I'm having trouble lately with getting a ways into making a game and realizing I don't have enough mechanics to keep it interesting. This made me wonder, do you get most of your design done up front or do you start with a vague idea of what kind of game you want and grow it out organically? And similar but related, do you try to milk all you can out of one mechanic before adding another one, or do you try to mix them up?

when i start something i make sure very early on that i can fully explore the core mechanic of the game. its the foundation of the gameplay experience so the core mechanic needs to be very “pliable” so it can be expanded on a great deal. thats why its best to keep your core mechanic very non specific. 

like with SMB it was platforming and platforming was broken down into, running, walking, jumping, horizontal and vertical dodging, wall jumping and time. its easy to design many many things around this very basic core mechanic.

and example of a very hard mechanic to design around would be something like the portal gun, its a lot more limiting in what you can do with it and required really good level/game designers to expand it into the game we know and love.

an even better example of a hard mechanic to design around would be the world spin in fez. if you actually look critically at its spin design you will notice that the world spin mechanic is extremely limited and can actually be designed around much at all due to the players in ability to memorize how a level looks and how it might look once spun, thats why when you play you usually just keep spinning till things are correct. to fix this phil designed around it and pushed exploration as the core and the world spin as just a visual mechanic and eye catching selling point.

explore your foundation as much as possible and if you hit a wall introduce another mechanic that stems from or complements your core mechanic. and if all else fales push your games theme more… killer 7 did it :)