Saturday, September 15, 2007

Depth

Few NHL fans give much head space to the issue of roster depth, but they should.

This photo was taken at the Pepsi Center last March. A game that the Avs won 3-2, making the playoffs seem possible for them again.

Joe Sakic would play 15.5 minutes of five a side hockey, a staggering 13.3 minutes of that against Jarome Iginla. Clearly this is really where the game was most likely to be won and lost, and on this night Joe scored against Iggy's line and was plus one on the evening.

Mark Giordano of the Flames is shown defending against Mark Rycroft of the Avalanche. Not a glamour matchup, but goals scored in this time in any game count just as much on the scoreboard. And over the course of the year there are hundreds of such goals.

There have been a several good posts regarding Giordano at Five Hole Fanatics and Battle of Alberta recently. The fans who cruise the internet searching for fantasy league tips were surely disgusted that so much virtual ink was being wasted on a depth player, but those linked articles are all terrific reads, and a pertinent issue for the Flames especially.

To provide some simple evidence, I wrote a script to scour the NHL.com event sheet for Game 1 of last season, to sort the players by their position and total ice time in this game, and then sum up their simple +/-'s by classification as top-six-forward, top-four-defender, depth-forward or depth-defender. So, by way of example, the four Buffalo defenders who played the most minutes in this game ... their cumulative +/- in this game was +2. So BUF's top-four-defender total was +2 after Game 1.

Then the same is repeated for the other three categories of player, then the same for the other team. Then the same for the other 1229 regular season games.

It should surprise nobody that depth players, by and large, get their asses kicked. They always have. I mean if you broke in with the Wings this fall, then generally you will play against a similar level of opposition, but sometimes you will get Datsyuk as a winger when he's double shifted, or you'll be getting some passes from Lidstrom because he plays half of the game. Plus you aren't likely to get minuses from short handed goals against, because you won't be playing on the powerplay. Maybe once in a while if you have a great game, show some hustle, get in a fight ... maybe Babcock bumps you up to play with Zetterberg for the third period to send a message to everyone else. Still, that won't be enough to save you, because those occasional shifts against Tkachuk's line through the season ... well they are a bitch. It's almost impossible to fall into the minuses with that gig, but chances are it still won't be good, because chances are you aren't very good by NHL standards.

The Avs peg in as a fairly typical example. Sakic didn't play much with Hejduk at evens, so they generally represent the top two lines there. During 5v5 hockey the Avs were +28 when one or both of these two were on the ice last season. And -16 when neither were on the ice. Plus they chewed up the vast majority of the ice time of the other team's best forwards and defensemen.

Or, alternatively, by the simple metric explained above:

That's typical for an NHL team, COL is near enough the median.

Now looking at the Flames (just ignore CBJ and NYI for now).

So, at even strength, the Flames have one of the strongest top six forward groups in the league, and on defense the top two pairings are kicking ass. To be fair, in the case of the defense, the really cherry EV minutes (offensive end draw, with Iginla, vs other team's 3rd line that just iced the puck) surely they fell to Phaneuf's pairing, presumably because they didn't have a Liles or Preissing for that gig.

In terms of results from the top of the roster; given the difficulty of their schedule you could rationalize bumping the Flames past OTT, BUF and DET into first place. The obvious area with room for improvement is PK, which for some reason was poor, but beyond that the 3rd and 4th lines, and especially the bottom defense pair, needs to be improved. Because if those players were on a weaker squad their +/- numbers would drop a bunch, and for guys playing in front of Kipper on a good team ... those are really shitty numbers to start with.

Was Giordano the problem? Is Eriksson the solution? I don't know, the rational Flames bloggers say otherwise. Eriksson has a standout +/- on that CBJ squad, Alan Ryder ranks him surprisingly high, but Hitchcock was terrified to play Eriksson against anybody who is good. I think the +/- numbers are what has grabbed Sutter's attention here, it certainly wasn't "saw him good". Akin to Sutter's signing of Zyuzin last summer, and really similar to San Jose's acquisition of Semenov a few weeks ago. I don't think this will go well for the Flames, nor for the Sharks, who (surely to no one's surprise) have top-heavy issues nearly identical to the Flames in magnitude.

I included the Islanders on the table above because they had the best bottom six forwards in the league, both by eye and by these numbers. Even better than Anaheim, New Jersey and Toronto. And they had a bottom D pair that was freaking terrible. Or at least that was the case in the fifteen or so Isles games that I watched after Smyth was traded there.

The Blue Jackets are the league's extreme exception here, with a bottom D pairing that put up terrific results considering the rest of the team. Granted poor Adam Foote played the toughest opposition every night, and he played nearly five hours of EV ice time with the Nasher. Let's face it, that was never going to go well.

