Background noise is annoying >_< I used to be able to read so well, totally ignoring everything going on around me. Now the exact opposite is usually true. Why?
Release info on a central site days before it is available by any other means. Data-mining and model rendering means that public testing is equivalent to a full release. It is harmful in itself itself to show things in a way which suggests that knowledge of new abilities or mechanics is worth displaying or talking about them in a poor way. Even if this attitude did not exist before, it will create it, and the attitude will be self-reinforcing despite being contrary to the state of a game. So even when dampening the competition-based factors which encourage this knowledge . . . primacy, the improper release and approach to new information can such habits to exist in those interested.
Accumulating information and releasing at the right times means less information at other times. Inconsistency will kill a habit. When a game relies on, tolerates, or even encourages an obsession with new details and the edge of change, it will not reduce the alteration of the spirit of the game by third-party sites and will not benefit those who are trying to enjoy the game in its present form, neither in the social environment it creates nor in the choices it will cause the game's designers to make, consciously or unconsciously, in response to the most insistent of players who do concern themselves with the flux and newness of a game. This mutual or self-delusion by avid fans and the creators of a game can lead to the game's . . . dullness, I suppose for those who are not interested in the grind.
Model-viewing screenshots tagged with the name of a site. Depends on having the object model in game files released to players.
Describing upcoming changes or newly released information in specific terms, or even copying all official information releases on formal or informal channels to encourage players to use a certain site as their primary source of information.
For example, a new boss is released on a public test realm. A third-party site uses a model viewer to create a screenshot of the boss and tags it with the site name. They data-mine the game files describing the boss's abilities and post all data on their site.
The following would make less people interested in that presentation by the third-party site, for a game developer interested in avoiding the knowledge-obsessive approach to their game. Days before the test patch is released, create a story about the boss with a 'role-playing' attitude, that defines who this boss is, why they do what they do and why they are important. This is accompanied by proper screenshots of the boss in certain points in the story, from someone who is actually good at photography and doing angles and background and so on (in other words not cr*p). In this way, the 'lore' of the game is created as the game is developed, and becomes important because of its realism and the emphasis on story from being the first information released on this boss, and allows the interpretation of a situation in terms beyond the format of an encounter or its rewards, prior to seeing any 'numbers' or encounter specifics. Especially when the story of an instance allows not fighting this boss, this makes knowing these specific numbers in advance unimportant because the player can have already decided that killing the boss is not what they want to do.
Mmm there's other things I thought about, such as being able to make all or at least many factions like you or to make them all hate you, and a 'long-term' reputation contrasting with a 'short-term' one that leads the long-term one such as when you betray a faction and they wait disapprovely for you to redeem yourself, and players being treated as a group with trusted players grudgingly making up for the untrusted players as if vouching for them (group effects or reactions is a theme as mentioned previously) . . . but don't know too much to say. Was something I wanted to talk about a few days ago but I forget. o_o Also was thinking recently about 'how to keep the boss from just killing the healers' as well as well as how to make sure someone CAN be killed by reducing the effects of healing when it is done repeatedly . . . this is not paradoxical if a boss's optimum 'strategy' is to kill someone in less time than it takes for diminishing returns on healing to take effect on a soft, healy target. 0.o