I was reading this interview with Bungie's Jason Jones on Destiny, and this bit really jumped out at me:
And then it evolved, as we evolved into sci-fi, because we felt the pure fantasy was missing all these things we loved. Literally, explosions. It's kind of a joke, but it's not really a joke. Explosions are awesome, just as a gameplay device and all kinds of other things. People understand them, they're a really easy metaphor for projecting power into the world.
There are two aspects about this quote that struck me.
First of all, metaphors are an incredibly powerful aspect of game design. Game design often (I would argue virtually always but that's an argument I'll try to make another time) involves melding mechanical aspects (rules) with fictional aspects into one coherent experience. Metaphors are the most basic and most powerful technique to do this. My personal hierarchy of interactive storytelling techniques goes something like: metaphors, world-building, actual stories.
(I'm not using the term 'metaphor' in a very precise way: 'analogy' would probably work as well. But most people roughly understand metaphor.)
Second, I've been working on mobile and casual games the last couple of years, and I really learned to appreciate and respect proven techniques, sometimes referred to as 'clichés'. Especially when we tried very hard to question, reinvent or abandon all well-known techniques in one particular project.
Using proven techniques often helps you understand and explain the game, which benefits players, the team, and yourself.
If you cannot explain the game to others on your team, you will probably encounter friction because it will be harder to imagine why it will be fun, and how you will build it. And really, that goes for yourself as well as for others on your team. "If you can't explain it to a six year old, you don't understand it yourself" and all that. (Although as any game designer knows, you can have a very clear vision and still not be able to explain it sometimes.)
If you cannot explain the game to players, and I would count the press as players in this case, you have a marketing risk. It's not impossible to make a successful game that's hard to explain, but it's a lot harder. And these days every developer should think more about marketing. (I always hated the demand to "explain your game in one line", but I changed my mind as soon as marketing became something I had to think about myself.)
(Should I mention that I think that players and developers learning what games are is how the medium evolves? Let's not put too many asides in parentheses.)
I'm not arguing you should use clichés, but that you should understand a rule (or proven technique) before you try to break it, and that originality purely for its own sake is a bad idea. (And those rules are themselves hardly original.)
In the context of having to ship a certain kind of game into the marketplace with a limited amount of resources, I prefer a risk-oriented approach to game development. Breaking rules you don't understand is risky.
Games are often criticized for being unoriginal. And it is not a bad thing to criticize: games are often unoriginal for bad reasons. But there are, in fact, good reasons why certain things, like explosions, come up again and again. (That doesn't mean they are essential, or that we will never find a better way.)
Update: In fact, after thinking about it a bit more, the power of metaphors and of proven techniques (patterns, elements) are probably two of the biggest things I've learned about games in the last five years.
Update: From the same interview (which I haven't finished reading yet):
And classes are a great short-hand so that when I look at you, I can have some expectations about what your abilities are and how you're likely to behave in the world, and what kinds of things I might depend on you to do, and a lot of games have done this very successfully. I would say that one of the origins of class is that it gives people some way to look at each other and talk about their abilities without actually talking.
Remember, this is one of the best and smartest studios in the world, who have enough money and creative freedom that they designed an entire fantasy universe then threw it away. (I was at their GDC art lecture, it was crazy to see all the art they did only to then go "Naaaaah let's do science fiction". Because of explosions.)
Also the classes they have, while not groundbreakingly original, are already different from the classical Holy Trinity that people have been talking about changing for 3 decades.
It's not impossible to use something else than the classes Bungie picked. They have constraints and biases others might not have. They have to please an enormous amount of players, including their own fan base, for one. But there are some good reasons why Bungie chose this approach.