The Immortal Challenge 1: Apocalypto

By the way, DaviddesJ upgraded my account to Pro for me. Let's give him a big clap :clap: Thanks for your solid support for the thread! Really appreciate it :)
 
The ultimate goal should be to have an AI that is equal to the best human player without ANY bonuses and call it Deity. Lower levels would still be without bonuses, but the engine would be "custom-lobotomized" to make stupider decisions.

Examples of this can be seen in chess programs. Crafty is a fairly strong engine, about 2660 strength (the top human player is 2800). On the Internet Chess Club, there are full strength Craftys, but there are also accounts WimpA, WimpB, WimpC and WimpD which, using the basic Crafty engine and obviously playing by the same rules, have been intentionally reduced to strengths of about 1900, 1700, 1500 and 1300 (the equivalent of a Monarch, Prince, Noble and Warlord, say) and which are tremendously popular for human players to practice against.
 
The ultimate goal should be to have an AI that is equal to the best human player without ANY bonuses and call it Deity.
I'm confident this is an impossiblilty for a game like CIV. Just think about the number of decisions you make over the course of a full game of CIV versus the number made in a game of chess. Not to take anything away from chess, but the level of complexity is orders of magnitude different. One of the biggest issues facing a CIV AI is simply defining what success looks like. Training an adaptive algorithm to succeed, when "success" is such a nebulous concept, is a monumental undertaking.

EDIT: I was just thinking.. a cool new wonder for CIV, in the same vein as the cure for cancer, would be "Competent Civilization AI".
 
Obviously brute force methods wouldn't work, but heuristic methods might not be inconceivable. Civ is more complex in some ways, but simpler in others, and in broad terms, I don't think success is that hard to define: bigger is better, more tech is better.

As I said, it would be an ultimate goal - we might not be able to actually achieve it with current technology and the resources devoted to it, but in striving towards it, we would reach a desirable approximation. The fewer bonuses we have to give the computer, the better.
 
Yey! Way to go, DaviddesJ! aelf: so, where's the update? ;)

Coming. I've finished playing the round. It was tough one in terms of the amount of micromanagement I had to do. Lower happiness caps and Sacrificial Altars make for a different pattern of gameplay. I will post the update later.
 
Round 4: 200BC - 780AD

Quite a long round this time, since the focus was on building up the economy and not much happened. The Sacrificial Altar really came to the fore as we whipped buildings and units up all over our modest empire. And the sacrifice must go on. We are the mighty people of the banner of the sun, and all that jazz ;)

First, we made peace with Louis:



I tried asking for Mathematics, but he wouldn't give it to us, so I took as much as possible from him.

Then I went to Augustus for a trade:



With Mathematics in the bag, we could go for Currency next, which would prove to be the right decision.

Then we switched religion for a while:



This was to make a tech trade with Louis. Might as well get as much out of Alphabet as possible.

And since we're going to be peaceful people for a while, we planted our GG as a GMI in the capital:



Why there? Is it going to be our military city? Yes. Why? Because it could be a lot sooner than anywhere else. The other city with potential is Orleans where there is plenty of food for whipping and hammers for hard production. However, it would be some time before that city could develop to play that role as effectively, so I decided that it would be our future GP farm instead.

I also followed lilnev's recommendation of not permanently running a priest in the capital for now, focusing instead on working cottage tiles to develop them and using excess food for whipping.

On the next turn, we could immediately trade with a cautious Louis:



Not a bad deal. It gave us Organized Religion, to which we switched:



It would help us as we focus on peaceful builds.

I tried to see if any of our two neighbours would trade Calendar for CoL, but none would. Ah, well. We needed monuments for a little while more to expand the borders of the newly conquered cities anyway.

Augustus Caesar soon switched his state religion to Christianity, as we feared. But, thankfully, it didn't stop him from trading with us:



Although I hate to be helping Augustus (who was beginning to run away technologically), we needed Hereditary Rule quickly to increase our low happiness caps.

After Currency was discovered, we went for Construction so that we could start building catapults that we would use for a second invasion of France.

Fast forward a little, I switced our state religion to Confucianism again to try and get Calendar from Louis. He wouldn't give it to us for anything, though, so I settled for the next best trade:



It's nothing great, but at least the gold would help us research Calendar ourselves next.

[to be continued in the next post...]
 
[continued from previous post]

Since Orleans was unable to grow quickly due to unhealthiness until we chop all those jungle, I decided to run scientists there for the moment to put in those GS points for the future:



Our capital meanwhile, gradually put hammers into settlers whenever it waited for whip weariness to go away. With this method, we were able to build a settler to build the lakeside/seaside city 1N of uberfish's spot (to keep the grassland hill so that it could be mined for hammers). And we were building a second one, probably for futurehermit's lakeside copper city. I also ran a priest occassionally to time the growth of the capital accordingly. Because of this, we would probably get our second prophet soon enough:



And, yes, we researched CS after Calendar. I think getting the Bureaucracy bonus online asap is important. We need the hammers and commerce for our war effort and to help catch up (or at least not fall behind too much).

