[PAYWALL] Modèle Dom’s Athletic : Canucks 91,7 points, 49% pour les playoffs

1 Comment

1. elrizzy 2 ans ago

   Some things from the article:

   https://cdn.theathletic.com/app/uploads/2022/09/25173734/screen_shot_2022-09-25_at_5.21.22_pm-1024×788.jpg

   His summary:
   > »The Canucks are betting on the core they’ve put together over the last few years and hoping the Boudreau effect doesn’t wear off anytime soon. This could be a defining year for this team, whether Vancouver shows that was the right choice and this is a deep enough team for a playoff run or their play just highlights what needs to be addressed to actually get there. Neither would be entirely surprising, but the certainty in either direction is what this team needs. Otherwise, the Canucks stay stuck in the middle of the basement dwellers and true contenders in a division they should be trying to take advantage of. »

   * 17% chance of 100 points
   * The model *really* hates the defence. Like absolutely despises it to the degree that the entire defence is worth less wins as a whole then a Cale Makar or the 4 other top defenders.
   * Schenn and Myers are seen as « not top 4 » in his model
   * The teams true talent ratings under Boudreau have increased, but doesn’t think it’s as simple as taking the « Boudreau run » and extrapolating from that.
   * The « wildcard » is how Pettersson and Hughes do
   * Thinks Boeser and Horvat have pretty much shown what they can do in the NHL and are rated fairly, though expects a bit of a bounce back from Boeser
   * Calls Horvat a « low end 2C on a contender »
   * JT Miller has a rating of 2.4 wins this upcoming season even though he did 3.4 last season, is dragged down by previous season’s play
   * OEL could be « top pairing caliber »

   Please remember these findings are from a mathematical model Dom uses for sports betting, and not personal opinions. His model uses data from multiple previous seasons, so it is not as « impressed » by a short positive run of play. Models can be flawed and don’t account for everything, always. Last year’s model was around 92 points as well IIRC.