks, plus our two playoff rounds. I think I went to NHL.com and went to the scores/schedule, looked at the page source, then used my super special text editor to slice off a narrow strip of text where all the Game IDs were, then trimmed out the rest.
The only way I can think of to do time of possession without watching video is parsing the play-by-play, but a lot can happen in between the recorded plays. I really wish zone entries (or puck dumps) was a stat… sigh
that's something that could be implemented
with sensors at each blue/red line. I wonder if that’s something that each rink would have to agree to do, or if the NHL could just make it … happen. I don’t really know enough about camera setups or whatever to hypothesize how this might be done, really, it’s just a geeky fantasy at this point.
It would be extremely valuable for a sport that values its zone play, though.
You do not even need sensors, just cameras
As long as you have at least two cameras (3+ is ideal) cover every part of the ice the puck can be tracked passively. MLB Broadcasts such as Fox and ESPN do this with baseball already to track a pitch AND hit the entire course of flight and tell its trajectory, break, velocity and acceleration.
But you cannot even get the NHL to install dedicated Goal camera’s but rely on Broadcast feeds, so I highly doubt something like this is much beyond a wet dream, since it would “appear” to only help stat geeks(God Forbid if teams used the data to help them selves out you know!).
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 15, 2010 10:11 PM CDT up reply actions
Yeah, you're right
It’s just cameras and some guys with programming skills and a good knowledge of image processing. Technically, if the puck didn’t get lost in the noise, and with infinite time, it’d be possible for me to write a program to track the puck on those Justin.tv feeds (or maybe you could just figure out the rink perspective… assuming the cameraman follows the puck, heh)
You would need at least two cameras on the puck at the same time
One camera just gives you a line of bearing (2D), to tell where the puck is you will have to give it a reference plane, which would be the ice. If the puck is airborne it would be placed further out from the camera perspective than it actually is. You can not use with width of the puck to aid in gauging distance since if you have a puck in the center of the neutral zone, and the camera is viewing the entire neutral zone, even in 1080 HD, each pixel would be 0.888" wide, for an 80’ wide perspective or 29.6% of the pucks max diameter or 88.8% of the min width (on edge). This would cause a missive error factor. Referencing to the Face off dots and Where the line intersect the board to know orientation of the view would be simple though.
You would need to have two cameras at min, and know what the orientation of the camera is while tracking the puck. This can be done either by Telemetry sensors in the camera, or being able to reference the view based on on-ice features (faceoff dots, line positions, crease, etc.). Two cameras will give two lines of bearing from two different planes,where they intersect is where the puck is. Additional cameras reduce the error area. Kinda like how GPS works, instead of using time-domain to determine satellites distance to calculate you position(which requires 3 for 2D and 4 for 3D based on spherical correlated intersections), you are using known sensor positions to calculate puck position using line of bearing intersections
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 16, 2010 5:42 AM CDT up reply actions
Yeah, the NHL stuff is great
I use the Game Summary for goals (the mutual +/- stuff) and the Event Summary for just about everything else, except anything TOI related, most of which comes from timeonice.com (which has Corsi numbers too). My program just helps me aggregate across games.
I’m fairly certain that BtN has something that parses the play-by-play data… I’d do the same (the logic would be similar to the Game Summary) but I’m too busy lazy.
Btw, just a quick thought I forgot to mention yesterday — I wonder if you could use whether a shot (and ensuing save; I think all goals have one) gets a video highlight made of it (see the NHL GameCenter and the little video icons) as a measure of whether it was a “quality shot.” It’d be crude and wouldn’t catch everything, but, it’s a possibility.
quality shot meaning something not sent directly into the chest pads
so as to require a more dramatic/interesting save or you know, effort on the part of the goalie?
I’d say it’d be one of those all osmosis is diffusion but not all diffusion is osmosis things. the highlight reel shots/saves would probably all be quality shots, but not all quality shots would be on the highlight reel.
Watch it with the jinxing!
Time for some thrillin' heroics!
I have a day full of family events today
I do not know if I will be back for the puck dropping (my blackberry will keep me up to-date though if needed!).
So I need to say it before I go here shortly
FUCK SCUM,
GO HAWKS, lets go fishing!
Get off my Land!