As soon as CS was discovered (it didn't take the full 18 turns, since our economy improved and we were able to raise our research slider for a while), we made the civics switch:



And our army, which we have steadily built up, is poised to strike:



This is where the round ends. A few decisions need to be made next. First, where and how should our army strike? I'm thinking of heading straight at Paris and then moving up to the French city on the iron (should we keep it or raze it?). I would stop there, leaving Louis with Lyons as a buffer between us and the powerful Augustus. What do you think?

We also have to decide on what to research next. We are substantially behind now and there's little to no hope of tech trading in the near future. Here's the sitch:



We have nothing on Augustus. I'm thinking of going Metal Casting -> Machinery so that we can build crossbows and our own macemen to guard against and possibly attack Augustus.

Lastly, we have a settler waiting in the capital. Should we use him to build the lakeside copper city? Or should we go for another spot first?

The power graph:



Yes, the Romans are soaring up there. They have macemen and longbows. Louis has longbows too, which explains the slight spike in his power, but I don't think he will be too hard to deal with.

The GNP graph:



What can I say? It's pathetic. Our economy has been strengthened, though. We're no longer relying on the gold mines as our sole source of income. And as I mentioned, we whipped throughout the round under auspices of the Sacrifical Altar (and it was tough trying to keep pace with how quickly our people forget their pains :lol:). As a result, temples, libraries, markets, Jaguars and axemen were built with the whip. Markets, in particular, proved useful, helping our economy and increasing happiness caps (thanks to gold and the fur we are getting from Augustus), which is why I think researching Currency was the right choice.

Here's the map of our continent:



After the second war, the larger part of our continent would effectively be ours. I don't know if that means we have enough land to catch up. I guess we could start expanding northwards and entrenching. However, if we're going to attack Augustus, I think we should do it asap (i.e. when we get trebs). Unless he has grenadiers by then :scared:
 

Attachments

  • ImmortalChallenge1_AD-780.CivWarlordsSave
    233.4 KB · Views: 91
gnp graph doesnt mean anything.
also you've been very lucky so far in that your alone with only two civs, one of which you had a chance to rush with chariots, so you dont have to worry about anyone attacking you. As far as attacking Rome with only jaguars and cats, well i havent actually looked at the save file, but if Rome has maces and longbows already, that might be a problem.
 
gnp graph doesnt mean anything.
also you've been very lucky so far in that your alone with only two civs, one of which you had a chance to rush with chariots, so you dont have to worry about anyone attacking you. As far as attacking Rome with only jaguars and cats, well i havent actually looked at the save file, but if Rome has maces and longbows already, that might be a problem.

and crossbows!
Seems a tough fight, if you ask me (which you shouldn't ;)).
Bring twice as many cats as he has units and you have a chance. But this will require the full abuse of the sacrificial altar all over the place.
Now would be a good time to lightbulb something, if you ask me (see above).
 
2/3 of this continent (Rome has a lot of bad terrain too) is enough to win if you're going the space route. Just be careful of potential Roman sneak attacks. If you want to kill Rome it's probably best to wait for renaissance as you will want to consolidate the northern half of your lands before some AI gets astronomy.

Going to nitpick, sorry. Texcoco wasting the 3f+commerce spice and a grassland just to get a mine was a bad trade for a slaver. As for the central copper site, you guys seem to have an unhealthy obsession with stealing food resources from other cities, but in this case 1S of copper stealing pigs is reasonable if it's going to be a production city. If you want a commerce city 1E is just better because it has more grassland (in this type of city I just put up a couple of farms to grow it out to maximum to work cottages and maybe whip 1-2 items then slowbuild most of the economic buildings.)
 
I'm not going crazy am I? Aelf said "lower happiness caps", but isn't the happiness cap the same on immortal/deity as it is on emperor (3)???

I agree on the whole "waiting for gunpowder" idea. I don't suppose there is much hope now of lightbulbing your way to liberalism, which is unfortunate. However, you could still probably lightbulb your way to chemistry and hope AC doesn't go for military tradition...
 
Going to nitpick, sorry.

At least I didn't get blasted ;) I thought things might not be going as well as they should. Well, your points are valid, but I think it's just a difference in perspective. I think it's better that Texcoco has some production capability to get some basic infrastructure up without having to whip. It doesn't have food specials so it takes more time to regrow. Besides, all those lake tiles are 3F3C, so the spice wouldn't be anything extraordinary.

futurehermit said:
I'm not going crazy am I? Aelf said "lower happiness caps", but isn't the happiness cap the same on immortal/deity as it is on emperor (3)???

Whoops. I assumed they were different. They were different in Civ3, right? I always get this kind of details wrong :p
 
I'm just thinking aloud, but I have the feeling a cultural victory may be possible. What do you think, all?

Are you serious? With Monty? :lol:
 
Top Bottom