Mrrph. It's hard to articulate what I want to say about this.
Some ideas are more generative than others. The word "generative" is hard to define exactly, but when a relatively simple idea gives rise to a huge variety of interesting and different effects, we say that the idea is generative. Go and Life and Chess for example have some simple but highly generative ideas.
I think that, for most people, games are fun while they generate new experiences or new problems to solve. A new idea, ipso facto, creates a set of new experiences or new problems, so it makes a game fun -- for a while. But most ideas aren't all that generative, so after a while the player has experienced whatever it is the new idea can produce, and after that the game gets less fun.
A good roguelike uses a combination of several different relatively simple ideas to generatively provide a huge variety of tactically different experiences or problems to solve. Different attack forms, different monsters, different equipment, different buffs, line of sight/targeting, speed, hunger, healing, resource management, etc... These things are not little subgames of their own, these are all interlocked. They are generative because each can change the meanings of the others. Each of these types of content provides gameplay (that is, new experiences and new problems to solve) proportional to the product of all the rest.
A new idea in a roguelike game, IMO, has to be evaluated in terms of this type of generativity. If the idea is generative, then it provides content proportional to all the other content that already exists. An idea that isn't generative may add content, but never multiplies it.
And that brings us back to 7DRL's. Some of these games have fun ideas. But sometimes the fun idea wouldn't be generative with the rest of the elements of the roguelike. Some of these games have original ideas. But sometimes the original ideas aren't as generative as the ideas that get left out, so we wind up with a "flatter" gaming experience. If the final combination of ideas is generative as a whole, then the game gets deeper and more interesting the more content is added. If the final combination of ideas isn't very generative, then it may be fun for a little while but no matter how much content you add people will eventually get bored of it.
Bear