ART.I§8-11; AM I-XXVII
James Madison is my Hero!
by Toews-makes-funny-faces on May 16, 2010 7:17 AM CDT reply actions
Good work @ gmh
I guess the way to beat Nabokov is to get the puck in the center of the ice close to the net. Most of the Hawks goals (as expected) are in this region of the ice. I agree with Verstig that’s weird that all of the goals are from one side of the ice, although there’s not much strategy wise you could do here, maybe put your better shot blocker there? but if it’s like baseball, you might mess up someone’s habits. who knows.
I wonder how many of these were glove side
You get sharper angles shooting from that side, so maybe it forces the goalie to use more peripheral vision (rather than straight-ahead) to track the puck?
the nature of those goals
if you go back and rewatch game highlights, don’t leave a lot of room for shot blocks. A lot of them were from broken plays or odd-man rush types, or people just blowing their defensive zone coverage. Transition plays, in other words, where it’s not like they’re setting up an umbrella and passing it around.
heh
every time I post, you guys make me feel a little less intelligent lol.. good stuff
aw, don't say that!
What you suggested re: blocking more shots on that side is really logical if you’re just looking at the plot, which is why I wrote out goal descriptions (which are linked to in the post). Sopel’s always going to block more shots on the PK than while backchecking, for example, just because both players are in motion and it’s hard to get set in that scenario. The Sharks play a really nice transition game, sort of like the Hawks, so it’s harder to predict how their forwards are going to attack.
Trust me it was a compliment of sorts
I’m new to hockey, watching the game doesn’t give me the type of information that it gives people who watch it daily. I used to listen to the Hawks on the radio in hte early 90’s, when I was supposed to be sleeping, then the Hawks hit the crapper… for a long long time.
Probably not?
Unless someone has been doing some illegal filming of me…
the radio call really helped me
when I first started watching hockey (only about four years ago, so I’m still quite new to it as well). John Wiedeman does such a great job describing the play that it’s really easy to visualize where things are happening relative to one another. I think it’ll get easier for you when you start watching more and more games. And as a baseball stat guy, you probably have an instinct already for what data will be more relevant or interesting than the usual ones.
Haven't listened to a game on the radio since 92-93'
I really only remember hearing “AND A SAAAAVE by BELFOUR” a lot. I’m not in the Chicagoland area anymore.
You can stream the radio calls over the internet!
if you go to the Hawks front page right now, there should be a link to listen to the audio.
I wish I could sync it up to my TV
But next season I will listen when I can’t watch
First try with my shot chart program
It doesn’t look like there’s a way to figure out PP’s without matching up with the play-by-play data, but here’s the game where we outshot the Sharks 47-14 (click to enlarge):
the only problem with culling numbers directly from PBP
is when they fuck up the distances. For example, Pavelski’s shot from the point. They inexplicably listed it as coming from 18ft, which might have been true if it had been deflected, but the goal was Pavelski’s, and unless they have cloned him somehow, he was very clearly the one who shot it from the blue line. This is why I did them all by hand — I still don’t fully trust the stat keepers in this league.
How bad is it?
I can definitely understand not trusting them, but if you want to aggregate data over the course of a season, it’d be sooooo much work
well, I estimated that Pavs goal as 60ft, when it was listed as 18ft.
So, PRETTY FUCKING BAD on that example. Other times it’s closer to maybe 2-5’ difference, depending on where you’re considering the “shot” to start. I usually look at the shooter’s feet relative to the puck, the stat keepers tend to miss too short rather than long.
I think I'd do a pretty awful job
I have a hard time mentally translating the rink size on TV into real-life distances.
That said though, I think with statistical analysis (as opposed to true illustration) is usually done assuming that errors are going to be made, and that over time there shouldn’t be too much bias in one direction or another.
I don’t think there’s much we can do about this… I guess you can always watch the goals and adjust those by hand, but for the other 56 shots…
I assume they train stat keepers to know this sort of thing
and there are enough markers on the ice to make the estimation pretty straightforward. Not as much as football with its hash marks, or basketball because of the smaller court, of course, but the biggest issue might be the pace of the game. The reason for stat keepers missing shorter is because of forward motion, so where the puck first leaves the stick is not usually where the movement ends, but a little further away. That’s the kind of stuff I’m looking for when I re-watch goals.
