The Athletic’s power ranking, Dom predicts the Blues get swept. He’s really down on the team.
I like Dom's data work a ton, but I think his analysis/interpretation of that data is sometimes lacking. I think that past/current data has much less predictive value in hockey than he does, so I'm not surprised that he looks at the underlying metrics and comes to the conclusion that it will be stunning for the Blues to win a single game in the series. With that said, he is absolutely correct that this Avs team is exceptionally good. They are one of the best 5 on 5 teams the league has seen in the last decade by pretty much any metric. There is debate about how the COVID alignment/schedule impacts these numbers, but you can't deny that their comparative advantage over the Blues is gigantic.
At 5 on 5, they are 1st in CF%, FF%, SF%, GF%, xGF%, SCF%, and HDCF%. They are 58% or higher in every single one of these metrics. Most impressively, their xGF% sits at 60.8%, which is a staggering 5 point edge on the #2 team in the league.
There is a larger xGF% gap between the Avs and the #2 team than there is between the #2 team and the #14 team. That is stunning. Natural Stat Trick goes back to 2007/08 for these types of metrics. This is the first time that any team has finished the season above 58% in xGF% at 5 on 5. I think that this is a result of our crappy division, but it is just undeniable that this team is incredibly good at 5 on 5. Vegas got the benefit of playing against the same crappy division and finished 7th with a 53.29 xGF%. Minny was 13th with a 51.21 and we were 26th with a 46.09.
Meanwhile, we are bottom 15 in every underlying metric and are bottom 10 in most of them.
As I said in my last post, I think we match up against the Avs pretty well. In terms of current ability, I'm not sure that there is a better 1C to task with shutting down MacKinnon than ROR. We have played Colorado pretty well at 5 on 5 this year (although our actual results have outpaced the underlying metrics by a healthy margin) and Berube's style is a good counter to the strengths of their blue line. Stylistically, I think our system is the best counter to the game the Avs play. I think that the underlying metrics hold less predictive value than Dom does, but the Avs absolutely should be big favorites in this series.
Edit: Dom's statement about this series perfectly demonstrates my opinion that his analysis/interpretation puts way too much stock in metrics. He says that "If the Blues somehow beat the Avalanche, I may just have to delete all my accounts. I will never recover from that." Hockey is far too unpredictable for this statement to be remotely sound. In a low-scoring-event sport like hockey, no one should ever have that much confidence in their predictive models over a 7 game sample. Moneypuck has our chances of winning at 15%, which is the most lopsided series their model has seen in 5 years. And those odds are almost identical to the odds of correctly guessing the outcome of rolling a normal 6 sided di. No one should ever stake their professional reputation on those odds and doing so is an indication that your interpretation of the data isn't all that great.