Apolyton University Mod (PTW version)

[alexman]

I can think of something to generate killer AIs at the expense of other AI civs, but I don't know if it will work.

Increase the optimal number of cities, and reduce the "percentage of optimal cities, so that their product remains constant.

This will increase the desire of the AI to expand, while keeping corruption unchanged.

Hopefully, this will make the AI more willing to fight for land when it already has many cities. We might even see some domination victories by the AI (imagine that!).

However, we have to be careful. If we overdo it, given enough land, the AI will try to expand more and more, and never develop its cities. (When in expansion mode, the AI builds settlers and units almost exlusively).

Any other ideas?

[Nor Me]

I do stand by my observations that the aggressive civs often come out on top, but with an addition: aggressive civs that are also either Industrious or Scientific seem to come out on top.

This is odd. I find it's the most peaceful civs with these traits that come out on top.
I have seen civs like Greece and France become killer. I infer from this that commercial is the 3rd best AI trait amd that less agressive civs tend to join the winning side in wars.

[Catt]

Originally posted by alexman
I can think of something to generate killer AIs at the expense of other AI civs, but I don't know if it will work.

Increase the optimal number of cities, and reduce the "percentage of optimal cities, so that their product remains constant.

This will increase the desire of the AI to expand, while keeping corruption unchanged.

Hopefully, this will make the AI more willing to fight for land when it already has many cities. We might even see some domination victories by the AI (imagine that!).

Yeah - I guess we've never tried it. Modders have shown that a radically lowered OCN causes the AI to cease expansion even in the face of available land. Could the converse be true?

I actually asked the question in the creation forum some time ago (look at this lonely thread) and despite a coupla dozen views, never got a reply (and obviously never tested it myself).

Probably worth a test during one of those lulls between AU games.

Catt

[alexman]

Another thing to consider is whether this trick will put the AIs constantly at war with each other when they have no more land to settle.

If so, this is something a human can exploit by staying out of the wars, building up, and grabbing land and resources as the AI cities change hands and have their cultural borders reduced.

I'm not sure if it's better for the AI to be aggressive or not. I guess it depends on the situation. I've seen AI civs get very powerful because they built up their infrastructure instead of building units and losing them in wars. I've also seen AI civs get powerful because they rolled over another civ and then built their infrastructure. It all depends on the swiftness of the wars, just like for the human. But the problem is that the AI doesn't know when to go to war and when to stay out of it. High aggression might make them be in wars of attrition too often, while low aggression might make them miss opportunities.

Let's keep looking!

[Dominae]

alexman, I would like to give a big thumbs up to the changes on the new version of the AU mod:



Specifically:

1. Doubled Entertainers

This doesn't appear to unbalanced. It may be because of the particular geography of the map in AU202 (SPOILER: abundance of Plains) which didn't favor the kind of "abusiveness" I was mentioning (irrigate everything to get cities way up in pop, then go back and mine). But, in general, you were right: I do not think it is possible to "break" the "super-entertainers".

However, dealing with unhappiness (especially during Anarchy) is easier for the human player. This is a side effect, nothing more, but one to keep in mind as a difference between the AU mod and stock Civ3/PTW.

2. AI Workers

The AI consistently uses mines more often than irrigation in my games. There are some times when irrigation is abundant, however, so I'm not sure what exactly is going on. Probably has something to do with the exact tile composition. I'll post some screens of particularly nice terraforming (SMAC term) shortly.

3. Tech Tree

I've found an unfortunate aesthetic consequence to attaching certain Diplomatic abilities to techs that come after Writing (Literature, etc.): when you obtain those techs, an Advisor asks you if you want to Establish embassies again (redudantly, since were already given the option with Writing). This is very minor, and is actually a nice reminder.

4. Taxmen and Scientists

My only negative comment: placing the pictures of these Specialists on the approriate techs in the tech tree looks horrible, IMO. Have them removed!


That's all I can think of for now.


