C-evo AI Forum

Posted 7-Mar-07 15:14 by Retudin

Capital AI

No pollution cleanup:

If the settler can reach the polluted tile, and is not busy doing something else, this is a bug; please send the game to r.janssen@paragon.nl. Note that the one settler may not be able to keep up with pollution. Draining swamps indeed needs implementation.

Riflemen 0/78/150 to weak:

The defense strength could be upgraded to 147 by developing composites mobile warfare and a new defensive unit; this would however cost valuable time, and make all 25 computing techs more expensive.
Note that Capital AI is build on the assumption that 1 city is enough to keep ahead with technology. The central assumption is that research tech is good enough to develop an advance each turn in the end game. In version 0 of civ-evo, the science multiplier good get to 400% (library + university + research lab) + 50*90% computing tech bonus ) * 2 (observatory) i.e. 9800% this was sufficient for CapitalAI to work.
In version 1, the bonus is (300+25*15)*2 or 1350, i.e. research is more than 7 times as slow after developing the future research techs. In addition trading advances has become much more important, since not only the traded advance becomes cheaper, but also all advances developed later. This makes scouting much more important, and having the technological lead much less.
In addition to all these rules changes, the other AI (at least StandardAI and Seti) have become a lot better.
Conclusion: If defense 78 isn’t enough, capital AI will not survive until the end anyhow, set it on a lower difficulty, or play civ-evo 0.14.
Note: just before evo1, I started writing global conquest by air assault, in evo0, it might have worked, but in evo1, it won’t (using only one city) , and Seti does it better. Since I cannot improve Capital AI significantly, I decided to start a new AI (don’t expect it to be ready soon) that does scout found cities etc.

Coastal fortress/ SAM:
When I might make a new version I may implement that (but see remarks above)

Diplomacy:
Refuse contact once, and you will not be contacted anymore.
Trading spaceship parts seems useless, unless you are producing such yourself. (Even paying 1 gold for parts will only help an advanced opponent)
I do intend to make a more advanced diplomacy handler though, and probably will put it into CapitalAI

Answer This Posting

User Name

New user

Password

Change password

Subject
Capital AI

Text
: No pollution cleanup: : : If the settler can reach the polluted tile, and is not busy doing something else, this is a bug; please send the game to r.janssen@paragon.nl. Note that the one settler may not be able to keep up with pollution. Draining swamps indeed needs implementation. : : Riflemen 0/78/150 to weak: : : The defense strength could be upgraded to 147 by developing composites mobile warfare and a new defensive unit; this would however cost valuable time, and make all 25 computing techs more expensive. : Note that Capital AI is build on the assumption that 1 city is enough to keep ahead with technology. The central assumption is that research tech is good enough to develop an advance each turn in the end game. In version 0 of civ-evo, the science multiplier good get to 400% (library + university + research lab) + 50*90% computing tech bonus ) * 2 (observatory) i.e. 9800% this was sufficient for CapitalAI to work. : In version 1, the bonus is (300+25*15)*2 or 1350, i.e. research is more than 7 times as slow after developing the future research techs. In addition trading advances has become much more important, since not only the traded advance becomes cheaper, but also all advances developed later. This makes scouting much more important, and having the technological lead much less. : In addition to all these rules changes, the other AI (at least StandardAI and Seti) have become a lot better. : Conclusion: If defense 78 isn’t enough, capital AI will not survive until the end anyhow, set it on a lower difficulty, or play civ-evo 0.14. : Note: just before evo1, I started writing global conquest by air assault, in evo0, it might have worked, but in evo1, it won’t (using only one city) , and Seti does it better. Since I cannot improve Capital AI significantly, I decided to start a new AI (don’t expect it to be ready soon) that does scout found cities etc. : : Coastal fortress/ SAM: : When I might make a new version I may implement that (but see remarks above) : : Diplomacy: : Refuse contact once, and you will not be contacted anymore. : Trading spaceship parts seems useless, unless you are producing such yourself. (Even paying 1 gold for parts will only help an advanced opponent) : I do intend to make a more advanced diplomacy handler though, and probably will put it into CapitalAI : : :

Capital AI

Answer This Posting

Administration