Are we talking playoff success as well?
I think his regular season model has worked quite well because offence and skill has proved to be more valuable in the regular season.
I think he’s been pretty accurate the past few years considering his competitors.
- Santo_44
Let's put it this way, I guess - overall the teams that his model identifies are top-tier teams is not that far off and typically has them somewhat accurate. I think a lot of this has to do with the way his model looks at individual players' contributions and how those contributions would roll up to an aggregated amount. It makes sense - a team with more star players would have a better sum. Also a team with more star players is likely a better team.
From the Leafs' perspective, no model was going to predict the Andersen we got in net, the fact that Barrie didn't even provide offense, which is really his only value, nor would it have predicted long absences from Rielly and Muzzin - nor how the increased role of negatively contributing players like Ceci would have affected the outcome. So you can't exactly throw that in Dom's face about how his model was so wrong.
That said, he did completely fail to predict the pandemic, so we'd be wholly remiss if we didn't hold him responsible for this glaring omission.
In general last year the following teams outperformed his model by 5 wins or more (considering proration to an 82 game schedule): Philadelphia, Boston, Colorado, Edmonton and Washington. The following teams underperformed his model by 5 wins or more: Toronto, Anaheim, New Jersey, San Jose and by a whopping 28.6 points - Detroit.
So by that token you can say that over 60% of his projections were within a relative limit. That's not bad. If I'm looking at the above - Carter Hart was probably better than expected in Philly, and the team outperformed it's expected goals by a fairly decent margin. As Pat said, and I think everyone agrees with - Boston continues to defy logic in terms of declining play. No one thought Edmonton would play that well, and I still doubt they'll do it again this upcoming season. I also thought Washington would decline, but they did surprisingly well. I did think Colorado would be a good team, so I was surprised his model was low on them. I also think everyone kind of expected a real poor performance from Anaheim and San Jose, so I don't know why his model was so kind to those teams.