Quoting: Tintin
I think the reasons why Dom's modelling is sceptical regarding the Blues is because they've been a terrible team in terms of xG% going back to the 20-21 season, like bottom third every single year.
Therefore the model has no banked confidence in the Blues being able to sustain anything close to its current form, the current lack of confidence being exacerbated
While the Blues have been spending the last two seasons and some change being a terrible 5v5 team getting carried by unsustainably good shooting and ridiculously good goaltending (see Husso), Minnesota (and Toronto for that matter) have consistently performed very well in terms of xG% which is why the model is more confident in them making the playoffs.
The legacy of model trust / distrust is also a variable explaining why playoff percentages doesn't rise and fall at the same rates for different teams in-season.
In any case, the reasons above are likely to explain a lot more as to why Dom's model doesn't think the Blues are making the playoffs than a not so secret conspiracy on part of Dom to intentionally knock the Blues down a peg or two.
If Dom were thumbing the scales, we would've found that out already expressed in terms of variable input vs modelling output discrepancies.
As far as I know, no one has been able to report and provide proof of such a claim however.
A few quick points of order:
- The Blues were +29 5v5 (and +36 at even strength) last season. Not sure how that qualifies as terrible.
- The Blues have shot have shot 10.2%, 10.3% and 12.0% the last three seasons. Considering they switched to a style of play that prefers to play for higher percentage chances instead of volume shooting (another factor that dulls their xG numbers), that sounds pretty darn sustainable. They simply have had and still have a lot of good shooters, and a lot of good passers, which allows the good shooters a lot of good looks at goal.
- It's kinda hard to point out specific output discrepancies without knowing the formula he has built the model on.
- Detroit, who has been brutal the entire time that his model covers and now stand well behind the Blues on xG share (44%) are higher than the Blues (20%). Vancouver, who have been rough for the last two seasons, have a 46% xG share, but he has them 27% to make the playoffs. Winnipeg who have been pretty meh in recent seasons are also below the Blues on xG share... 56%. The Isles, who were bad last season are below the Blues in xG share this season... 30%. Nashville hasn't been anything special the last couple seasons and has a lesser record, but hey, their current xG share is <1% higher... 36%.
I fully get the whole xG thing, and it would make sense if the playoff percentage disparity was a lot smaller. It doesn't explain a 60% edge for the team with a lesser record. Especially considering the Blues have 49% 5v5 xG share this season despite a bad shooting slump during the losing streak (and uncharacteristically weak special teams thus far). That 5v5 xG share is two whole points better than Colorado's this season. And Minnesota's is 50%, hardly anything to explain the wild disparity. Meanwhile, everyone else doing these projections with a model has the Blues anywhere from 26% to 53%, and the 26% is a big outlier. All the rest have them between 38% and 53% (or 2x-2.8x higher).
It's not adding up, my friend. Seems fairly obvious, too. He makes no secret of his disdain for the Blues, openly admitting that is due to how poorly he bets on them, and their fans. It's a nauseatingly constant theme.