• Join Us |
  • |
  • Sign in with:

Learn to Design Games Really Fast

by Chris Remo, Oct 27, 2005 9:00am PDT
Related Topics – Valve

You may have seen my post two days ago about Carnegie Mellon graduate students who each design, prototype, and build a game over the course of a week, ending up with about 50 games over the course of a semester. Well, if doing that sort of thing appeals to you (maybe not once a week for an entire semester, though) then you can read the guide that four of those guys conveniently published on Gamasutra yesterday. It gives a good look at the thought processes behind designing games that are focused around a single strong gameplay idea, as opposed to games which attempt to simulate reality by including a broad range of detailed situations and features.

You only need a few days. It seems like a natural and comforting thing to say, "Hey we made a great game in one week. Therefore, if we spend TWO weeks, it will be TWICE as good!" Of course this isn't the case. We found that generally any gameplay idea can be prototyped effectively in less than one week. Extra slop time tends to yield diminishing returns. Some prototypes, for instance, took just a single evening to throw together, while others got an extra week or two of love. Surprisingly, we found that there was no correlation between time spent in development and how successful the game ultimately turned out.
Insights like that may not be quite applicable to large-budget game development, but in terms of mod authorship they do dovetail nicely with comments made by Valve's Erik Johnson, who suggests that the most successful mods are the ones with a single strong idea, and are released as soon as possible in order to guage reactions and make improvements. He also suggests that the lack of such an approach is hurting the current-day mod scene. Anyway, the Gamasutra article is a good read. It's written in a casual and disarming style, so despite its instructional content it flows easily. There are plenty of examples and screenshots of games such as Matt's Potato, Tower of Goo, and Super Tummy Bubble. It also explains why Alien Hominid, Super Puzzle Fighter II Turbo, and Super Mario Bros. can be described as "juicy."




Comments

10 Threads | 29 Comments
  • The best games I've worked on are games where the core idea (and the fun) take root quickly and are nurtured over the course of the project. The worst games I've worked on have vague (but often cool) ideas that just seem to languish, waiting for an undisclosed amount of content before they become "fun."

    The most disappointing games are the ones where your prototype is great fun, but it's both challenging and rough around the edges. So when you try to make it more accessable and less buggy, you kill the best part of the game. (For example, a racing game where you can skillfullly exploit bugs in the physics to go faster, but since they are obviously bugs, the publisher demands that you squash them.)