I've been thinking about how I come up with the designs for my games. It's tricky to do that sort of introspective analysis but I think I've got a novel handle on it. I think it's possible to think of each phase in the design process as a person with a different personality. See what you think of this:
Phase 1: Imaginary friend
In this phase I play the game in my imagination to get a feeling for how the game will look and play. I am trying to understand the core game play and hopefully spot any major defects in that game play. I want to understand what's exciting and what would make me want to play it. It's tricky to do this but over the years I've got better at it and be reasonable certain that when I finish this stage I'll have the core of a good design. That doesn't mean that this idea will form the basis for a fun game though, I have to implement it before I can know for sure. I have a lot of games at this stage and I'm often thinking about them. When my wife asks what I'm thinking about, when I'm staring off into infinity at dinner time, it's usually a game I'm playing inside my head. At this stage the game is like an imaginary person who you are rehearsing a conversation with. As we all know things which often sound OK in your head don't always work out so well when spoken aloud which brings us onto the next phase...
Phase2: Introverted Geek
Once I'm reasonably happy with the game I'm playing in my imagination I make a working version of it as quickly as possible so that I can see how it plays for real. Because I'm not going to let anyone else see it I don't worry about graphics, sound, tutorials etc. The point of this phase is to prove to myself that the core game play mechanisms work and that the game is fun to play. If I'm having fun with this version of the game then there is a chance other people will too! The important thing to realize though is that at this stage the game can get away with being a really bad communicator. That's to say that a lot of what is happening behind the scenes isn't communicated to the player. This is OK though, because I know what is going on so I can understand it. It's as if the game was a really introverted person, if you know them really well they can be a lot of fun to be with, but it takes a lot of effort to get to know them. I won't move onto phase 3 unless I'm having fun with this game.
Phase 3: Ambivert (look it up!)
By now I'm confident that the game can be fun to play but I'm yet to prove that it translates to something which others will be interested in. So the challenge is to turn my introverted game into one which can communicate what is happening and communicate it's needs to an audience. To do this I start to add better graphics, graphical effects, various HUD elements such as score, lives etc, sounds, all the things which players need so they understand what is going on as they play. In other words the game is becoming more intuitive. Note at this stage I'm assuming my players are experienced testers, and I'm going to explain the game too them, so there aren't many tutorials and the game isn't polished. In this phase the game is like the sort of person you get on well with once you've been introduced by a mutual friend and they have started the conversation for you. But they are not the heart and soul of the party yet. I won't move onto phase 4 until my testers are having fun and usually there is quite a lot of iteration in the design and art.
Phase 4: Extrovert Party Animal
This is the final phase of the design process. The game has to stand on its own feet now and go out into the world and make friends without me introducing it. It has to shout out its merits and not be afraid to talk to strangers at parties! Tutorials must be in, splash screens, fancy graphics, all elements of the game play must be completely transparent to the user because I won't be there to explain what is going on. There can be bugs in the game but the important thing is that complete strangers have to have fun with the game without help from me. In this phase the game is a true part animal, it is like that person who when they walk into a party suddenly everything comes alive. If the game passes this test then it's good to go into production.
Not really a design phase but it's worth mentioning that once a game makes it to extrovert it still isn't finished. What we have at this stage is a short demo, proof of concept, which anyone can play and enjoy but we will still need a lot of additional content and testing before it's ready to be released as a commercial product.
So that's how I see my design process working. At the end of each phase a game can fail the test and I think it's important that if it fails we say a sad farewell and move on. Each stage takes progressively longer than the previous and the investment in effort and money is greater so there are less games at each stage and the stakes higher. Not every game which plays well in my mind will be able to make the long journey to extrovert and it's important that I don't see that as a failure. After all we can't all be extroverts. Probably only one game in a hundred, which I play in my imagination, actually gets released to the public.