AI Trade Calculations and game performance

Big Bopper

Prince
Joined
Oct 21, 2011
Messages
307
Location
New England
I have a total of 25 tradeable resources, 17 strategic and 8 luxuries. There are 31 civs, but 7 civs will only ever have the prerequisite of 2 strategic and 3 luxury resources and the other 24 will never have the prerequisites to two strategic resources. I am wondering how much this will affect game performance.

My Questions are...
1. Does the AI calculate resources that it has not discovered the prerequisite tech for?
2. Does the AI calculate resources it already has access to?
 
As in the end this is a question of the Civ 3 code, I think the only person who has the answer is Flintlock.

My guess is, that the AI only counts resources that it already has access to by having them in the tradenet and having the proper techs for it.

The reasons for that guess: All resources in a Civ 3 game are placed on the map since turn 1. This can easily be recognized by having a look at the map of a game in DEBUG mode. Therefore, if the AI would also include resources it not yet has discovered them in the calulation and not has the techs for those resources, the "basic" calculations here should be the same with the small exception of the disappearing probability of some resources.

But from my feeling those times for the "basic" calculation during a game are not constant, but rising.

So it seems, that the "basic" calculation for the disappearance of resources that are placed on the starting map of a game is for all resources on the map, even if not a part of the trading net and not having the proper techs for those resources, there must be a minimal additional calculation time for the new location where that resource is reappearing again on the map.

Therefore in my eyes the calculation time for resources should be rising by the tradenet of the AI and the technological knowledge of resources inside that tradenet and an additional minimal calculation time for the new location of reappearing resources - but this is only a guess.
 
Thanks for taking the time to respond.

I hadn't even thought about the disappearing probability being a factor. Even if its guessing, your guess will be closer to the fact than mine. Have you had many problems with game performance using this many tradeable resources in your mods/scenarios? I've always tried to limit mine to about 16.
 
Have you had many problems with game performance using this many tradeable resources in your mods/scenarios? I've always tried to limit mine to about 16.
To have "problems" with the game performance is a subjective feeling. The high number of tradeable resources exists since more than two decades in the mods DYP2 and RAR.

The C3X mod by Flintlock speeds up the tradenet calculations massively and it also contains a tool to measure the interturn times. The results are astonishing. In my tests with that tool, the most massive rises in the calculation time were not by normal trade deals, but by diplomatic actions (mutual protection pacts):

https://forums.civfanatics.com/thre...rd-and-much-more.666881/page-87#post-16544741

https://forums.civfanatics.com/thre...rd-and-much-more.666881/page-87#post-16544800

https://forums.civfanatics.com/thre...rd-and-much-more.666881/page-87#post-16544800

To have a modern strong pc is also very helpful to speed up the performance of the game. In the last CCM3 succession game (now finished -congrats to the team) on a big archipelago map, the turns became very long in the later stages of the game. I played some of the turns on my nine year old pc (which was very strong when I bought it) and the same turn on my new strong pc, bought in the last black friday sale 2024, and the difference in speeding up the calculations was enormous.
 
Last edited:
I have to finally give Flintlocks mod a try, once I get this scenario finished.
I think it could be better to give C3X a try first, as there is a big chance, that you will rework your scenario after you have tried Flintlock´s mod. :)
 
Last edited:
That's what I'm worried about. :lol:

So close to finishing it. Once it is in a working state, then I'll see what can be changed for the better.
 
Last edited:
Back
Top Bottom