I think the prospects we have picked as "reaches" reflect the scouting staff's opinions as to which prospects might turn out to be better players than the rating agencies' forecasts.
If the Senators draft a player in the later rounds who turns out to be a star (e.g., Stone), then I wonder if the scouts were that smart because if they projected him as a potatntial star should they not have drafted him much sooner as a "reach"? Or were they just "lucky"?
Lucky. Any 3rd round or later pick that makes it as a regular NHLer much less a star is some degree of luck. Especially CHL players which have the crap scouted out of them.
As for reaches, there can be a couple of reasons. A big one is view bias, where a player gets viewed a lot more due to the team already having a player on a team they are really tracking (i.e. our fascination with ND, although it has served us well). Another is teams trying to build an identity (See Boucher), which I disagree with, draft the most talented players and let that build your team identity. This hurt Edmonton for years as outside of the top of the first round they drafted for identity not talent. Reaches also happen when a team is trying to fill a hole in their team with a high first rounder. I have no issue if a team outside the top 15 drafts for organizational depth issues, but reaching because you are short RHD for example tends to backfire. Also falling in love with "character" players. While I think character is important, its junior level hockey. Drafting character over talent is a road to the lottery. 99% of NHL players, even role playing character players, were elite point producers in junior and often in the NCAA. And lets not foget some some people just want to draft different than consensus to prove they are smart .....
Truth is, rating agencies have always been the best barometer. They are independant and enough different eyes that biases come out of the equation. Some teams can look like drafting geniuses by getting two lucky picks outside of the top 10, but there is just not enough data to rate them vs an agency which rates 300+ players, as opposed to teams that are just rated on the 7 they pick.