How to Develop an Idea

This post examines taking an abstract idea and turning it into something useful. It is a many-step process.

1. Record the idea. Depending on the idea, make a note of it, or put it into an appropriate media (sound file, drawing)
2. Examine the idea. 
• Is it viable? Is there some simple thing that makes it unworkable?
• What type of idea is it? My game-related ideas generally fall into one of two categories: World and Mechanics. World ideas include characters, settings, and plots. Mechanics are the specifics of how a game works.
• I will be focussing on Mechanics-related ideas in this post. World ones will be done in a future post (probably)
3. Flesh out the idea; build on it until you have something workable. Start as bare-bones as possible - what do you need in order to get through the first turn of the game?
• Using Shadows over Japan as an example; I needed all the pieces to set up the board, as well as the board itself, the cards, and tokens. In addition, I needed rules dictating how the board was to be set up, and at least a basic turn structure.
• Answer the following questions: (they may seem familiar)
• How do you conduct a given turn?
• What is the overall strategy?
4. Test and Review
• Once you have a basic idea of what you are going for, review it again for viability.
• Are there any crippling exploits that ruin the experience?
• Is any strategy particularly overpowered or underpowered?
• At this point however, you do not even necessarily have a win condition well-established. The game length is probably not yet determined, and gameplay itself is still in the Alpha (subject to massive change based on testing)
• Ensure that the win condition fits with the overall strategy - if the game is based around massing up resources, ensure that the player who does this best will win. 
• Again using SoJ as an example, I initially had a game length that took more than 3 hours to complete; based on this, I revised the turn structure and number of turns to something more appropriate. I also found it was annoying to move the turn tracker each turn, so I revised that to the card-per year method that it has now.
• You do not have to play out the whole game; if it starts getting boring, or you already know what you definitely want to change for next time, end the game early, do your revisions and start again.
5. Go into Beta
• Once you have at least a couple of games played to completion in a row without revising the rules, you are probably ready to go into Beta.
• I would recommend making a proper prototype at this point; return any borrowed pieces, and get a dedicated set for this game. If required, look into publishing companies to make it such as Game Crafter to produce a prototype. 
• Play the game, repeatedly, and with as many different people as possible. It is important to get many different perspectives. If possible, have people play a game without you there (or possibly without anyone who has played it before) in order to see how clearly your rulebook conveys its contents. 
• You may also want to record either a video of people playing the game, or just the relevant numbers involved in the game (for example, the income you make each turn).
• The amount of testing necessary depends largely on how complex your game is. Deeply strategic games with little chance involved need to be played hundreds of times by highly skilled players in order to ensure proper balance. 
• Simple trivia games (i.e. Cranium) do not need to have their mechanics extensively tested, but instead need to have their questions reviewed, in order to determine what percentage of people know each question, according to the demographics relevant to your game (or you could just wing it and hope for the best, as seemingly most trivia games do).
• Many games offset deterministic strategy and the AP associated with it by introducing chance into the game; chance comes in many forms, but it helps move some games along. Not being able to agonize 50 moves in advance helps some players finish their moves faster. If chance is a factor, determine what your probability curves should look like - do you want a lot of chaos, or just enough to keep high level players guessing?

What should you do from there? Your guess is as good as mine. See if you can get it published; either through a professional publisher or DIY. I have remained an amateur in this industry, as I am more interested in the design decisions than the business ones. 

But more on that later.

-Colin