Ed Cogburn
Posts: 1979
Joined: 7/24/2000 From: Greeneville, Tennessee - GO VOLS! Status: offline
|
quote:
Originally posted by Paul Vebber:
Breakthrough AI will take awhile, our initial crop of games will have decent AI, hopefully better than decent!
....
AI is one of those things that can rapidly reach dimishing returns on investment. Some of teh most popular games have had rudamentary AI (Sp a prime example) but was so much fun folks mostly accepted that.
....
Then back it up on our end with commitment to the evolution of the product lines.
I do like to hear about a commitment to the evolution of your product line, as long as that commitment is deep enough to matter. Every company tries to profit off of its best games by coming out with "BestGame-II", or "Advanced BestGame", but do they ever come back to a game and bugfix it more than 2 years after its release? A basic update to make it playable on newer equipment for example (not improvements, just fix-it-enough-so-it-works-on-new-gizmo)? No, ... they'll support it for about 1.5 years, and after that it becomes abandonware. There are a few exceptions, like Gary Grigsby coming back to PAC to make an unofficial update. Of course, it was unofficial because SSI had lost interest in the game, so that partially makes my point too.
This is the main reason I like Open-Source so much, although it has not made a significant impact on the computer gaming industry yet. With open-source you are basically guarantteed not to end up with abandonware, if the game has a decent popularity, there will always be someone working on it or at least maintaining it. The FreeCiv project has a lot of the Big Mo (momentum) behind it. It may turn out to be the first open-source project that proves itself to be viable in the gaming community.
As for AIs, don't kid yourself. I and I'm sure *many* others recognized the stupidity of the AI in SP, but with enough cheating the AI could at least *appear* reasonable, and thats all we needed since the game was so fun we played it anyway, stupid AI or not.
I dream of the day when a game company has the guts to produce a popular game where it opens up the AI to the user. By this I mean designing the game so the AI can be, in effect, a plug-in module. Modules using some kind of embedded scripting language like Python or equivalent, with the source script for the default AI provided, allowing the users (the fraction of the user base that are also programmers, professional or otherwise) the chance to experiment and improve it over time. If the game is popular enough, this would work, I'm sure of it, but I also figure I'll have to wait until Hell freezes over first though ....
_____________________________
|