Sam Gagner, Dustin Penner And Ales Hemsky, The Oilers' Top Line?
In my article on Junior-aged players and entry-level contracts, I used a scatter plot to show Corsi against qualcomp. That plot contained an unintentional view of Sam Gagner's career progression and it spoke volumes about Gagner as a player. Because of that, I decided to study the same statistics for the ten forwards that have played for the Oilers each of the last three years. I gathered all of the numbers from Desjardins' Behind The Net, and placed all thirty points on a scatter plot. The results may give us a window into Tom Renney's brain as training camp begins.
Lowetide did an excellent workup on Renney and his coaching style months before he was hired as an assistant coach last season. He revisited the box after Renney was hired and not only found that Schrödinger's cat was still alive, he also found this:
How did he handle his top 6F in NYC? Played the hell out of them. In 05-06, Jagr (who loves Renney) played 22 minutes a night, including 6 on the PP. Other forwards who were above 4.5 minutes per night with the man advantage were Michael Nylander, Martin Straka and Martin Rucinsky. Looking at the TOI totals for that season it is very easy to see how he handled the forwards. Top line (Nylander-Straka-Jagr) 19 or more minutes; 2nd line (Rucchin-Rucinsky-Sykora) 15+ minutes; 3rd line (Moore-Prucha-Ward-Nieminen) 11 or more; 4th line (Betts-Hossa-Ortmeyer) about the same as the 3line. The PK men (Rucchin, Moore, Betts, Ortmeyer, Ward) were exactly the men who weren't on the PP. I suspect we'll have something similar here, a distinct skill 6 plus powerplay and a flat 6 for checking and the penalty-kill.
Will the scatter plot give us an idea of the top six forwards in Edmonton this season?
The plot below shows Qualcomp on the vertical axis and Corsi on the horizontal axis. The qualcomp metric that I'm using in this chart is Corsi Quality of Competition by www.behindthenet.ca The qualcomp figure is normalized for a twelve forward grouping. Those in the top right portion of the plot were on the ice against players that outshot their competition, but those OIlers carried the play. Those in the top left were out against outshooters but were buried. The bottom left shows players that were outshot by weak competition and the bottom right shows players that outshot weak competition.
(Click on the chart to see a larger version)
This actually looks like a chart that our resident astronomer should have posted, but it shows a few interesting points, nonetheless. Sam Gagner's progression is obvious here. He's qualcomp is increasing each year and his raw Corsi numbers aren't suffering. Dustin Penner and Ales Hemsky are the only fellows pushing the play. I've created a Taurus-like constellation for Zack Stortini in the corner to show that the fourth liner's been under water.
One notable point - and I've made it before - why was Robert Nilsson kicked to the curb again?
However, the Oilers have been a terrible team recently, and the raw numbers would reflect that. If we delve into the relative Corsi (Corsi for a given player MINUS Corsi of his team when he’s on the bench) numbers and compare the plot, we can see how important the player is to the team.
(Click on the chart to see a larger version)
In this plot, Gagner's progression is even more pronounced, and Penner and Hemsky's work more obvious. Stortini is still Taurus the bull in the bottom left corner and Ethan Moreau and Fernando Pisani still have a monopoly on "run over by tough minutes" in the top left.
This plot also brings Shawn Horcoff's role to focus. We know that Horcoff was held back because he was made to tow J.F. Snowplow around the ice. Remember this paragraph?
J.F. Jacques is outshot 2-1 without Horcoff. When paired together, Horcoff is able to stop some of the bleeding as they are outshot 4-3. Without Jacques, Horcoff is actually outshooting this year at 169-154.
Horcoff, as long as he's not saddled with Jacques, should have a bounceback year, even against the tough minutes.
I should also bring up Andrew Cogliano. In both plots, his points seem to meander around the screen, like a rudderless ship. It's an apt description for the young man's career to this point. I said it last season - that the team need to move him to the wing and let him sink or swim - and I'll say it one more time this season.
So Penner, Gagner and Hemsky are the only guys on the team constantly pushing play up the ice, and at the same time moving their plots to the right of the graph. Given their success as a line last season, it seems that these three may make a formidable first line. The problem for Renney is that there aren't any other players on this team, other than Shawn Horcoff, who have any history of success against any minutes, let along tough minutes. If Shawn Horcoff is going to take the tough faceoffs and the tough matches, he's going to need a couple of forwards that can, at the very least, tread water. Ryan Jones and Gilbert Brule are candidates, but neither have ever faced tough competition. Tom Renney has his work cut out for him - someone might want to throw some aspirin in that box.
30 comments
|
0 recs |
Do you like this story?
Comments
I suspect we’ll have something similar here, a distinct skill 6 plus powerplay and a flat 6 for checking and the penalty-kill.
I just did a quick rundown on who would be my choices for the PP 6 and the PK 6. The PP guys are pretty easy.
Who is going to kill penalties on this team? After Horc and Fraser, it gets pretty thin. Jones? Zorg? Cogliano? Brule?
Ugly
I’m a chart fiend and these get me off. Ahhh. Thanks Derek.
Corsi QoC may have some uses but I think it is inferior to the original QoC. Players capable of shutting down the other team’s top lines aren’t lower competition than soft minute killers, yet they would show that way… Penner tracking only above Cogs, Zorg, and Poo doesn’t seem to pass the smell test. Stone and Nilsson tracking with Horcoff also seems awry. I assume there is some literature on the subject that I am missing, but I think the original QoC would be better suited for these graphs.
by till_horcoff_is_coach on Aug 19, 2010 1:10 PM MDT reply actions
Corsi QoC may have some uses but I think it is inferior to the original QoC.
You’re a pain in the ass, but your request will be fulfilled tomorrow at 1430.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Well I wasn’t trying to be difficult… just really intrigued by the graphs and this series of articles.
If you don’t think there is value in QoC instead of Corsi QoC then don’t bother… it just seemed to me that it would be.
by till_horcoff_is_coach on Aug 20, 2010 10:17 AM MDT up reply actions
You’re not being difficult at all. The graph is forthcoming. Another chance to get off, I suppose.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Another chance to get off, I suppose.
Sweet – then use hot pink – I’ll wait till I’m home to check it out.
by till_horcoff_is_coach on Aug 20, 2010 10:28 AM MDT up reply actions
I like that...
in nearly every advanced metric this site posts Gagner looks better and better. It definitely reaffirms the belief that one day that kid is gunna be a helluva player. Though, all these advanced stats also bring up the interesting question “Why did we buy out mini-magic?” (the answer probably being that management doesn’t track adv. stats, but Nilsson is a saw him bad player on most nights too).
Great article Derek.
in nearly every advanced metric this site posts Gagner looks better and better. It definitely reaffirms the belief that one day that kid is gunna be a helluva player.
I wish they’d have left in him the O for one more year.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
This graph really helps to show that Hemsky is an incredible player. All three of his dots are in the upper right corner, and he’s progressing as well, not at the rate Gagner is (obviously), but Hemsky is still getting better and better each year. He’s extremely underrated as a player in terms of his ability to carry the play.
It would be very interesting to see a graph like this with all players in the NHL, or at least the top 20 scorers or something to see how Hemsky compares to them. I would guess he would fit right in with the elite players in the league. He seems to be severely underrated in Edmonton the last few years.
This graph really helps to show that Hemsky is an incredible player.
I’ve got an article coming up about Hemsky that shows a bit more of his amazing numbers.
He’s so under-appreciated by Edmonton fans.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Maybe if the hack would go and score 90 points he’d earn some respect!
In theory, there is little difference between practice and theory, but in practice there is!
Ah yes. The awesome 90 point barrier.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Ales better get 90 if he plans on keeping up with Hall.
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
by Bruce McCurdy on Aug 20, 2010 11:55 AM MDT up reply actions
I put Super Sam on top-line in my head long ago!
Great post!
You love this site, right?
Love mine too: www.hockeyfanland.com
by hockeyfanland.com on Aug 19, 2010 4:13 PM MDT reply actions
I think there are a lot of reasons why Horcoff with Hemsky and Penner makes a lot of sense. Granted, not all of them are hockey reasons.
1. With Penner-Horcoff-Hemsky we have a known commodity that can produce while tying up top competition on the other team.
2. Gagner has shown steady progression in his ability to push the play. Why not attempt to use him as the stabilizing influence on a secondary scoring line? Say with Hall and Brule.
3. Horcoff is locked into a contract. It makes sense to maximize his productivity and effectiveness now. I say pair him with linemates where we know this can happen.