I agree with shots, you can figure that those will still give you relevant information over the course of the season, but at least double- or triple-check the goals after the games, there are only a handful and those are the ones that matter the most, anyway.
Heh, tell THEM that
I’m not sure if my program will be all that easy to use… the coordinate system in the data is not even scaled to the size of the image they use (I don’t know if you had a chance to look at the numbers… I’ll try to dump them into a spreadsheet or something… though I just realized, it might be in feet, which might make my job easier) so I had to come up with a way to scale them.
I guess the best way will be for me to make my program dump everything to a spreadsheet, then you can edit the spreadsheet (and make the necessary adjustments) and feed it back in.
also, wait
wasn’t this the game where Hossa scored a late goal from along the goal line to the left of Nabby, and I don’t see that reflected in this chart. CONFUSED, HELP.
I think it is
if you mouseover the goal here, it is a Panda goal… is this another case of the stat keepers f’ing up?
oh, you know what
the goals are plotted in mirror image of what they actually look like from the tv camera. The Sharks goals have been flipped as well. That’s just unsettling.
Mine? Or the Gamecenter ones?
Or both?
This is easily fixed though, give me a sec
(I hope the 2nd period ones aren’t flipped the other way though, that would be a huge pain)
yeah, that looks more like it
although the Hossa goal should still be more along the goal line than it is, and the Thornton SHG is not nearly as close to the crease as it should be — it looks like that was one of the goals that the stat keepers missed long on, because 23ft is more like where he got the puck (on the plot, its right at the middle hash marks), but he held on to it for a few feet and got closer to Huet before shooting (in between the hash marks and the top of the crease), which I didn’t adjust in the goal description but did on the plot.
Just adjusted it again
but I can’t push the stuff further out without shots coming from behind the net (could those possibly count as “on net?”)…
I’ll try to make this spreadsheet and send it to you, see if you can’t make some sense out of it maybe
if they take some sort of deflection out front
then they can be on-net? Off the defending team, that is.
I don’t think it’s anything you can adjust further on your end, because the NHL shot plot has it wrong and I assume you’re just extracting the data from what they’ve got.
looks like it... sigh
I just sent you the data that I grabbed to make this, maybe you can make some sense out of it, but I’ll try to make it so that I can feed any adjustments back in, so that the dots are movable.
oh snap, I just realized — I think the coordinate system they’re using is just in feet… I’m going to go back and fix my program so that it works with this (plus I can use that huge rink image you sent me :D)
I thought it was some crackpot pixel mapping
I just multiplied the numbers by 2 and shifted them by a tiny factor… turns out the NHL image isn’t even the proper proportion, go figure
still a bit small
I guess I could save it and make it zoom but it’s a bit small even when you click to enlarge
It's the size that they have on the NHL GameCenter page/feed
I ripped the image from their site… gmh sent me a huge rink graphic (bigger than my screen!!) so I’ll work off of that later on.
All these great stats and analysis
And so now I wonder — what do you think the coaches use? Do they already have a way of obtaining this sort of information? This info seems like it could definitely help plan a game and specific plays and match-ups.
The sun never sets on a badass
the use of stats in the NHL seems to have gone a long way
although some teams in particular seem to favor it more than others. It all seems pretty secretive though— if they’re using it, we won’t really be able to tell. I remember an article talking about stats as they are used in the NHL, but I don’t recall seeing Chicago as one of the teams that uses it. Buffalo and San Jose use it most, I think.
But when it was suggested to him that Toews v. Kane seems likely to become a sidebar to every future international hockey tournament, he smiled and said: "I'd like us to win something together, too."
(Tweets @ChiBlackhawks and blogs at Blackhawks Down Low.)
by chiblackhawks on May 17, 2010 9:56 AM CDT up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
In Association With
Latest Puck Daddy - NHL from Yahoo! Sports
- Monday’s Three Stars: Hanzal leads Coyotes; Kessel, Leafs keep rolling
- Puck Previews: Oilers visit Maple Leafs; Red Wings go for Coyotes sweep
- Howard Baldwin, ex-Whalers owner, offers critical look at ‘marginal’ NHL markets
- Ray Emery’s chance to start equals goaltending migraines for Blackhawks
- Puck Headlines: Sam Gagner didn’t score 8 points; Stewart waived; Crosby update
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.
