Tuesday, April 23, 2013

GDC2013 - Prototyping: A Developer's Best Friend

Simon Darveau of Spearhead Games talked about the value of prototyping.
  • The most important lesson is to "follow the game's will."
  • Simon worked on Petz Sports and the Assassin's Creed series. Petz Sports was designed entirely on paper first. 90% of the stuff did not end up working in the digital version.
  • He believes that a revolution is coming. There are four steps in creating a game... 1) identify the high level goal, 2) build core mechanics and experience, 3) defining the most amazing content, and 4) producing the game. Simon is best at the third step.
  • 12 advices for prototyping
    1. A game has its own will. Follow the fun and it's not always where you thought it would end up.
    2. Think inside the box. Start with familiar mechanics and add a twist. Don't rely on mechanics that haven't been proven yet.
    3. Escape from prisons of paper. Don't use a game design document because it's the wrong medium to express interaction. It's not only boring to write, but it's too abstract. Destroy any paper process and get the team talking and working together by day one. It promotes ownership and faster development.
    4. Prototype in the same game engine you are using. Don't play with another beast and realize that it cannot be ported back into your engine. Also, don't make several different parts outside of each other and then stick them back together because there are lots of integration issues.
    5. Iteration time is of the essence. Don't make one map of your three-week schedule; make 3 maps per day. it will force you to simplify what you are trying to do.
    6. Try everything, generate chaos.
    7. Merge game and level design. If you're a game designer, you should not push any feature without proving it out yourself first.
    8. You don't know what your searching for until you found it. Prototyping is not a validation tool, it's an experimentation tool.
    9. Refine your diamonds. After you find the mechanic, iterate and polish it so you can communicate your ideas to others.
    10. Make your prototype open source. Playtest it a lot with everyone and get as much feedback as possible.
    11. Be humble and confident. Don't be afraid to fail and be ready to fail all the time.
    12. You need ninjas, not warriors. Warriors see their games as products and will just do what they're told. Ninjas are the true developers, not afraid to ask questions and explore.
  • There's no official prototype validation process. Just play internally with the team and the gems will rise to the top.
  • Prototyping is not shooting in a random direction. You must know where you're going.
  • Don't use paper documents. It frees you up to discuss changes as a team.
  • Convince management that prototyping is worth it. The best way to do this is just make the prototype and show them the value of it.

No comments: