Every time you design a game, you have to make several conscious decisions in response to certain questions. How will the player interact with the game? Is the game 2D or 3D? Does the game have a skill/leveling system? How does that work? What should the HUD show? Are there doors in the game?
Similarly, you will make several decisions I consider to be unconscious decisions. For example: Gravity always points downwards. Visible walls are solid. The game will always tell the truth to the player. These things are so ingrained in conventional game design that we don't even think about them - they simply are the way they are because we've always done them that way. But I think it's unhealthy to take these things for granted. Like a child wanting to understand the world, it's important that we ask why. When you're designing your next game or iterating on your next idea, I'd like you to consider the following:
Find an unconscious decision where X happens.
Ask Why does X happen?,
and then ask What if Y happened instead?
Taking various common-sense game design conventions, and rejecting them outright can give rise to some really interesting and innovative ideas. Even things as fundamental as time and space can be subverted to truly incredible lengths. I've listed a few notable examples here:
VVVVVV - What if gravity doesn't always point down?
Portal - What if space is non-euclidean?
Monument Valley - What if space is non-euclidean, and gravity doesn't always point down?
Antichamber - What if space is non-euclidean, and it changes behind you?
Braid - What if time doesn't always go forward?
SUPERHOT - What if time doesn't move at a linear pace?
Fez - What if our 2D game actually takes place in a 3D world?
Miegakure - What if our 3D game actually takes place in a 4D world?
The Witness - What if we don't tell the player how puzzles work?
Hidden In Plain Sight - What if players aren't told who they are on-screen?
Rocket League - What if our football game doesn't have humanoid players?
The Stanley Parable - What if the characters in our game are self-aware? ...and space is non-euclidean?
[Ed: apparently I really like non-euclidean geometry]
This is not a new approach to game design. The examples above are proof enough of that. But it's an approach I'd like to see more people taking to their work. If we as developers simply rest on our laurels then we'll stagnate. There's no shortage of shooters, but I'd far rather play something like SUPERHOT that changes the medium than play yet another Call of Duty.
So, fellow developers, I call out to you. When designing your next game, what twist can you bring to the table?