Thinking about this gave me the giggles: 

-6 Beneath Contempt
-5 Contemptible  
-4 Loathsome 
-3 Hideous
-2 Miserable 
-1 Awful 
0  Bad 
1 Neutral / Average
2 Fair 
3 Good 
4 Excellent 
5 Awesome 
6 Incomparably Fine

If one wanted to have additional layers, we could initiate additional
categories, such as for 'originality.' A lot of games are original but
have bad gameplay or unoriginal but with good gameplay (I am reminded of
Ben Good's essay here about Omega Chess). Still other categories for
'fun-ness,' presentation, appearance. Categories could be optionally
listed according to ratings and categories with overall negative ratings
should perhaps be shelved into different sections of chess variants after
each receives a fair number of votes from the community of users (as
opposed to just members). 

There is one thing that disturbs me most of all about how people rate
games and I fear that there is sometimes a tendency to judge games without
playing them, trying them out. Sometimes, it is not necessary to playtest a
game, but I think too often a game is judged too much by certain
superficial aspects that have little to do with worth of gameplay (as with
books by their covers.) 

If one has a separate category strictly for rating 'gameplay' (as
opposed to other aspects), it could be a category that could only be
filled out after actually playing the game. If nobody is willing to play a
game, that would usually imply something about the nature of the game. I
suggest that as long as a game maintains a positive gameplay rating, it
not be shelved to the negative ratings section. Because a game can fail
every other mechanism or gradation of analysis, but if people enjoy
playing it, that's probably a pretty good test, in my opinion. 

'Confusing presentation, ugly appearance, highly unoriginal concept, but
amusing gameplay.'