4. Gagner is not locked into a long-term contract and by placing him with Penner and Hemsky there is a possibility that his performance statistics will be inflated above his actual capabilities. Gagner strikes me for various reasons as a pretty serious fellow. As a personal opinion I hope that the team doesn’t hurt itself by giving him the means to ask for more money than he’s worth.
5. Gagner is closer in his development arc to players such as Hall and Brule. There might be some peripheral benefit in having him make roots with the guys who are more likely to stay with the team.
2. Gagner has shown steady progression in his ability to push the play. Why not attempt to use him as the stabilizing influence on a secondary scoring line? Say with Hall and Brule.
Because they will be destroyed on the road. Somewhere one of the commenters on this site said something about drunk girls and prom. Go with that.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Gagner and Hemsky’s styles don’t mesh at even strength. Gagner needs wingers whose strengths are on the cycle or in the give and go. Hemsky plays more of a one-on-one, hog the puck looking for the open guy game. Power play…sure. Even strength. Bad idea.
Eberle has a much more complimentary skill set for Gagner than Hemsky.
Gagner has average physical skills. His strength is his vision and passing, and he needs to play a give-and-go game to create space for himself.
I think that is one thing that people forget a bit when they think about Hemsky. He’s a hard guy to play with in that you need to be a support player for him as opposed to a linemate.
I think with the emergence of Dustin Penner (in otherwords his ability to carry a line), seperating him and Hemsky give the Oilers more options.
In theory, there is little difference between practice and theory, but in practice there is!
At least at EV, Tom Renney was running a top nine more than he was a top six in his last two seasons in New York. This is what his forwards got for TOI per game at EV (minimum 40 GP):
2007-08
Fwd 1 to 3 – 14.21
Fwd 4 to 6 – 12.26
Fwd 7 to 9 – 11.23
Fwd 10 to 14 – 8.91
2008-09
Fwd 1 to 3 – 13.91
Fwd 4 to 6 – 13.22
Fwd 7 to 9 – 11.77
Fwd 10 to 13 – 8.07
2008-09 was really a top eight and a bunch of scrubs, but in both instances the biggest difference is between the 7-9 group and the 10-13/14. In 2008-09, there’s a larger gap between the 7-9 group and the 10-13 group than there is between 1-3 and 7-9. I’m not at all convinced that “Top Six” is a useful way to proceed when we’re talking about roles at EV. The fact that the Oilers need to create three lines to take a regular shift is one of the reasons that I’d like to split Horcoff, Penner and Hemsky completely, at least to start each game, especially if they decide to run with three rookies. I think I’d like to go with something like:
Hall-Gagner-Hemsky
Penner-Brule-Eberle
Paajarvi-Horcoff-Cogliano
The actual combinations are less important to me than the idea of a vet, a young vet and a rookie on each line so that none of them would just get their asses handed to them if they got out in the wrong situation. Plus, this way you’re not asking a young vet to babysit (Gagner out with two rooks just seems horrid to me). You also wouldn’t need to match so judiciously, since it would be more about getting them all experience against a variety of competition. That said, it would be really nice to have a “Pisani” instead of a “Jacques” on the fourth line to sub in for d-zone draws and protecting leads to provide some shelter within that context. It’s still something the Oilers can try with Fraser, who if I have my way, will be doing a lot of fogo this season (like Brodziak a couple of years ago).
Is that third line ever a god damned mystery going into the season.
I don’t know. I think I would be inclined to start Penner with Hemsky to see if what they did last October was a fluke or not. On the other hand, Hemsky would probably be fine regardless and Penner did show chemistry with Brule. Plus, that sort of forces Horcoff up a line, which in turn forces the third line to not play anymore.
I can really see a roster like that on opening night, actually. I’m still not sold on having three rookies, but I know that it’s going to happen. Maybe it will work better since it’s not like the team actually has the pieces to form a legitimate third “checking” line anyways, so it will be varying degrees of scoring lines (although that hinges on the mystery of Horcoff and Cogliano), and if you’re doing that you need to make sure any one of them isn’t going to get murdered too much more than the others.
you need to make sure any one of them isn’t going to get murdered too much more than the others.