Dominae

[Dominae]

Although the choice of Wonders could be better, I'd say this is progress:

[alexman]

Originally posted by Dominae
alexman, I would like to give a big thumbs up to the changes on the new version of the AU mod:

Are you sure you're the real Dominae?

The AI consistently uses mines more often than irrigation in my games. There are some times when irrigation is abundant, however, so I'm not sure what exactly is going on. Probably has something to do with the exact tile composition.

This is actually not a AU mod fix. It's Soren's fix, new in 1.14f. See the Why, oh why thread for some screenshots.

when you obtain those techs, an Advisor asks you if you want to Establish embassies again (redudantly, since were already given the option with Writing).

Yes, thanks for pointing this out. There is an easy way to fix it. Alliances, which is another ability of Writing (with equal AI value) doesn't produce this popup.

My only negative comment: placing the pictures of these Specialists on the approriate techs in the tech tree looks horrible, IMO. Have them removed!

They get placed there automatically when you attach them to techs. If they look so bad, we can definitely undo the changes. Their AI value is relatively low anyway.

Thanks for the report!

[alexman]

• Increase aggresion of Chinese to 4: These guys are one of the most dangerous civs when used aggressively (can you say "Archer rush?")
• Offensive units as a build-often for the Chinese: To build those Archers and those Riders. Especially now that they're more aggressive.
• Reduce aggresion of Babylonians to 2: This is the ultimate builder civ. Reduced aggression should stop them getting into wars instead of building.
• Trial: Increase optimal number of cities for all map sizes by 50%: As discussed above, this might make the AI more aggressive in grabbing land. When it runs out of room, and is strong enough, it will hopefully attack its neighbor. The goal is to generate a couple of "killer AIs" in each game, instead of a bunch of pretty good ones.
• Trial: Reduce percent of optimal number of cities for all levels by 2/3: To keep corruption unchanged after the increase in OCN.

I will test the OCN/percentage change by playing two OCCs (one with the change and one without).

Comments?

[Dominae]

If we're going to make the Chinese more aggressive, we might as well do so for the Iroquois as well (again, up to 4 this time). In AU202, they really are not taking advantage of their UU. Just a thought...


Dominae

[Dominae]

Also, could you explain again why are higher OCN ultimately leads to Killier AIs? I understand that they will want to expand more (with Settlers), but they do this quite well already. Does the AI attack another civ in part because it's under the OCN?


Dominae

[punkbass2000]

Originally posted by Dominae
Also, could you explain again why are higher OCN ultimately leads to Killier AIs? I understand that they will want to expand more (with Settlers), but they do this quite well already. Does the AI attack another civ in part because it's under the OCN?


Dominae

Yes, though I would phrase it differently. The AI always wants to reach its OCN, and thus, if it cannot do so peacefully, will resort to war. This isn't to say it won't go to war while it can still expand peacefully, incidentally.

[Dominae]

Originally posted by punkbass2000
Yes, though I would phrase it differently. The AI always wants to reach its OCN, and thus, if it cannot do so peacefully, will resort to war. This isn't to say it won't go to war while it can still expand peacefully, incidentally.

That's what I meant; thanks for the clarification (English is my mother tongue, really!).


Dominae

[Ozymandias]

Originally posted by ducki
Hey guys!

I thought of a potential "fix" for the Great Wall, if the current one - granting free Walls in all towns - is insufficient.

Change it to The Aqueducts, a new great wonder that grants free Aqueducts to all cities. I think that would make it worth building for both the AI and the human - we'd have to make it more expensive, probably - and it retains, IMO, the "flavor" of Civ as well as posing the player with a strategic choice by giving yet another valuable beeline in the Ancient Age.

Thoughts?

Would the Civ Engine then place Aqueducts in cities which didn't have water nearby? -- Sorry, I'm a bit behind the curve in a few of these threads, but IIRC there was something about a Wonder placing Harbors in landlocked cities.

