Sunday, December 07, 2008

Scoring Chances So Far


This is a graphical presentation of the scoring chances for the Oilers this season at even strength, as recorded by Dennis and summed by Scott. This is up to and including the home game vs Dallas, but does not include the game in St. Louis.

The blue bars represent the scoring chance +/- per hour. So for every hour of even strength (EV) play that Visnovsky has been on the ice, the Oilers have generated 3.8 more scoring chances than they've surrendered. So far this season, at EV the Oilers have scored about one goal per seven scoring chances.

The red bars are the Corsi+/- rates for the players (+/- of all shots directed at net), which serves as an indication of the end of the rink in which the player is spending more of his ice time. I suspect that the true value for most players is living somewhere between the ends of the blue and red bars.

The guys who have the better chances to shots rate (i.e. drive shot quality) are Brodziak, Strudwick, Cogliano, Moreau and Staios. Excepting Cogliano, that's presumably largely because they are spending much time in their own end, so a greater share of their few on-ice chances are coming in transition. Skill and the inevitable bounces obviously come into play as well. On the whole everyone is bunched really close at this, however, and I suspect they will become even more tightly grouped as the season wears on.

And as with all hockey stats, they have little value if context is ignored. The quality of line mates and opposition, as well as the situations in which they were sent over the boards, these are obviously huge factors in any hockey player's results.

25 Comments:

Blogger andy grabia said...

So what the heck is going on with Hemsky and Penner here? And is Cole getting screwed by lineup placement? I haven't watched enough to know. Help me out, Vic. :)

12/07/2008 11:29 am  
Blogger Jonathan said...

But really, Vic, what does Marc Pouliot bring to the team ;)

12/07/2008 8:05 pm  
Blogger rananda said...

i'll admit that im shocked souray has been keeping his head above water.

12/08/2008 10:19 am  
Blogger Traktor said...

How is Pouliot's EV scoring chances and Corsi opposite?

Is that telling us that he creates a scoring chance almost every time he's in the offensive zone?

12/08/2008 10:34 am  
Blogger R-Gib said...

Jeez, this just makes that Viscoff guy look like a monster of a player again!

12/08/2008 11:03 am  
Blogger Jeff J said...

The guys who have the better chances to shots rate (i.e. drive shot quality) are Brodziak, Strudwick, Cogliano, Moreau and Staios.

I'm not following you here, Vic. Do you mean those whose chance rate is significantly better than their shot rate? Because from the chart it looks like Cogliano, Pouliot, Pisani and Gagner are at the top of that list. Incidentally, if you'd done one of your quizzes and asked whose chance rate exceeded their shot rate the most, I would have guessed Gagne.

12/08/2008 11:53 am  
Blogger Jeff J said...

OK, never mind - I think I figured it out. By "chances to shots rate" you mean literally "chances for to shots for," not the "chance rate to corsi rate" depicted. My mistake.

12/08/2008 11:55 am  
Blogger Vic Ferrari said...

Andy:

First off, good to see you posting, hopefully Mike Winters and Pleasure Motors decide to talk to us again as well. But if they don`t they don`t, it`s their perrogative and it`s understandable.

Back on point: Corsi +é- seems to be a bit more volatile, so a couple bad games sees the territorial advantage (i.e Corsi) swing that scoring chances seems to take a few games to catch up to. Or so it`s seemed to me when I was tallying them game by game earlier in the year.

The other thing is luck (or randomness, which seems less truthy to me as a word, but is probably going to be more readily embraced). When I was looking earlier the kid line had an absurdly low number of scoring chances against to shots against. And to my mind at least, there was nothing driving that other than the bounces. That`s been driving itself back to the norm at speed according to these numbers, but it isn`t there yet.

The other huge thing is linemates, which I suspect will explain a lot of the Corsi volatility as well. I could factor that in using Johnsons time-with-other-players stuff, but the missing STL game would mean a lot of manual work, so I`ll wait until somebody emails Dennis the STL game and he gets a chance to go through it.

With that in, I`m sure that we`d see Gilbert leapfrog Souray and Grebeshkov. That would make sense to my eyes, anyways. Plus using the `Viscoff`script you can simply see that Grebs and Souray are fractions of themselves when they aren`t playing with Visnovsky, at least by the underlying numbers.

