Just a quick question:
Is there any plan for updating the AI? I just played two games to about 50 turns, and it was painfully apparent that the AI pretty much has no clue how to design ships, build them, or deploy them. I was able to crush every AI player in about as much time as it took to send my fleets to them.
An easy question to ask, and hard to solve, but is there any plan to work the issue?
AI Plans
Moderator: Oberlus
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: AI Plans
It is known that the AI is no challenge.
We had some contributors who were submitting AI patches, but IIRC they haven't been active on the forums recently.
We had some contributors who were submitting AI patches, but IIRC they haven't been active on the forums recently.
-
- Space Krill
- Posts: 2
- Joined: Fri Sep 07, 2012 3:12 am
Re: AI Plans
If I wasn't going to school right now and raising kids while working my day job, I'd offer to get some ideas and/or code in.
But alas, I am... I'll be done in 2 years
Thoughts run along the lines of operations research "optimization" of value:
1. Identify AI opportunities
a. colonization (planet value, proximity to empire, proximity to enemies, defensibility, base of forward operations)
b. defense (which owned systems are vulnerable, which are not, how to defend them, when they're "safe")
c. offense (which opponent controlled targets are valuable, which are not; how to take them, when it's achievable; destruction of military vs. conquering of fleet)
2. Identify opponent priorities
all of above, run for opponent
3. Identify strategy
a. balance promotion of own 3 goals vs. denial of opponent 3 goals (favor self-promotion over competitor denial)
4. Identify AI response
a. technology (which are valuable for achieving above, how well supported)
b. industry (which builds support objectives above, how well supported, how to tactically adjust to technological opportunities)
c. movement (how to allocate existing resources to balance above, knowing what builds are in the queue)
d. all planned over coverage of how well goals are covered over 1, 2, 3, 4-turn horizon
Just a brain dump.... More to come in 2 years
But alas, I am... I'll be done in 2 years
Thoughts run along the lines of operations research "optimization" of value:
1. Identify AI opportunities
a. colonization (planet value, proximity to empire, proximity to enemies, defensibility, base of forward operations)
b. defense (which owned systems are vulnerable, which are not, how to defend them, when they're "safe")
c. offense (which opponent controlled targets are valuable, which are not; how to take them, when it's achievable; destruction of military vs. conquering of fleet)
2. Identify opponent priorities
all of above, run for opponent
3. Identify strategy
a. balance promotion of own 3 goals vs. denial of opponent 3 goals (favor self-promotion over competitor denial)
4. Identify AI response
a. technology (which are valuable for achieving above, how well supported)
b. industry (which builds support objectives above, how well supported, how to tactically adjust to technological opportunities)
c. movement (how to allocate existing resources to balance above, knowing what builds are in the queue)
d. all planned over coverage of how well goals are covered over 1, 2, 3, 4-turn horizon
Just a brain dump.... More to come in 2 years
Re: AI Plans
Hi, I've been working on AI upgrades and am pretty pleased with them. I'll upload a zip file & make a separate post for it.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0
- Geoff the Medio
- Programming, Design, Admin
- Posts: 13587
- Joined: Wed Oct 08, 2003 1:33 am
- Location: Munich
Re: AI Plans
Be sure to note in your post that you release the changes under the GPL 2.0 or later.Dilvish wrote:Hi, I've been working on AI upgrades and am pretty pleased with them. I'll upload a zip file & make a separate post for it.