That’s pretty much the idea. None of those lines would be taking on tough comp on purpose (depending on what the opposing coach decides to do; it’s likely that whichever line has two of Gagner, Horcoff, Hemsky and Penner draws the most attention). It gives each group a chance to play against varying degrees of competition, and each rookie would have a “mentor” to lean on.
by Scott Reynolds on Aug 20, 2010 5:50 AM MDT up reply actions
Thanks Scott. It’s actually encouraging to think of it in the manner you have been showing.
There seems to be enough talent in the top 9 to make something with it but line combinations tend to expose some players. The concept of playing the top players more and making lines on the fly to suit the situation give some hope (at least to me).
Until a penalty at least.
by till_horcoff_is_coach on Aug 20, 2010 10:24 AM MDT up reply actions
Two Capable Lines
xxx – Horcoff – Hemsky
Penner – Gagner – xxx
Or reverse the two, if you’d rather. Given the dearth of capable forwards in the bottom six, I’d say run those two lines at the tough comp and sent out Fraser+ for every defensive zone draw gainst lesser competition.
A posse ad esse.
The Copper & Blue|OilersNation|Hockey or Die!
Twitter: @JonathanWillis
Mail: jonathan.willis@live.ca
by Jonathan Willis on Aug 19, 2010 7:09 PM MDT reply actions
Outstanding post, Derek. Love those graphs, gotta admit I saw Taurus before I even read the caption, and that’s no Bull.
The deterioration of Moreau and Pisani is painfully apparent here, as both took a serious dive to the lower left. Both guys have been getting fed for three years straight, come to that, although obviously a lot of that is role-dependent. A guy like Nilsson is on the other side of that coin, tho’ I agree he would have been the one of the three to keep around.
I was more surprised to see the apparent deterioration of Pouliot, who I thought had played alright once he finally got healthy in ’09-10.
That said, I would dearly love to see a graph like this for a couple other teams (say Calgary and Vancouver?) to see if this distribution is in any way unusual. It would seem to me that you would expect to see your top line guys in the upper right, your soft-minute munchers in the lower right, your tough-minute checkers in the top left, and your fourth-liners and scrubs in the lower left, and indeed that is pretty much exactly what we see here. Kind of squares with some research I did in this area a year ago, though I have to admit your graph makes its similar points much more eloquently, not to mention elegantly. The three-year time lapse makes for a very nice addition though, gives one a sense of career curves.
The one point I made in that old post that I’ll stand by here, is that this distriubution would look quite a bit different if presented as a cube with the third dimension being Quality of Teammate.
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
Outstanding post, Derek. Love those graphs, gotta admit I saw Taurus before I even read the caption, and that’s no Bull.
I figured you would.
The one point I made in that old post that I’ll stand by here, is that this distriubution would look quite a bit different if presented as a cube with the third dimension being Quality of Teammate.
You figure out how to present a third dimension, and I’ll plot it.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
MatLab FTW.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
The Third Dimension
There are many ways to do it, but the one that seems most effective to highlight the key relationships visually is this:
1. x-axis: Quality of Competition (to the left, lousy competition… to the right, skilled competition)
2. y-axis: Quality of Teammate (to the bottom, lousy teammates… to the top, skilled teammates)
This graph alone would illuminate much. At a first pass, one would expect players to fall along the diagonal from lower left (4th line vs 4th line) to upper right (1st line vs 1st line). Deviations from this would show who is being protected (lousy competition, good teammates — they’d be in the upper left), and who is being thrown to the wolves (skilled competition, lousy teammates).
3. The third dimension can be accomplished by the size/shape/color of the marker. For this to work, you’d need to bin the Corsi numbers. For example:
Dark red = -10 corsi or worse
Light red = -5 to -10 corsi
Grey = -5 to +5 corsi
Light green = +5 to +10 corsi
Dark green = +10 or higher corsi
(Of course, you could pick the bins that highlight the data the most)
Or use the marker size for magnitude, and green/red for the sign.
With a setup like that, one would expect greys along the diagonal, greens in the upper left, and reds in the lower right. Anomalies would POP. Reds in the upper left mean you being given every chance to succeed, but are failing. Greens in the lower right mean you are something special.
@Derek: ^^ Like that.
Thanks, oilerBC!
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
by Bruce McCurdy on Aug 20, 2010 8:50 PM MDT up reply actions
See, that makes way more sense than my idea. ;)
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.

by 





























