I've been committed to a game idea I've had for years, literally over 20 years.
I've been writing the code for an engine to be run in Chrome for the past 4 years in Javascript.

I'm looking for feedback and comments on my proposed development strategy:

1. Finish the engine.
2. Begin programming a simplified economic simulation using thousands of AI.
3. Purchase a powerful desktop to run the simulation.
4. Create a website for users to view the simulation with the hopes of generating interest.
5. Create game content that exposes the game's concept and play style to players.
6. Allow players to play for free on a website.
7. Appeal to developers to see if anyone is interested in helping with the project using the Cathedral Model for open source. ( I'd use this for insights into interest in the project, goal would be 2 casual contributors )
8. Begin developing a client for release on steam and advertise it on the website ( better graphics, more power ).
9. If by this point there is sufficient interest in this game say ~5000 people ( no idea ), launch a kick-starter campaign to raised funds to significantly expanded content, In which case I could responsibly commit the time necessary.

What do you think?  Thoughts?

good to know.  Thanks

20 hours ago, swiftcoder said:

This feels like you are putting the cart before the horse. Items #8 and #9 in your list seem like the first steps you should be taking (ideally before investing years in engine development).

I've been told that the time when a development team could successfully raise capital from a kick-starter campaign without having an actual game is over.  That now you need to have a playable prototype for people to try before they're likely to commit.  

20 hours ago, swiftcoder said:

Have you considered short-circuiting the process of finishing the engine by using an off-the-shelf engine?

I shouldn't call it an engine.  How about a custom library that allows the functionality I'll be needing for the game I'm making.

20 hours ago, swiftcoder said:

Consider leasing cloud compute rather than investing in an expensive desktop. Desktops don't scale very well in the long run.

So the simulation server(s) will easily use 5GB of ram or more, I'm anticipating more.  I've looked into costs and the cost of hosting a cloud computer for 1 year with those requirements seems to be about the cost of a high end desktop.  Do you know of any good sites for hosting?

20 hours ago, swiftcoder said:

It's hard to convert free users to paying users. Consider the other way round (i.e. paid alpha/beta access).

Others have said the same,  Hmmm.  What do you think about maybe providing a limited gaming experience available for free on the web, and then if you want to play the full version you'd have to buy it on Steam? 

Thanks for the feedback, any other ideas?