Thursday, November 20, 2008

What Is Carbo After?

Lines That Make Sense Vs The Crapshoot

Today's article from The H Does Not Stand For Habs has prompted me to do a piece I have been putting off for a long time. I find her revelations similar to mine, but a little out of kilter.

A little while ago, I found this website that provides some great statistical breakdowns to look at. One of my favourites, and the one I have been gravitating to recently is the analysis of goals scored while a given player is on the ice (only available at even strength).


This statistical obscurity may sound trivial, but I think there is something to it. For on thing, it tends to confirm what I see from a subjective point of view. For another thing, lines are built for even strength – and so these stats are highly relevant for the discussion of line pairings. It looks at team goals when a player is on the ice rather than individual points. Ultimately this is more reflective of my goal for the player – to be on the ice for (and thus part of) a lot of Habs goals, regardless of whether he touches the puck.

In her article, JT had two revelations that took me aback.

The first, and this was news to me: Andrei Kostitsyn is clearly a right winger:
He has always been a right winger, and the only reason he was a left winger last year was because Alex Kovalev is a right winger. On Lang's line, Kostitsyn is the right winger, and it was evident in his play, as one of the very few bright spots in a dismal display, that he prefers to play on his natural side. He's better there, and he should not be moved back to the left.

Forgive me for being a skeptic, but despite Andrei's decent play the past few games – he was far better at point last year when playing the left wing! Her conclusion that Kostitsyn is more vibrant with Lang and his brother was plain to see, and the statistics play it out too. But could it be due to something other than a switch to right wing. After all (and good scientists in the readership will have noticed) a lot of variables were changed in the move – different centre, different winger, more responsibility, less shadowing, etc.

Personally, I don't think Andrei is better as a right wing. He is better when he is free to do what he wishes, and when he gets a bit of the puck. At least, that's what I've seen.

Her second revelation was equally as puzzling to me – Kovalev is back to 2006-07 form:
As it stands, he's managed to mess up the Plekanec/Kostitsyn combo. With him on the wing last night, the Koivu/Tanguay duo had its worst game of the year. If he's going to drag down whatever line he's on, he's either going to have to look after his health/mental issues or sit.

Now forgive JT again for being contrary, but I thought Kovalev actually played pretty well two nights ago. If I were to be entrusted the keys to the dome (won't happen anytime soon, I've been told), he was on my list of contenders.

Her observation was of course correct form the statistical point of view – but then again there was only a solitary goal on Tuesday to feed the stats. But I still can't accept that it is Alex dragging down every player he plays with, even in a slump like this one.

Looked to the stats
Having been in these arguments before with bloggers (particularly about Kovalev), I know I have to bring more to the table than a subjective view on the game. Else I only get the retort: "I watched the game too!". For that reason, I summoned the statistical breakdownery of that wonderful website I mentioned before. It being kind enough to show scoring rates at even strength (per 20 minutes), productivity of pairs of players with and without each other, and keep an archive back as far as I require it today. Without further ado:

Goals on ice for per 20 min (ES all linemates)



Player2006-07P2007-08P2008-09
Kovalev0.7261.0420.637
Plekanec0.9211.0170.794
A Kostitsyn1.0951.0571.359


While these stats are certainly interesting, and give credence to the Kovalev drag theory. Take a look at the same stats for the three together and apart:

Goals on ice for per 20 min (ES with specific linemates)








PlayerTeammate2006-072007-082008-09
Kovalevwith Plekanec0.8151.0360.841
without Plekanec0.6171.0640.000
Kovalevwith A Kostitsyn0.2651.1801.155
without A Kostitsyn0.7660.7140.000
Plekanecwith Kovalev0.8151.0360.841
without Kovalev1.0470.8660.571
Plekanec with A Kostitsyn1.3821.1161.250
without A Kostitsyn0.8190.7590.000
A Kostitsynwith Kovalev0.2651.1801.155
without Kovalev1.4420.4751.966
A Kostitsynwith Plekanec1.3821.1161.250
without Plekanec0.4870.8421.773


The first thing I notice from the stats is that both Kovalev and Plekanec are down. The season is early, so we'll have to see if they're just slumping or actually deteriorating.

The second thing I see is that Kovalev was not dragging down the "first line from last season". Playing with Kovalev this season, Kostitsyn's pace was actually up (I look at Kostitsyn because he was the first piece to be removed from the line, both through injury and line change – removing him was the end of the line).

Last year for the first time in living memory, we had a first line that worked. Each player (more or less) played better with the linemates than without. particularly striking is how well Kostitsyn and Kovalev fed the offence when on the ice together. This probably cost Andrei some goals, but helped the team totals massively. Can we ignore the success of last year after a few bad games?

I also look at the Kovalev drag differently. In fact, I'd say that this season, rather than Kovalev dragging the line down, Kostitsyn was dragging it up. Evidence from team goals at even strength show they happen more often when he is on the ice with most players. Look at Kovalev and Plekanec who haven't been on ice for an even strength goal in his absence. That's 73 minutes for Plek and nearly 100 minutes for Kovalev.

So by taking Kostistyn off line 1, he lost a line ticking at about 1.25 goals every 20 minutes (so a goal a game at their ice time) and completely neutralised Kovalev and Plekanec – his two best scorers from last season. He's made Lang and Sergei Kostitsyn better.

You tell me, is that maximising your assets??


The real question for Carbonneau goes deeper. Plekanec and Kovalev won't put up a duck for the rest of the season, so their dramatic difference with and without Andrei will close. But, Guy needs to ask who the focal point of the attack should be.

I see a few clear choices:

1) Kovalev
2) A Kostitsyn
3) Koivu
4) Tanguay

If the stats above tell us anything, it's that you can't make a super line without A Kostitsyn. He is an enabler. So, if it's super line we're after, why not Andrei with the suddenly dysfunctional Tanguay/Koivu duo. If it's a 2-line attack he wants then he must reunite Andrei with Kovalev and Plekanec and tinker with Tanguay/Koivu (maybe, I dunno their partner in success Latendresse?).

His waffling is costing the team right now. The one thing that was going right early on (the offence) has now come into question and all through needless tinkering. He talks about someone taking the team by the hand right now, but isn't that his job?

So what do you all think? Reunite the old line or AK46 with Saku and Tangs?

Let's beat those Sens, regardless of the answer.

No comments: