As far as I can tell, there are three premises that fans of the Montreal Canadiens are using as justification for trading Andrei Markov. Not all three are used all the time, but it's always variations of these three:
The Montreal Canadiens are not currently contenders for the Stanley Cup but will be in 2 years.
Andrei Markov is replaceable within two years by free agency or from within (possibly by Nathan Beaulieu)
Andrei Markov is in decline and will not be good enough to help the Habs in 2 years.
Premise one is something we've been calling BS on all year, even going back to last year. Removing the Cunneyworth era from the last three seasons of NHL action, the Canadiens have been a 52.4% Fenwick team, one of the top teams in the league. Most of that time was spent without Andrei Markov, the Habs' second best defenseman.
Most was also spent without young possession drivers like Alex Galchenyuk and Brendan Gallagher. The Canadiens have been getting better and better every season. This year they were lock and step with Boston the entire season until Carey Price had his troubles, which was the only major difference between the two teams as far as play goes.
You're free to believe that the Canadiens aren't contenders, but you're dead wrong. All evidence points to them being contenders right now. What confuses me about this denial of the obvious though, is the insistence that the Habs will be contenders in 2 years. Based partially on Galchenyuk and Gallagher improving, but based just as much on prospects making huge impacts. This isn't a smart thing to bet on if you're sacrificing your assurances. There's no guarantee that Sebastian Collberg will be able to do what Brian Gionta does in 2 years, or that Beaulieu can replace Markov. If anything, those are extremely lofty and unrealistic assumptions.
That brings us to premise two, which demonstrates how insular fans can be. When watching a team you have a vested interest in, you notice every mistake, amplify it, and it becomes all you remember. We dealt with this same stigma with P.K. Subban being "too risky" when in fact he was the least risky of any player on the Canadiens. Let's break it down a little so you can understand how valuable Andrei Markov is.
In his last three full seasons played, Markov has been paid an average of $5.75M against the cap. Closest comparable salaries in those years were Bryan McCabe for two, and most recently Tobias Enstrom.
During those last three full seasons, Markov finished 5th, 2nd, and 4th in scoring among defensemen. He finished 5th, 1st, and 2nd in powerplay points among defensemen. In short, he's a top 5 offensive defenseman in the NHL and has been for a long while.
During those three seasons, he finished 16th, 18th, and 22nd in the league among defensemen in average ice time per game. This is a highly relied upon top pairing defenseman who also happens to create offense like nobody's business. Putting this pressure on Beaulieu seems not only premature, but flat out irrational. If Beaulieu ends up being that good, lovely, but don't put him in a situation where he has to be.
That brings us to premise three, which is that Markov is in decline.
Of course, in broad strokes, this is true. Markov is 34 years old and a defensemen will begin to decline at that age. Nicklas Lidstrom did, but did you notice? Probably not, because he was still an excellent defenseman.
The assumption seems to be that Markov's decline will be especially sharp, but the facts don't support that at all. Most defensemen of an equivalent talent to Markov don't decline overnight, and Markov's latest season does anything but support the notion that he's in sharp decline.
Luckily for me, I can actually prove this with data gathered over the last while. We can go all the way back to 2007 and look at Markov's career, and create trend lines for his performance, his role on the team, and the team's performance:
The y-axis is set at the beginning of this past season, with the rest of his career before it. What's interesting is that Markov has never outperformed team average for extended periods. However, Markov's performance level has gradually gone up, to the point where this year was his best Fenwick year ever.
Part of that is team effect. In fact, Markov's performance has stayed with the team's performance almost exactly on average throughout this period. That he was able to continue this curve in spite of missing the better part of two full years to multiple knee surgeries is pretty astounding and speaks to his talent level.
Granted, when he came back from injury most recently, his role changed. Whereas before he started most of his shifts in the defensive zone, his last 60 games he's been starting more often in the offensive zone. This is how to ensure you get value in an aging offensive defenseman. Lidstrom started 57.1% of his shifts in the offensive zone the year before he retired,
However the effect is exaggerated by the trend line due to Markov being extremely sheltered when coming back from knee surgery in both 2010-11 and 2011-12. This full season, he started just 50% of his shifts in the offensive zone, and after the first few games where Therrien was making sure he was okay, much fewer.
Even more impressive; around game 28 Therrien absolutely buried Markov in the defensive zone, and although he struggled there a bit with Alexei Emelin, and a lot with Raphael Diaz, by the end of the season he was playing the best hockey he had all year, especially when given an elite partner like Subban.
When I was talking to Brian Wilde on twitter, he mentioned that Markov's +/- in his last 19 games was an obvious sign of decline. While it's true that the Canadiens were outscored at even strength while Markov was on the ice during his last 19 games of the season 16-9, that had very little to do with his play. What it had to do with was not being able to get a save. During this time, both Carey Price and Peter Budaj were struggling mightily. When that happens, you're going to wrack up the minuses.
Markov's play however, was perfectly fine. During those 19 games the team managed 51.1% of Fenwick events while Markov was on the ice at even strength. In the playoffs, Markov had a 54.9% Fenwick, boosted no doubt by Subban, but impressive nonetheless. These are not the stats of a player in sharp decline.
Why use Fenwick over plus/minus? Simple. Fenwick correlates better with scoring chances, and has far more events. In this sample over 20 times the events. Broadening your sample gives you a better statistic for predictability of future performance.
Markov's closest comparable in career trajectory is likely fellow Russian Sergei Gonchar, which is funny because many who are screaming that Markov should be traded after Gonchar's big contract don't realize the irony that Gonchar is still pretty effective at 39, and that's why he was given that kind of money. Is Gonchar good enough to actually deserve that contract? Not anymore, but at 34 he was, and more.
Markov has one year left at $5.75M, after which he will likely command the same contract or less, as he'll likely move downward in value as he ages. However the Canadiens played Francis Bouillon on their third pairing for most of the season, a soon-to-be 38 year old with a history of even more injuries than Markov. And say what you'd like about Bouillon, and I bash him all the time, he wasn't completely out of place at even strength. There were better players pushed down because of Therrien's liking for him, and he should never have seen a second of powerplay time, but he wasn't actively poor.
Does anyone really believe that a 38 year old Andrei Markov would be out of place on any good team's third pairing? As he ages you shift him out of his tough minutes role, you pay him less, and he continues to provide value. The chances of getting fair value for a player of Markov's calibre in today's game is essentially nil.
Marc Bergevin stated when he took over the Montreal Canadiens that he doesn't like trading players if it will create a hole in his lineup. The hole caused by a Markov trade would be cavernous. We saw last season for the first time what Markov and Subban can do together, why mess that up now?