AU: AI creation contest

[sabrewolf]

(bump)

seems to be interesting again...

http://apolyton.net/forums/showthread.php?s=&postid=4079481#post4079481

ps. about time for the AU forums *hint* *hint*

[LDiCesare]

Might be interesting, but time-consuming. I wonder how much of the ai can be modded in Python, too. After all, CtP2's ai could be heavily modded in SLIC. I may do some ai tweaking in Python because I want to learn the language and I need something to do with it.

[Tassadar500]

That got me thinking. Perhaps we could let people have partners to do it with if they so wished?

[sabrewolf]

lord nuclear, in my first post:

[...] everybody/every team creates their own AI, either by modding the existing AI or taking ideas from others [...]

it is clear that some AU guys will join forces while other people might go on their own (eg. thinking of people like TETurkhan or the DyP mod creator, etc)
the C4-AC guys will surely also have some input.

but i don't think this will be priority. although dominae sais we should discuss from begin on, most people's modding will be more experimental (eg. make ultra-agressive/ultra-builder AI, etc)

[Fried-Psitalon]

If it's open source, then people can build on each other's ideas and continue to improve the AI. Good stuff.

[nbarclay]

An AI that does well against other AIs may not do as well against humans (for example, because it is too predictable), or may not be fun to play against (for example, if it is so aggressive that it's no fun for players who prefer relatively peaceful games). A contest sounds like a good way to find candidates for AIs or AI features to incorporate into the AU Mod, but we'll also need to see how effective and fun AIs are as opponents for human players before we make any final decisions. Eventually, we may even want to offer versions of the AU Mod with different AI personalities that fit better for players who prefer different types of games.

[Fried-Psitalon]

Erg. Really hope there's not an AU mod for CIV. What's the point of teaching people to get better at the game if you're going to change how it functions? At least give the game a chance for six months or so to learn the deeper details first...

[Tassadar500]

Originally posted by Fried-Psitalon
Erg. Really hope there's not an AU mod for CIV. What's the point of teaching people to get better at the game if you're going to change how it functions? At least give the game a chance for six months or so to learn the deeper details first...

Wow, you've totally missed the ball on this one. Not only have we come to the consensus that we will wait a bit before we will start making changes, the AU Mod isn't about making easier for the player. It's about improving the AI, balancing the game, and in the end, making it tougher for the player.

[nbarclay]

Originally posted by Fried-Psitalon
Erg. Really hope there's not an AU mod for CIV. What's the point of teaching people to get better at the game if you're going to change how it functions? At least give the game a chance for six months or so to learn the deeper details first...

As one of the panel members for the Civ 3 version of the Mod, my expectation is that development of the AU Mod for CIV is going to be a gradual process. There may be a few things we change quickly if we discover serious balance problems that human players can exploit. But for the most part, it's going to take time for CIV strategy to evolve to a point where we have a clear picture of what kinds of changes are valuable enough to be worth deviating from the stock rules.

[Theseus]

Don't worry FP, it'll be quite some while before an AU Mod actually gets done and is available.

The discussion in the interim, however, should also be helpful in terms of patches and whatnot.