With samples this small those models are also going to be noisy given the assumptions they make about shot quality.Those naturalstattrick numbers need to be taken with a big grain of salt.
They had the canucks like 5th in XGA.. some private models had them bottom ten. And based on the eye test of what we’d been giving up in the middle of the ice up until the last game against Anaheim, there’s absolutely zero world where we are top 5.
That said the Canucks are also top five in CA/60. There’s been some glaring breakdowns but on the whole they are limiting shots against really well. Could also just be some bad luck in the number of plays that have turned into those ugly GA.
Plus the difference between their xGA and their actual goals against is basically just the difference between Silovs and an average-ish goalie. To my eye that’s more on Silovs than the Canucks giving up higher danger chances than the xG models suggest.