Those who have been reading me for awhile know I’ve been waiting for this: the collapse of the Colorado Avalanche. The NHL’s most uncanny team this season has been riding a wave of fortune so extended and unlikely, I’d almost begun to wonder if Joe Sacco had somehow discovered a new, paradigm-shifting hockey coaching strategy…
To say that the Avalanche were a surprise this season is a gross understatement. After finishing 2008-09 as one of the worst teams in the league, the org avoided spending money in the free agent market, plunked a bunch of kids onto the roster and then hid under some coats and hoped that somehow everything would work out.
Funny thing is, it did. (Sort of…)
To be fair, the Avs did manage one especially prescient foray into the UFA pool by snagging Craig Anderson (former Flames draft pick Flame Anderson!) for pennies. That particularly low-risk venture turned out to be high reward given his thus far 10th best .917 SV%; an especially impressive number given the fact that Colorado isn’t exactly the NJ Devils of mid-90’s in terms of defense.
That’s where the genius ends and luck steps in however. Boasting no less than two teenagers, five rookies, a pair of sophomores and more holes than a politicians testimony, the Avs line-up this season has completely flown in the face of conventional team building. If the object is to win, of course. In fact, the roster much more closely resembles your typical re-building patchwork of kids and cheap stand-ins that will give the youngsters some reps and not cost the owners a lot of money. And, of course, finish low enough to add another high draft pick to the pile.
If the latter was the managerial strategy, the Avs failed in that regard this year. Or, to be clearer, the fates failed them.
You see, Colorado didn’t just happen upon a great team where a lousy team should be (like, say, The Phoenix Coyotes did this year). Almost all of their underlying stats are universally bad. Not just average. Awful. They have one of the worst shot differentials in the league (-4.6/game or -352 total), spend lots of time in their own end at even strength and have routinely been outchanced. In terms of volume (shots, chances, possession, etc.), the Avs have been completely underwater all year, to the degree that they have frequently been bottom two by those various measures. Measures, I hasten to add, that correlate strongly with outscoring and winning in the long-term.
What has sustained the Avs seeming success over the course of the season has been frequency, or the percentages. A combination of Anderson playing lights out and the skaters scoring on more than 10% of their shots (10.6)* has allowed the Avs to float above their apparent weaknesses on a cloud of favorable bounces. It’s a cloud, I think, that’s finally beginning to dissipate…
*(To put that rate in perspective, the Chicago Blackhawks have scored 239 goals on 2611 shots this year, for a SH% of 9.15%. That doesn’t sound terribly significant, but a single percentage point across thousands of shots makes a big difference. At 9.15%, the Avs would have only managed 190 GF this year, or 31 less goals than they have now).
…because percentages eventually regress towards the mean in the NHL. It’s a hard fact the Flames themselves learned this year after shooting lights out for the first couple of months. Calgary SH% was up around a league leading 14% percent at the end of October, prompting Darryl Sutter to famously sneer "that’ll show the people who said we wouldn’t be able to score." (oops). Of course, the Flames percentages train crashed a lot earlier than the Avs…thus the temperate December and bitterly awful January that followed it (and subsequent mid-season dismantling of the roster). The Flames didn’t simply and suddenly "get less accurate" in the space of a few months. They were just fortunate to start – more of their shots happened to go in. It’s counterintuitive, but it happens.
The ES PDO mean in the NHL (that is, the sum of ES SH% and SV%) is 100. Most teams in the league are within a half a percentage point of that average. As are most players, in fact. Any number above that spread and we’re talking the province of chance. And the further one ventures away from the mean of 100, the more unsustainable the number.
With that in mind, the Avs have no less than 14 skaters with individual PDO’s of 102 or more. In fact, eight of them hoever at 103 or above. For context, consider the fact that the Flames have only one skater currently operating at a PDO of 103 (Mikael Backlund) and six around 101. Noted Stats guru Gabrieal Desjardins had this to say about Colorado back on March 16th:
But goaltending isn’t even half the story: not only do the Avalanche have one of the best save percentages since the lockout, they also have one of the highest shooting percentages.
in the long-run, teams simply do not sustain high shooting percentages. If Colorado’s shooters had merely been league average, the Avalanche would have 72 points today and a playoff spot would feel very much out-of-reach.
He notes that the Avs PDO (up to that point in the season) was the 7th best the league had seen since the lockout. That’s simply remarkable. And it won’t continue indefinitely.
The point of all this is that the Avs recent struggles are in fact long overdue. Or, to be more accurate, the Avs have actually struggled all season and only now are the results matching the performance.
They are catchable prey.