So thinking about rogue-likes made me wonder about this.
So Joe has said when a player checks a walkthrough to breeze through a challenge they didn't want to do it's probably a result of bad game design.
And as far as I can see in order to be a really competitive high quality nethack player (and other games as well, lots of rogue-likes, Kingdom of Loathing, other looser defined examples I'll mention at the end of this statement) you have to draw on the collected experiences and discoveries of lots of other nethack players. And if you are that high-quality of player, they'll of course be drawing on yours as well. There's a similar phenomenon I believe in games like Zack and Wiki, or Professor Layton, where you ought not NEED the help of other players, but you're probably in the same room as friends while playing, and it just makes the game a lot more fun (which games should be, thank you John Kovalic, you can stop repeating that) and uses different skills (socializing and sharing ideas with other people versus hacking your way though alone), and using different skills is a good thing for games.
So what makes those things good and running off to gamefaqs wrong (no matter whether it's because the designer made a mistake or the player is making one)? Sure, there's much more of a reciprocal thing going on the "good" examples, except even walkthroughs get expanded as readers volunteer their own information that they've spaded out.