Pouliot is the hard one to figure. My guess is that he`s a very good player. He`s hard on the puck, and good stuff always eentually happens to players like that. Still, I suspect that the scoring chance +é- is flattering him here. Probably passes have found their way through defenders in a manner that can`t be sustained.

And just generally, I think the opposite has happened to Penner and Hemsky. I think we`ll see the Corsi and SC+é- gravitate to one another. Because they must.

To me the thing it shows, re Penner, is that if you`ve only got one trick as a hockey player, it`s good if it`s ``keeping the play alive in the offensive end of the rink``, because Penner doesn`t look any hell in the other two zones, and the same stuff happened last year in terms of uél numbers.

12/08/2008 1:26 pm  
Blogger Vic Ferrari said...

R-Gib

Scott will need to give us this, but I would bet that the Viscoff scoring chance +é- is off the hook. It has to be.

12/08/2008 1:30 pm  
Blogger Vic Ferrari said...

Ya, you figured it out right.

Good to see you hanging with us, Jeff.

12/08/2008 1:31 pm  
Blogger Hawerchuk said...

Vic,

Very nice stuff. The NHL adds a (!) to flashy plays that go up on the highlight reel...It would be great if they could use something like that to highlight scoring chances.

One question. Have you looked at shot quality-adjusted Corsi? I wonder if that would match up closer with observed chances?

12/08/2008 2:35 pm  
Blogger Vic Ferrari said...

Possibly Gabe. By the looks of Ken`s stuff, the gain would be very marginal.

And just generally, I like to stay close to simple truths when I can. Just my thing, doesn`t mean it`s right.

BTW, I assume that you wouldn`t mind me writing something to strip the Oilers underlying numbers off of BTN daily for Dennis.

12/08/2008 2:54 pm  
Blogger PDO said...

Vic, hit ctrl and shift on your keyboard - it'll fix the comma's and such.

What I really wanna know... is what the hell happened to Erik Cole. I mean, you had to know the numbers would take a hit coming out of the SouthLeast, but the guy has been positively cranked. His underlying numbers are no hell. His counting numbers certainly aren't anything to right home about. He draws lots of penalties, and hits a lot... but he rarely drives the puck to the net, he doesn't use his teammates...

12/08/2008 3:31 pm  
Blogger andy grabia said...

I don't get Cole either. From what I have seen, the coaching staff has no idea what to do with him. But I haven't watched a ton of Oilers hockey this year, so I could be wrong. I did watch him a ton when he was in Carolina, though, and it wasn't just the fact that he was playing in the Southleast. That guy's a good hockey player. Something else has to be going on, even if it's just adjusting to the WC style of play.

12/08/2008 8:25 pm  
Blogger slipper said...

Last season in Carolina Cole lead that team with 43 penalties drawn while he only took 10. This year he's leading the league in this category.

First, I think that creating powerplay opp's is pretty valuable, and shouldn't go without notice. Second, I think it says that this is a player who is hard on the puck and wins battles.

That said, he has shown no finish. I watched an interview with Cole where he said that it was a struggle playing his opposite wing at the start of the year. He and Hemsky have been on the ice for zero even strength goals this season, while he and Horcoff scored at a rate of almost exactly 1G/Hour.

I don't expect as much production from Cole as he had in Carolina, where benefited from more PP time and playing on his natural wing with a more talented center.

Whether he gets some more puck luck or not going forward, this is a useful player. God forbid, if the Oilers luck themselves into a playoff position, we could talking about Mike Peca 2.0;)

12/08/2008 10:56 pm  
Blogger Vic Ferrari said...

That all seems fair to me slipper. I think he'll be fine.

I think he would benefit from playing with a centre that is harder on the puck, though. And just generally better on the cycle. Moreau seems like he should be a guy like that, but he's more of a shooter and net crasher I think. The play dies with him quite a bit in the offensive end.

I think Cole will do well with Pouliot when MacTavish gets around to giving that a run.

It's important to get the Cole line going. The 05/06 Oilers really took off, in terms of outplaying the other guys at evens, when the RPM line got rolling imo.

12/09/2008 4:14 pm  
Blogger Bruce said...

