Suggestion: More active industrial AI

Have a feature request for SRGW? Post here.

Moderators: Balthagor, Moderators

Post Reply
amynase
Lt. Colonel
Posts: 212
Joined: May 02 2014
Human: Yes

Suggestion: More active industrial AI

Post by amynase »

With a 1914 start, and maybe even the implementation of a unit limit as suggested in another thread *whishful thinking* , the average game of SR:GW could take way longe than just a couple of years.
Right now in SR:U, the one field in which the AI can truly not compete with a player at all is the economy. Even if you play the game for 10+ years, most AI countries will barely build anything besides maybe a few land unit production sites or Agriculture.
I think it would make the game a lot more intresting if the player had some true economic competition from the AI, meaning an AI that actively constructs new production sites for any goods that it either has too little production itself or can sell with profit on the world market. For this, it should of course consider if enough of the required goods are availible for the construction.

It would really add an intresting new way for the player to compete with the AI besides the pure military way of competition the AI offers right now.

Thank you for considering this :D
nick-bang
Colonel
Posts: 348
Joined: Sep 07 2010
Human: Yes
Location: A dark and ominous room - only illuminated by the eerie light of a computerscreen

Re: Suggestion: More active industrial AI

Post by nick-bang »

Very hard to model.

What should the computer try to build:

That which it can make money on, areas where its lacking ressources?
User avatar
Zuikaku
General
Posts: 2394
Joined: Feb 10 2012
Human: Yes

Re: Suggestion: More active industrial AI

Post by Zuikaku »

nick-bang wrote:Very hard to model.

What should the computer try to build:

That which it can make money on, areas where its lacking ressources?
1. increase production of resources it lacks if available.
2. increase production of resources demanded on world market ("good oportunity" resources)
3. increase production of lacking goods
4. increase production of goods in high demand of world market
5. build only if money, resources and manpower available.
Please teach AI everything!
nick-bang
Colonel
Posts: 348
Joined: Sep 07 2010
Human: Yes
Location: A dark and ominous room - only illuminated by the eerie light of a computerscreen

Re: Suggestion: More active industrial AI

Post by nick-bang »

OK, but how should it prioritize?

Also at some point you have to make up your mind if you believe in a Keynesian model. Or make a science fiction game by building it onb a communist idea.

Or - hwich is probably the only one you can actually model realistically - a Liberal model.
User avatar
Zuikaku
General
Posts: 2394
Joined: Feb 10 2012
Human: Yes

Re: Suggestion: More active industrial AI

Post by Zuikaku »

nick-bang wrote:OK, but how should it prioritize?

Also at some point you have to make up your mind if you believe in a Keynesian model. Or make a science fiction game by building it onb a communist idea.

Or - hwich is probably the only one you can actually model realistically - a Liberal model.
Secure electricity => secure resources => secure goods => loop to start ?
Please teach AI everything!
nick-bang
Colonel
Posts: 348
Joined: Sep 07 2010
Human: Yes
Location: A dark and ominous room - only illuminated by the eerie light of a computerscreen

Re: Suggestion: More active industrial AI

Post by nick-bang »

Zuikaku wrote:
nick-bang wrote:OK, but how should it prioritize?

Also at some point you have to make up your mind if you believe in a Keynesian model. Or make a science fiction game by building it onb a communist idea.

Or - hwich is probably the only one you can actually model realistically - a Liberal model.
Secure electricity => secure resources => secure goods => loop to start ?
I could buy that but we would need it to adjust to supply and demand. It makes little real sense to have the AI sell goods below production value.
way2co0l
Brigadier Gen.
Posts: 687
Joined: Nov 29 2010
Human: Yes

Re: Suggestion: More active industrial AI

Post by way2co0l »

nick-bang wrote:
Zuikaku wrote:
nick-bang wrote:OK, but how should it prioritize?

Also at some point you have to make up your mind if you believe in a Keynesian model. Or make a science fiction game by building it onb a communist idea.

Or - hwich is probably the only one you can actually model realistically - a Liberal model.
Secure electricity => secure resources => secure goods => loop to start ?
I could buy that but we would need it to adjust to supply and demand. It makes little real sense to have the AI sell goods below production value.
Agreed wholeheartedly. I've long been a proponent of the AI restricting its production when its production costs exceeds market value. That should apply here as well. If the cost to produce agriculture exceeds what it costs to get it on the world market, and they are able to purchase sufficient quantities on the world market, then they shouldn't make an effort to produce more on their own. There are arguments to be made for self sufficiency, but for simplicity sake, it's better for the AI to buy it on the market when the option to do so cheaper is available.

However, if production costs are cheaper than world market prices, they do not currently produce enough to fulfill demand, or there is a shortage on the world market and there are sufficient financial resources available, then begin construction. Construction location choices should be weighted to prioritize hexes which have no other resources possible, as well as highest local supply. Any hex where the total number of resources for all combined resource options is lower than 6 is a location that should be prioritized based on local supply. Any hex where resource amounts exceed 6 should only be chosen when all other options for that resource have already been utilized. When the AI is considering construction in a hex with low supply, they should first identify a nearby hex (within 4 hexes distance) with the fewest resource count to construct supply depots to improve supply.

I can expand on all of this to clarify my vision more if desired, but I think it's probably pretty clear how I'd desire it to work in this regard.
SGTscuba
General
Posts: 2544
Joined: Dec 08 2007
Location: Tipton, UK

Re: Suggestion: More active industrial AI

Post by SGTscuba »

I think this would be useful especially with regards to the shortages of oil, mil goods and rubber that seem to always occur. Normally I end up bleeding the Ai dry in big wars as they cannot buy or produce enough mil goods and so they halt unit production and still run out of supplies.
My SR:U Model Project, get the latest and post suggestions here:

http://www.bgforums.com/forums/viewtopi ... 79&t=28040
nick-bang
Colonel
Posts: 348
Joined: Sep 07 2010
Human: Yes
Location: A dark and ominous room - only illuminated by the eerie light of a computerscreen

Re: Suggestion: More active industrial AI

Post by nick-bang »

So with a delay - it seems that we have reached a consensus ?
Fistalis
General
Posts: 3315
Joined: Jun 23 2009
Human: Yes
Location: x:355 y:216
Contact:

Re: Suggestion: More active industrial AI

Post by Fistalis »

I'm a big proponent of having the AI randomly have a variation in levels of competence.. More stuff in the AI stances etc to determine what they would focus on. Those who are focused on economy or internal issues etc SHOULD be better at building etc than they are now though.. I'll say that.

Edit: Let me expand. While i'm all for improving the overall AI in this aspect I strongly believe the specific focus and level of proficiency should be closely tied with with the leader of a country. Whether that's determined by a random table or tied to specific leaders based on their historical records i'd leave open to debate. I would love to see a tangible difference when there is a leader change of a country.. and this is one of many things that would help achieve that.
Si vis pacem, para bellum
my Supreme Ruler mods Site - May it rest in peace
Post Reply

Return to “Suggestions - SRGW”