AI tournament: Season 3

[Pangaea]

I like the idea of using more 'normal' BtS settings, so wouldn't prefer balanced resources. Doesn't it give 'everything' to everybody? Or did it remove some resources? I don't recall. Everybody say it's crap though.

This game is all about randomness, but by playing more games the best AIs should still shine.

It seems these games with normal settings favours peaceful wins more than the previous series, so I think we'll see more space and cultural wins, and probably some diplomatic too. I'll go for Cyrus this time. If he can expand well, he should get a ton of land. Block off the others, mate.

 

[vranasm]

Yes.


I'm not too familiar with balanced resources... what does it do exactly?
I do expect AIs to require to trade with each other as that is part of their personality.

balanced resources should guarantee every player iron/copper/horse (not sure if all together, but I guess yes) nearby starting location

I think it's easy to check how it exactly works (if one of those 3 or all of those 3) if you run couple of random maps and check WB

edit:
this settings is generally banned in HOF since back in the day it was viewed as favoring human player over AI (I guess over time the perception changed, but that is another topic for discussion), it was percieved as allowing human rush techniques each game, but I think once you run into highest levels you can quickly see why this problem is not bad and actually it starts to favor AI since you can't find easy target for horse archer rush, which is first really good rush on highest levels of difficulty.

 

[Gumbolt]

Is the intention for Toku to wipe out Gandhi?? Toku if he plays sensible could have 3 gold resources.

Gandhi start looks less than favourable. No food resources outside his BFC for 8 tiles. (I ignored the pigs tile that would require a galley. He also has the worst BFC in my opinion. As if he will chop forest? Edit - Actually he can still use the northern pigs and a cow resource. Maybe he will be okay.

Ragnor land past his first city is pretty poor. He will be settling a lot of poor cities. Let's hope financial traits pulls him through. If he uses a lot of sea tiles he could pull through. I suspect he will be backwards by the time his borders meet other AI.

In terms of starts Mansa and ottoman seem to have the strongest start. It will be a race to see who grabs the gold resources between them.

 

[Gumbolt]

Just completed a run through and if my account limit allows I could post once Kossin is done.

All I can say from my play through is these Ai have no idea how to win a game, fight wars or settle cities so they can grab any resources.

I wonder if playing this deity makes the window for wars too short?? Does this impact some AI if they tech much faster?

 

[Izuul]

In my experience Deity is the only level where you consistently get good wars. Even Immortal AI rarely stack enough troops to win offensive wars.

 

[Ray Patterson]

I don't think it reduces the possibilities of war. The first tech to slow down war is feudalism, until then teching only makes wars more possible. Pre-catapult conquests, as in proper rushes, are very hard for the AI to pull off anyway. I don't think it would be a lot more succesful with more time.

Besides deity doesn't only speed up teching, it speeds up expansion and production as well.

 

[MoneyManey]

Delurking here. Awesome thread Will follow with interest. Gandhi looks toast !

 

[kossin]

Group A Game 1 - Turn 50

The doings of each individual AI, 2000BC.

Gandhi - stuck on 3 cities but with 2 settlers on the way. Buddhism was no contest of course. Moai is a stupid move as usual this early but we can expect an early shrine maybe. He might be beaten by Washington to the spots close to his capital... hopefully not wasting too much time moving around in the process.

Spoiler :

Ragnar - 3 cities and a settler?! Pitiful ! Moai Statues strike again. My guess is he failed a Wonder attempt.

Spoiler :

Mansa Musa - wonder spam at its best AND very good expansion, grabbing the nice gold and flood plains area. He can get plenty of great cities - question is: will he be able to keep them?

Spoiler :

Tokugawa - expand all the directions! We have BFC gold, no need for 2 more - let's go to that desert I hear there's an Oasis to refresh our troops later.

Spoiler :

Cyrus - 6 cities is pretty standard for Cyrus. I looked at his troops but couldn't find evidence of a stack other than 1 unit set for attack_city near Pasagardae. Alphabet is a smart choice if he gets there first. Unfortunately it also means he won't get any wonders likely, bad for his tech rate.

Spoiler :

Mehmed - thankfully Iron Working was an early priority. I really don't understand how the AI can be so horrible at choosing city locations. He needs to gobble up land quick because his neighbours aren't waiting.

Spoiler :

Washington - I'm actually surprised by Washington. Granted, Gandhi isn't really trying to slow him or anything but his city locations are pretty good and he's expanding in the right direction.

Spoiler :

Various demographics

Spoiler :

Of interest, the real cost of Moai Statues is 150H (60%), so the time spent building them is not as devastating as could be

Land overview. Too early to tell the whole picture but...

