A.I.; Does it cheat?

Jrat · January 22, 2002, 23:29

I have Apolyton & Goodmod installed on my system... I have played my first game till the year 1750 A.D.. However in the year 1100 A.D., I noted that I had only enough technolegy to build Knights & Summeri while my enemy was building Infantry & Infantry men... This did not make sense to me, considering in real life when these units became available... Did I make a blunder in scientific research, or is this a bug in the game? And, I have noticed that sometimes, my science research time drastically increases for no obvious reason... ie. from 30 turns to 270... Is there a way to improve research other than the percentage of gold dedicated to research option in the Empire manager? Thanks...

and if it matters, I played my game on the "impossible" option...

Martin Gühmann · January 23, 2002, 03:25

You should max out the Food/Production/Gold sliders. That is what the AI now in GoodMod does and I also do, in my last GoodMod game I was able to catch up the AI finally. But I have also the problem that I am not far enough to build the Gaia Controller between 2200 and 2300. By the way I set the ending date in GoodMod back to 2300.

BTW in higher difficuilties levels you have less turns so your tech develope will be finished later.

And I think the AI is hurt by the pollution happiness effect to much therefore a game without pollution would allow the AI to max out the slider the entire game.

-Martin

Jrat · January 25, 2002, 03:50

Thanks... I'm playing a new game, and though I'm still the world's weakest nation, I am certainly doing better this time than the last... Practice makes Perfect, as the saying goes!

... After all, this is only my second game!

Pedrunn · January 26, 2002, 05:25

Quote:

Originally posted by Martin Gühmann
And I think the AI is hurt by the pollution happiness effect to much therefore a game without pollution would allow the AI to max out the slider the entire game.

I have notest that and i am looking forward to decrease the effect of pollution on happiness. I have look all the way through the const.txt and citysize.txt several times but i just could find nothing to make this change. Do you have any idea how can i manage that?

Martin Gühmann · January 26, 2002, 09:05

Pedrunn the Apolyton Modification Pages will you lead to the govern.txt. Here is the improtant stuff from these pages:

PollutionCoef - Government modifier to Pollution rate. Pollution is multiplied by this number.

PollutionUnhappyCoef - Unhappiness amount caused by Pollution. Pollution is multiplied by this number to get Unhappiness amount.

Most governments have a PollutionUnhappyCoef of 0.02. That means a city with 600 pollution has a happiness reduce of 12 and a city with 400 pollution of 8. I usually compensate the effect with entertainers. The AI can't use entertainers in this way, that's the reason why I saw in MedPack2 a civ that destroyed itself, because of this unhappiness effect. In my last GoodMod game the same happend with a very productive city. Just a revolutiion, because of high pollution unhappiness and there were no way to compensate it with sliders by maxing them out concerning happiness. But if you reduce the happiness effect you should increase the effect on the invironment. In Apolyton Pack you can have cities with a very high pollution, but the effect on the city invironment is minor. You can adjust this in the const.txt. That shouldn't be a problem for the AI now if you combine it with the latest terraforming fix.

-Martin

Pedrunn · January 28, 2002, 06:54

I had found this line yesterday and i already managed the chance. I hadnt see your post before because for some reason the whole apolyton site could not be accessed (page not found) but thank you for you help i already decrease the pollution effect on happiness.

Martin Gühmann · February 17, 2002, 14:15

Pedrunn what are your values that you now use?

-Martin

kaan · February 20, 2002, 07:23

please do not forget what really stops the AI from cleaning up pollution, the bad values in strategies.txt .

in STRATEGY_DEFAULT i found these values defined:

// rounds to wait before fixing roads
TimeToFixRoads 10

// rounds to wait before fixing polluted tiles
TimeToFixPollution 20

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 2000

this means that the AI spots a polluted tile in its area and then wait for 20 tuns (after waiting for 10 turns the AI might consider building a new road around the polluted spot if it had one), then it looks at its PW and if it has plenty it is fixed.
In other words: this will never happen!
now the problem is not that the default personality never cleans up pollution, but that the values for the rest of the personalities in this file are the same (this is especially absurd for the ecotopian).
i would suggest these settings:

STRATEGY_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 10

// rounds to wait before fixing polluted tiles
TimeToFixPollution 20

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 2000

STRATEGY_SCIENTIST_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 5

// rounds to wait before fixing polluted tiles
TimeToFixPollution 5

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 500

STRATEGY_MILITARIST_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 3

// rounds to wait before fixing polluted tiles
TimeToFixPollution 20

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 2000

STRATEGY_ECONOMIC_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 5

// rounds to wait before fixing polluted tiles
TimeToFixPollution 7

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 700

STRATEGY_ECOTOPIAN_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 10

// rounds to wait before fixing polluted tiles
TimeToFixPollution 1

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 100

STRATEGY_DIPLOMATIC_DEFAULT

// rounds to wait before fixing roads
TimeToFixRoads 10

// rounds to wait before fixing polluted tiles
TimeToFixPollution 5

// stored PW must exceed cost to fix + threshold
FixPollutionThreshold 500

These values should have a lot more playtesting that it has but i can tell you that it works quite well for getting the pollution removed and this in itself reduces pollution and thereby it should increase happiness. another observation i made reguarding the effect of these changes is that it seems to make the AI work better througout the industrial age when it comes to economics and happiness control.

i hope this is usefull for you guys

klaus

Martin Gühmann · February 20, 2002, 14:53

Dead tiles are not the problem, actual in Apolyton Pack they are very rare. The real problem for an AI is the pollution of 600 per city that cuases a lot of happiness reduction. As the AI can only deal with it by using the gloabal sliders instaed of local entertainers. That is a major problem for the AI if the it has to sarcifice its empire efficency for the sake of happiness. So lowering the ubhappiness effect of pollution will reduce this problem.

By the I decided just to half all the pollution unhappy coefficients for all gouvernments.

-Martin

kaan · February 20, 2002, 16:05

then i misunderstood the problem (a rare example

)
well the mayors should use entertainers to make people happy no matter what the cause of unhappines is.
how do i spot the problem in the game? (mods dificulty etc.) i want to find the reason

klaus