I think there is a risk that "recommend a game" questions are "bad subjective".
Can I suggest that we recommend that any "recommend a game" question should be tagged recommendation and should try to limit the number of possible answers, perhaps by answering the following questions
How many players is it for?
Do you want a competitive game, a team game or a co-operative game?
How long do you want a game to last?
How complex would you like the game to be - from "playable by pre-literate children" to "need to read multi-hundred page rulebooks"?
And probably other questions that will come up in the answers - can others please suggest other questions that people might use to limit the results.
In particular, I propose that anything you can do with just an advanced search at boardgamegeek is not a sufficiently specific question.
From the looks of it, one common type of recommendation question will be "I like X. What are some other similar games?"
When the question is about mechanics and the design of the game in this way, I'm not sure your list of questions is appropriate. The asker won't necessarily care how many players it's for or how long it lasts - they just want something with similar mechanics to game X.
I also think we should keep a close eye on the meta-discussion of this issue at Gaming, as both sites are going to have similar patterns for recommendation questions.
I don't think there's a problem with suitably specific questions asking for game recommendations (e.g. this one). I know there's no one definitive answer, but this is game site, not StackOverflow, and there are going to be a lot of people interested in recommendations for good 1-/2-/4-/10+ player board/card/mix-of-board-and-cards games to play with teenagers/blind players/over-competitive spouses in-the-car/on-a-plane/in-a-tent.
Marking them as CW just discourages people from asking or answering these useful questions, which I don't think is what we want.
This discussion is repeating itself on every friggin SE site. I personally believe that game recommendations of sufficiently specific nature ought to be on topic.
They represent an central desire among gamers - to find new games they might enjoy.
They also allow new and casual members to bring in their own experience and thus foster the community
For these reasons I believe that these kind of questions should:
Be allowed on the main site.
Remain outside CW as that actually discourages answers, because you don't get anything for it.
Be answered with only one game per answer so the voting mechanic can be put to good use.
Be filtered and overly general inquiries closed. I agree with closing "Best games of 2010" as it is mere chit chat and not aimed at solving an actual question.
But in any case: I strongly recommend to vote on this issue as soon as possible. I've seen this question drag along for weeks on other sites. We need a decision soon or all hell will break loose.
I've been involved with the game-rec discussion on gaming.SE. Endless talk and indecision. I eventually left it be because I got sick of it. IMO we are better off allowing these questions and to spare ourself the flamewar. If they really turn out to be so disastrous for quality as some think then we will see and the SE community will finally have it's answer. But until then our job is to try and make this site as good as we can - not a copy of the others.
The question has generated some discussion about the suitability of the question and two sides of the argument are quite reasonable.
I worry that many people will come to this site looking for ideas about what game to buy next, and the community are all too happy to help answer the question. My concern around closing the questions too quickly is that we stifle the usage of the site and it fails as a result.
Yes, this sounds like a good idea to me. I think this may well be one of the major difficulties we encounter, so we should be strict in recommending an edit of the question/voting to close whenever we feel this recommendation is being violated.