Spoiler :

Have your predictions changed? Mehmed is disappointing me with those early jungle cities

 

[Thrar]

MM is going to become a beast if Ragnar doesn't start expanding soon.

 

[Gumbolt]

Ragnar looks a bit nerfed. Maybe a slower expansion will speed him up tech wise. Somehow i think Mansa/Mehed will have LB before Ragnar gets a chance to go war mode.

Mansa has picked up the double gold site. He could tech fast now. Same for Mehed once the triple gem site goes live.

The AI seem to settle cities based on best site given BFC. The problem is they don't take into account border pops/ techs required or food at times. It's quite amusing really. Can you really teach a computer these things?

Cyrus and Americans looks pretty good early on. Wait and see if Toku can pull off a decent first war. Once he has a metal he will 90% of time attack someone. Let's hope the stack does not include too many archers.

I still fear for the Indians. 3 cities and looks like Americans will beat him to the pigs site.

 

[enKage]

I like the way Cyrus missed gold in BFC

For now: Wash/Mansa/Mehmed. Curys dissapointed me for now, but has open way to get better

 

[lymond]

The bottom half looks interesting to me. If Mansa spreads Hindu to Mehmed and befriends him before Mehmed gets to spamming, then that could form an interesting alliance. I predict a Mansa peace vassal to Mehmed at some point if no war beforehand. Ofc, that could end up benefiting Mansa as well.

Gandhi has a very awkward position. Toke or Cyrus will probably DOW him at some point.

Wash looks strong here, but I never see him excel normally. If he stays strong late though you never know.

Rags looks pitiful over there

 

[vranasm]

rag put dissapointing performance here... maybe you nerfed him too much with the map edits?

I really like Washs expansion he is in good place to snag the double gold spot... I expect some hot heads between him and Toku soon...

Maybe with the lowered costs of Moai they are much better build then for human would be? would be interesting to hear some math about this...

I don't think Mehmed can easilly vassal Mansa since mansa has a lot of land to go still and Ragnar seems too slow to block him off.

the next 50 turns should be qi.

 

[Aje]

<- Agree with lymond and vranasm analysis.

Cyrus or Washington looks to be in good shape, but has anyone actually seen them "win" a game?
Mehmed could be strong if he starts chopping. Is the AI always this slow at chopping?

I think that Mehmed and Rags will clash at some point and that will give Mansa the chance to tech away. Offcourse the same could happen in the north depending on the religion spread (as usual). Who will get Conf and/or Juda?

Which wonder did Rags lose out on? He usually builds GLH in my games and that looks like a good choice in this one aswell.

EDIT1: I think that Gandhi will settle the northern 2xGold spot (as Washington will beat him to the coast spot). I like kossins resignated comment on Tokus desert city

EDIT2: Gao(Mansa) is one crazy city: 9xFP, 2xGold. IMO, this shows the strengh of a good early UU that can capture or defend key city spots. Go Skirmy!

 

[Ray Patterson]

Mansa's rocking and Mehmed's not, but Mansa's power is already getting too low. Toku will prob DOW Gandhi pretty soon, though I doubt his land will give him a good enough army to actually win anything.

Do barbarians in borders prompt ai to build units? If so, then Mansa's building the GW will cost him dearly.

Hard to call a frontrunner yet. Cyrus seems in an interesting position though.

 

[Fippy]

Rags should still be strong later cos his cities so far have good long term locations, and there's still much land for him.

 

[nfw]

I think that inland sea confused Ragnar.

 

[Izuul]

Looks like a hurricane forming off the coast of Nidaros. Ragnar could be in trouble.

 

[kossin]

rag put dissapointing performance here... maybe you nerfed him too much with the map edits?

I really like Washs expansion he is in good place to snag the double gold spot... I expect some hot heads between him and Toku soon...

Maybe with the lowered costs of Moai they are much better build then for human would be? would be interesting to hear some math about this...

I don't think Mehmed can easilly vassal Mansa since mansa has a lot of land to go still and Ragnar seems too slow to block him off.

the next 50 turns should be qi.

I had to delete some of it. He had room for ~6 cities towards his north. The rest of the land I removed was snow and tundra to his southeast.

I'll post next update tonight.

 

[shakabrade]

Looks like a hurricane forming off the coast of Nidaros. Ragnar could be in trouble.

Hurricanes form near equator and then move to the poles (Coriolis effect). So, Nidaros should not be hit as much, while that other city is inland enough to be safe. Phew. Hurricanes this early could really cripple one's economy. Almost like Vedic archers.

I had to delete some of it. He had room for ~6 cities towards his north. The rest of the land I removed was snow and tundra to his southeast.

I'll post next update tonight.

Did it affect domination limit? Great concept, btw. kossin, and Ray Patterson.