I think there are sort-of two ways to do this: (1) Ranking players within limited contexts (teams, seasons, peak years, etc.) and (2) ranking players overall, which means considering the entire career, the longevity, etc.
The thing is, we all have somewhat different ideas about what constitutes a substantially impressive period of time for (1). Obviously one season isn't enough. Two seasons isn't enough. Three not enough. Four? Probably not substantial enough. Five is getting more impressive. Six is pretty good. Seven is good. Eight is amazing. Etc. Call it "The Lindros test" -- if the player had considerable but not overwhelming team success, reached the level of NHL scoring leader/Hart winner, but only lasted six or seven years at prime level, can that player be considered for the Top-10? Guy Lafleur is similar, except he won a bunch of championships in the NHL.
Then, regarding (2), some people punish players with longevity but bigger fall-offs in the (usually) latter parts of their careers. Take Mark Messier vs. Joe Sakic, for example. Messier had basically nothing but elite player-seasons from 1981 to 1997, winning 2 Hart trophies and 6 Stanley Cups, both on two different franchises. Sakic also had incredible longevity from around 1989 to 2007, before injuries and age stopped him. But because Messier had no team success from ages 37 to 43, people seem to hold that against him, which they don't do with Sakic, who played on a much worse team (in fact, one of the worst of all time) from 1988 to 1992, and which missed the playoffs five of his first six seasons. Personally, I would say if anything ages 19 to 25 is a more vital part of the career than 37 to 43, but for some reason a lot of people can't get over the player's final seasons.
So, in conclusion, I dunno.
I will say that Wayne Gretzky from January 1981 to May 1988 is the greatest player ever, given a context.