The "Quantum Ogre" is a highlight from - and a codeword for - discussions about player agency and encounter design in RPGs.
Relevant links:
http://www.korpg.com/blog/the-fallacy-of-the-quantum-ogre-argument/http://hackslashmaster.blogspot.com/2011/09/on-how-illusion-can-rob-your-game-of.htmlhttp://www.emacswiki.org/alex/2011-09-08_Player_AgencyEssentially:
The DM has an Ogre encounter prepared.
The DM gives players a choice: Forest A or Forest B, alike in everything but name and location.
The DM intends to drop the Ogre encounter into the game no matter which forest they pick.
For some, this is akin to GM railroading, obscured by a meaningless choice for the players. It is seen by others as an oversimplification, or an accusation of "ur doin it rong".
My moderate take on it is that this would probably be less of an issue if the encounter wasn't a blind one. Which is to say: the GM can drop clues or rumors to establish that *something* is up in Forest A (without screaming out "an Ogre is here"), and then giving the players a motive to trigger the encounter (and an opportunity to either plan for or avoid it) rather than simply removing the reward/penalty for choice and simply making an encounter assured.
It seems like an attack on GMs with limited time and resources, but in actuality, it is more of an indictment of giving prominence to non-story encounters (whether random or prepared, I feel, makes little difference).
Now if that "quantum ogre" is (or holds) a clue for the overall story, a stepping stone in the mystery, then yeah, it is important to the story, and the GM needs to get that clue out there. But the GM should also prepare 2 or more *other* vectors by which to deliver that or a similar clue.