Good things for prototype systems:

  • Based on a high level programming language
  • Smallness
  • Cross platform
  • Mature toolchain to simplify asset generation
  • Ease of moving to a production engine

Making games with little time, people or costs →

How to Prototype a Game in Under 7 Days: Tips and Tricks from 4 Grad Students Who Made Over 50 Games in 1 Semester

A good rapid prototyper would realize that failure is ok! That's what prototyping is for, so go crazy! If you fail, there will be dozens more, and chances are, you'll learn something anyway. By embracing the possibility of failure, rewarding experimentation becomes possible.

Eg: plant game


  • Low tech forces focus on the ideas - this is really important I think
  • Ease of running/viewing/playing
  • Use of commonly installed things - web browser/flash plugin, blah.


  • Have to run everything in a virtual machine/interpreter so a bit slow
  • Not much in the way of graphics features (not really a problem for prototyping)
  • 3D possible but is not really the focus in these systems, so it can be awkward

processing →

haxe (compiles to swf/js/…) →



  • Prototyping in the same system you produce the game in saves time (arguably)
  • You can have access to the same features in the prototype as a final game


  • Complexity kills creativity - game engines tend to have a lot of complexity
  • Have to install stuff, more chance of platform incompatibility

Blender game engine → blender game engine notes

  • Blender is a good option as it's the same tool used to create the artwork - there may already be a familiarity
  • Scripted in python which is a good prototyping language

Fluxus → fluxus notes

  • Fluxus is scripted in Scheme which is a good prototyping language
  • Deliberate perversion of games industry 'best practices'
  • Not working on windows
  • Difficult to install
  • Designed for fast graphics prototyping at this point - rather than gameplay prototyping
  • Frisbee, which is a functional reactive programming/animation system for fluxus, might provide a much better, smaller, more creative way of writing games, but it's unproven (to me) and highly experimental at this point.

PyGame →

Mainly sprite based examples, there are some 3D games, but most seem to involve calling opengl commands wrapped in python, rather than a nice high level entity model (I might be wrong these days tho).

Panda3d →

Free software python based game engine for windows & linux from Carnegie Mellon. Looks really good.

some more notes can be found in the game engines node

Blitz Basic → who make Blitz3D →

Lots of games written in this, very popular in the indy game dev scene.

Virtools →

Does online, visually programmed, and looks very suitable - just also kinda expensive.

Unity →

  • game_prototyping.txt
  • Last modified: 2008-07-18 08:30
  • by