Highly unlikely the Caps win a Cup in 2018 following NHLe those prior years so I'm good riding out any expected regression since then. (There do seem to be some discrepancies between this analysis and HockeyProspecting's re-drafts so I'm not sure which to go by but this one has more open data.) This comparison even understates Washington's top-ranked success over that period by not including goalies. Of course they were due to regress but I doubt it's because they decided to suddenly start drafting low ceiling players. They shot for upside with Lapierre & Miro. McMichael was a finesse play. They went D a couple times early when there wasn't great junior production but...eh. The failed Mantha trade has been a bigger issue given the cost and expectations in what should have been at least a fair top sixer. Additional factors like the flat cap thanks to COVID also hasn't helped and an inflexible coach that hardly had time for less systematic freelancers also didn't help avoid becoming too same-y. Maybe via NHLe they'd have more regular season pop the past couple years thanks to one more productive piece. But I'll take the events and process that led up to a Cup any day over trying to be paper champs and sort of taking that earlier run for granted. Whether MacLellan has improved their draft process is fair to wonder but it's the same people running it. Unless one believes McPhee brought special sauce to the table it's more likely just expected regression. Hard to sustain hit after hit for 15+ years. Also more difficult to finesse rough edges without dominant play-driving headliners.
It seems like NHLe's success is largely due to a handful of players, mostly Russians and smaller wingers. That seems to be the basic advice: don't sleep on either. But I'm not sure that's a viable forward-looking strategy to the point of full-scale adoption even setting aside the glaring goalie factor. The Russian Factor is arguably as dead as it has ever been despite ongoing geopolitical tensions so I wouldn't count on it accounting for much of its margins going forward. The cat is largely out of the bag after Kucherov, Panarin, Kaprizov, etc. Three Russians went top 12 this past draft thanks to Arizona and there hasn't been a significant slide for some time. Smaller productive juniors do still get slept on but I also wouldn't count on consistent massive gems to the point of ignoring other player types. Judging by the early returns in some more recent drafts one thing that sticks out in terms of categorical over-performers is re-entry players rather than those two categories.
I buy the premise of at least shooting for upside past the first 40-50 picks when it becomes a relative no-brainer. I can also buy NHLe as a better indicator of draft slot over-performance but beyond that are some questionable weighing effects that would make me hesitate to use it as a primary source with more at stake. Not ranking Mikko Rantanen in the first round in 2015, for instance, begs for further refinement. He was a blue chip lock...end of story. If data suggests otherwise to that extent I definitely question it. Having Bracco & Fischer from a loaded NTDP team higher (Bracco significantly so) sticks out as perhaps struggling to weigh team effects. There are multiple instances every year in HockeyProspecting's first round re-drafts where common sense suggests some factors may be weighed problematically. Taking goaltenders out of the equation altogether makes it not a viable singular source.
It's easy from the sidelines to suggest going with a supposedly superior performing model but any team with serious interest in this area should just roll their own. That way they can take ownership of it (and, again, we don't really know). I would be more surprised if anyone doesn't employ something along these lines. Teams like Colorado, Carolina and Winnipeg often seem informed by similar lines of thought at least and I'm sure there are many others that at least consult modeling projections to help craft strategy.
www.actionnetwork.com