It's important to get the Cole line going.

That's for sure. The current version (with Brodziak and Moreau) struggled mightily in the Shark Tank. According to the redoubtable Timeonice.com, with Cole on the ice at even strength the Oilers were outshot 17-0. With Moreau, 18-0. With Brodziak 16-1. The line did give up the one even-strength goal, and Roli saved their bacon a few other times. They couldn't get out of their own way, let alone their own end.

12/09/2008 7:35 pm  
Blogger Vic Ferrari said...

Yeah, lots of lines have looked brutal playing in the shark tank this year Bruce. And while a -21 Corsi in one game (Damn! hadn't checked that, but that's stunning) is something else, they weren't alone in negative double digits. Hell, Visnovsky was -20, and he's arguably been their best EV player this year behind Hemsky and Horcoff.

Plus MacTavish is settling into a Hitchcockian groove here now. Horpensky is going power vs power. And they got the lion's share of the Thornton ice time in that one (though pretty clearly MacLellan was gunning for the kids) and the Cole line got a stupid number of own zone draws.

The Hockey Gods were not fair on this night, thankfully :) It was a good night for the Oiler PK to show up, as well. San Jose looks scary good.

I still think that any line with that gig is going to need to be able keep the puck in the offensive zone for long stretches. Because no matter how good you are in your own zone, bad things always happen when you play their too much against good players.

Pouliot seems the obvious candidate to me. Though according to Stauffer (via LT) he isn't chomping at the bit for the gig.

Things are slowly getting sorted though. And in the nick of time, they really need to make hay here, because every team in the west is staying in the hunt. And the teams at the back (COL and DAL) have the potential to go on a tear if they get healthy and their goalies start stopping pucks.

12/10/2008 4:46 am  
Blogger Vic Ferrari said...

Also, if absolutely positively can't wait to see Matt's baseball style standings on Fridays, you can go here.

Detroit and San Jose are in a class by themselves no doubt. But beyond that it's crazy close in the west. The Oilers are tied for 8th, and are two games out of third place and two games out of 15th place.

It's going to be a great year, I think.

12/10/2008 4:51 am  
Blogger Jonathan said...

Vic, I know you don't really take requests, so I'm just tossing this out there -

How hard would it be to write a script to figure out what percentage of shifts starting in the defensive/offensive zone end in a goal for/against league wide?

I'm going to take a crack at this one day doing things the long way, but I think it would give us a really good idea of the value of things like zoneshift.

12/10/2008 9:59 pm  
Blogger Matt said...

Jonathan: Vic might correct me if I'm wrong, but I believe he once figured that -- on average, league -- a D-zone faceoff is worth about -0.6 of an attempted shot (Corsi point), O-zone faceoff worth about +0.6.

Use with caution, but that's enough to illustrate the fact that (e.g.) Kesler and Wellwood's comparable SF/SA numbers are really not comparable at all.

12/11/2008 10:25 am  
Blogger B.C.B. said...

Hello IOF,

I have a quick request, if you have time.

I have been trying to create some new stats around PIM: PIM/GP, PIM/60, and two subjective stats (aggressive penalties/GP and Lazy penalties/GP). I don't know how useful (if at all) they are.

Could you take a quick trip over to my blog, and rip me a new asshole? Since you are the smartest stats people I read, I thought I would ask.

Thanks
Brendan at
http://bringingbacktheglory.blogspot.com/

12/12/2008 1:14 pm  
Blogger Vic Ferrari said...

This should give you a good idea of how it shakes out, jonathon.

http://vhockey.blogspot.com/2007/11/starting-in-your-own-end-of-rink.html

12/16/2008 12:59 pm  
Blogger Vic Ferrari said...

BCB

I won't speak for the other guys, but I don't go looking for fights elsewhere, enough troublemakers come to my neighbourhood to keep me busy :)

Doesn't sound like my cup of coffee, though.

Desjardins (behindthenet.ca) does some good penalty stuff. He shows who is drawing and taking the penalties that matter (the ones that cause PPs for one of the teams).

12/16/2008 1:18 pm  
Blogger Jonathan said...

Thanks, Vic.

12/16/2008 1:48 pm  

Post a Comment

<< Home