So two teams with an obvious weakness in the bottom pairing picked up two of the three guys (A.Johnson, Tollefsen, Eriksson) who qualified as the same for Columbus. Curious is that.

I think the team that made the best decision in this case was Columbus, because they kept Tollefsen. He looks to be the best of the bunch, and he's young, big, and even fights. Aaron Johnson seems like a nothing pickup for the Islanders, but he has such an unreasonably good penalty differential (25 MORE of his shifts ended in CBJ PPs than opposition PPs, yowza!) that maybe he deserved better numbers, maybe he was driving the results more than it appears at first blush. It's enough of an oddity that you would think that Sutter, Snow, Wilson and a few others would have downloaded a bunch of CBJ game video and reviewed Johnson/Eriksson/Tollefsen. But somehow I doubt they did. Hearing so much of Doug MacLean over the summer has destroyed my faith in the competence of NHL GMs.

9 Comments:

Blogger Devin said...

Terrific stuff as always, Vic. Do you have the numbers on the Pens last year, by chance, specifically the D-pairing metric?

9/15/2007 5:39 pm  
Blogger Vic Ferrari said...

Devin:

The PIT D by this simple measure:
Top 4: +13
The rest: -3

Up front: +28 for the top six and -23 for the rest.

So that seems very ordinary.

Again, this metric is hardly bulletproof. You're a Pens fan, no? How does that mesh with what you saw, Devin?

9/15/2007 6:57 pm  
Blogger MetroGnome said...

Excellent post and thanks for the love.

9/15/2007 6:59 pm  
Blogger Lowetide said...

Excellent post, Vic. Logic tells us there's some drastic dropoff, but the numbers are startling when placed in this context.

9/15/2007 7:52 pm  
Blogger Devin said...

Vic- thanks for those -- they really make their money on the PP in Pitts. This was yet another excellent post by you (you always get me thinking in new ways about what's really happening out there -- I would seriously pay $ for a subscription to this site, LT and MC, no question).

So yeah I'm a Pens fan (non Pens-watchers, skip the rest, it'll bore you to death), have been forever. I was a Lemieux and Jagr guy growing up (I was too young to really "get" Gretzky and what was going on then). I stuck with them casually through the awful years, but now that I can watch Crosby do these ridiculous things every night it sure makes it easier :)

The reason I wasn't sure about their results is the Eaton injury this year that forced Melichar up with Gonchar for a lot of games. He was with Scuderi on the bottom pairing but Nasreddine took over his gig when Eaton went down (and apparently it went down OK, but seriously the forwards have to be responsible for a lot of that). They have Gonchar playing just crazy minutes there, and I can't help but think jettisoning Orpik, signing say Markov, and having a Whitney-Gonchar 2nd pairing would be a sight to behold. Gonch would murder the lesser lights. Markov would add a sorely needed PK veteran. I dunno why they signed Scuderi and not Melichar instead, though. Pens fans hate Melichar for some reason (call it the Marek Malik effect, I dunno) but he did just fine playing some awful minutes.

Early reports from camp have them using Sydor with Gonchar, and Eaton on the third pairing with Letang. Maybe that will be enough to make a difference. You can actually watch the scrimmages online at http://penguins.nhl.tv/team/console - it's great.

9/15/2007 8:49 pm  
Blogger Black Dog said...

Great stuff Vic.

9/16/2007 5:05 am  
Blogger Cam Fraser said...

Again, just incredible analysis, Vic. It makes me sad that the dumbasses that are paid to provide the same end up telling us about the player's wives or mother's pasta.

9/16/2007 9:09 am  
Blogger Vic Ferrari said...

LT: I think for us Oiler fans it's more surprising, because this team had such strong forward depth for a few years. I mean it was common for people to discount the solid EV+/- numbers of Georges, Horcoff, Bergeron, Dowd, etc because they were near the bottom of the roster.

Maybe the thinking was "if they were playing against Jagr they'd be getting shelled" ... which is probably true. Still, fairer to compare them to their peers I think.

Nolan talked about the Oilers as a model for the type of team they wanted to build on Long Island. This was last summer, before his first season. And there was probably some cachet in that because the Oil had made it to the finals a couple of months earlier.

Still, in terms of the type of player, the aggressive style of game, and the not-much-for-stars-but-no-weak-links lineup ... very similar to the Oiler squads of MacTavish/Lowe last year and this excepted of course.

9/16/2007 2:11 pm  
Blogger Lowetide said...

Vic: Your statement after lineup in the final paragraph rings true and is the one reason why this season is likely to be very frustrating.

9/16/2007 8:09 pm  

Post a Comment

<< Home