Ptw Ai

[DrSpike]

Is it the same as civ3 1.29f?

[Jaybe]

Nope. Soren made some "tweaks"!

Not much, just terrain improvement, building, and military differences. And a total revamp of barbarian behavior.

All from a few tweaks.

[DrSpike]

These would be undocumented 'tweaks' I take it.

[=DrJambo=]

anyone care to elaborate a little more on this?

[alva]

Take a look a the strategy forum, there a few trheads about this

One of them is not irrigating grassland in despotism.(unless needed)

[Skeeve]

I suppose that the AI should have been looked at with regard to the new rules. For example, when playing on Elimination the AI strategy is quite poor.

[DrSpike]

I haven't played much since 1.21f, and I'm too lazy too hunt through lots of old threads. What's the current situation on the 2 old hot topics.....corruption and AI tech trading rates?

[alva]

None of the new styles,elimination and others, are
ment to be played against the AI. Use these only in hman versus human games.

corruption: is in IMO just about right.
techtrading: has been altered in several ways, you can do you own research now

[Stuie]

Originally posted by DrSpike
What's the current situation on the 2 old hot topics.....corruption and AI tech trading rates?

I'm not sure either is considered a "hot topic" since they opened them up in the editor. That said, I find corruption to be perfect as is, but I modded the AI tech whoring down to a more managable level (I think I set it at 110).

[alva]

hmm, I agree;

The game 'needs' alot of corruption to function, if not, taking cities becomes overpowering.
Granted, it can be annoying at times, but it does act as a sort of equilizer between the top-dog and the little guys IMO.

[DrSpike]

Well I started a game last night, so I'll soon find out.......but does the AI bombard properly yet?

It's a deity game so I'll follow the classic buy techs until quite late approach......maybe tech trading isn't so high now but I expect it is still very fast.

[Stuie]

Originally posted by DrSpike
Well I started a game last night, so I'll soon find out.......but does the AI bombard properly yet?

No. AI use of artillery is seriously lacking - I saw the AI use one catapult once (and I mean, once) in the open field in my last game. I think this only happens when the AI manages to capture artillery and then come across a target of opportunity while taking it back to one of its cities.

I'm a big proponent of softening up targets with a nice bombardment prior to attacking, so I feel like the AI puts itself at a disadvantage by not using it.

[zulu9812]

I've actually noticed that the AI isn't really capable of handling diplomatic negotiations with all those extra civs. I was playing on a world map with 27 civs, and the AI civs were constantly declaring war, making peace, and declaring war again. One time, England made peace with Spain and then made a military alliance with Russian against them - all in one turn-change! The AI also doesn't give a rat's arse about keeping agreements for 20 turns, and doesn't seem to suffer a huge degradation in reputation either.

[FNBrown]

Originally posted by zulu9812
The AI also doesn't give a rat's arse about keeping agreements for 20 turns, and doesn't seem to suffer a huge degradation in reputation either.

This is the most frustrating part of the game for me. I really like the idea of forming long-term alliances with certain Civs for strategic reasons, and I'm very careful about keeping my end of the bargain on all treaties, agreements, and alliances I sign.

I really wish the AI would do better about keeping treaties.

Of course, the problem is that a human player could exploit the "sucker" AI, knowing the AI would keep a treaty and not be capable of "treachery," but I think if you have a solid diplomatic reputation, the AI should be programmed to keep its treaties with you.

I also think long-time friendships should count for something. A neighbor that you've shared military alliances with in the past should not suddenly decide to attack you for no reason.

[Catt]

Originally posted by DrSpike
It's a deity game so I'll follow the classic buy techs until quite late approach......maybe tech trading isn't so high now but I expect it is still very fast.

It can be a little harder to execute that well since 1.29f made changes to the tech pace.

Long and short of it is that techs "act" as if they are more expensive (without digging into the mechanism by which this is accomplished); it's easy for AI civs to fall out of the race (whereas under 1.21f my view was, small or large, each Ai is going to be pretty close to tech parity); and it's easy to find yourself far enough back that buying and conquering your way into modernity is not quite the cakewalk it used to be.

EDIT: All good IMHO because it forces a more balanced approach rather than the rote standard regardless of circumstance -- but it makes it harder on the human player.

Catt

[DrSpike]

Well I knew the game inside out 1.21 and before......my best deity landing using this strategy was stupidly early, around 1050, so i'm hoping now there is a little more challenge.

[Inverse Icarus]

the AI has a better sense of overlal strategy these days.

[Catt]

Originally posted by DrSpike
Well I knew the game inside out 1.21 and before......my best deity landing using this strategy was stupidly early, around 1050, so i'm hoping now there is a little more challenge.

You'lll probably still whip the AI, but you won't launch in 1050 AD.

The mechanism by which the tech valuations have changed (which I didn't go into last time) was simple in conept, but produced wildly different tech pacing. In 1.21f and prior versions, the human always researched at a "base rate" of 10. At Deity, the AI's rate was 6. At Chieftan, 20. Now, the AI base rate never gets below 10 -- at levels higher than Regent, the human's base rate increases (instead of staying at 10).

You might still coast off of the AI's research -- but the AI's research doesn't move nearly as quickly as in 1.21f and prior.

Above Regent, I think this does two principal things: (1) makes individual research more valuable, since the "floor" price for any tech is higher than prior versions (and which means, since gold geneeration opportunities are unchanged, more expensive techs and the trend I mentioned earlier about more easily falling out of the race completely); and (2) prolonged the game. Point 1 makes it harder to do the zero research strat and tends to force a more balanced approach on the human. POint 2 makes it a bit easier -- more turns for the AIs to drag themselves (with a little prodding) into fruitless wars of mutual stagnation.

I'd agree with much of what others have said in the thread and overall I find it a more challenging and engaging AI. But if you were regularly beating Deity in the early 1000's - 1200's AD before, it's not as if you're going to have to drop back to Monarch and "retool" your gameplay skills.

Catt

[DrSpike]

The thing is early landings are often less impressive on deity.

Once you have the strategy down pat it isn't that hard. Having said that there are still exceptions: I played a fascinating comparison game on a smallish map with 3 AI civs, one of which died early (the one on my continent. ) where a guy who contributed to the research pool early beat me by quite a few turns.

Of course, since defence is often stretched in such games it would make things harder if the AI upgraded, and I have seen some favourable comments on this whilst browsing the forums.

[Saint Marcus]

And a total revamp of barbarian behavior.

meaning? I'm anxious to know.

[alva]

Barbs have researched intelligence
They won't attack you just about anywhere.
No more setting a spearman on the mountaing and waitnig for it to become elite.
They will hunt down you're workers, .....