-Oz

[Ozymandias]

Originally posted by nbarclay


Ironclads don't belong in the battleship age. That's why providing upgrade paths for frigates and ironclads is important. And while Civil War ironclads weren't much good in deep ocean, successors that were came close enough after (if I recall correctly) that it's reasonable not to penalize ironclads in deep water given the time span they cover.

After the Monitor/Merrimack-Virginia clash, EVERY European navy immediately cancelled all non-iron-armored warship construction.

An historically accurate upgrade path might (post-Monitor) be --

Ironclad steam "frigates"
Early battleships (Remember the Maine!)
"Modern" battleships (dreadnoughts and the pre-WW1 arms race)

-Oz

[alexman]

I tested the increased OCN to make the AI more land-hungry, and it worked as expected. However, I'm not sure that this is what we want.

Have a look at the attached picture. The one on the left is with a 50% higher OCN, and the one on the right is standard OCN. Both with the AU mod, otherwise.

In the Western continent, the Japanese dominated in both cases. In the modified OCN version they were more aggressive and immediately got into a war with China, whom they destroyed. In the standard version, they got into a war with the nations more to the South (Persia, France), leaving their Chinese neighbors alone.

In the Eastern continent, everyone is pretty balanced in the standard version, and stayed peaceful until the industrial age. But in the modified version, the Iroquois dominate.

In other words, what we expected, happened. The AIs try to gobble up each other. But was the result a more "killer AI"? In the modified version we got a bigger Iroquois empire, but it came at a very high cost because they were almost constantly at war. The AIs were fighting with tanks in 1750 when I stopped the standard game. In the modified OCN game, they were using Cavalry and Infantry, so they were half an age behind.

To summarize, increasing the OCN results in more wars. Strong empires will dominate and prosper, but empires with no significant advantage over their neighbors will enter long wars of attrition and stagnate. Communism will be used more frequently, and the research rate will drop.

So what do you think? Should we make the change? Is the human better off dealing with many smaller AI civs and face a faster research rate, or dealing with with fewer, bigger AI civs at slower research rate? I am inclined to say that the former case is presents a greater challenge, but it depends heavily on the map and starting locations.

[punkbass2000]

I think we'd be better off with the OCN change, at least o nthe higher levels. That way the tech rate will be high enough anyway, as well as creating Killer AI.

[Theseus]

25% ?

[Catt]

Originally posted by alexman
So what do you think? Should we make the change? Is the human better off dealing with many smaller AI civs and face a faster research rate, or dealing with with fewer, bigger AI civs at slower research rate? I am inclined to say that the former case is presents a greater challenge, but it depends heavily on the map and starting locations.

Thanks for testing I think its worth a few more tests. I am inclined to agree that smaller, tech-robust civs may present a larger challenge due to the "come-from-behind" and "help-the-trailers" concepts built into the game, as well as the AI advantage of the AI-AI trade rate. This assumes (1) that the increased OCN does present a tech slowdown (over a series of games), and (2) the tech slowdown can be avoided by the human (!). If the slowdown is worldwide (including the human), then I think the increased OCN makes sense. [Did you play these as OCC's? so it is tough to gauge the possible human tech rate?]

I'b be inclined to play an AU game or two (or a couple of side games) to get a sense of the effects across a larger number of samples and in a "regular" game environment.

Catt

[alexman]

Catt, the worldwide slowdown affects the human, just because he can no longer be helped by the AI's research as much. Otherwise the OCN itself does not directly change the tech rate. Since corruption stays the same, the only reason for the slowdown is the fact that the AI is more often at war.

By the way, I didn't do any research in these OCC games. I just plopped a city on a one-tile island, produced wealth, and had the enter key pressed until 1750 A.D.

I think Theseus' suggestion of increasing the OCN by 25% makes sense. Soren made the AI stop expanding at the OCN mainly for corruption reasons. This behavior has not changed since the game first came out, but what has changed is the amount of corruption in the game. After the patches, police stations increase the OCN by 25%, but the AI does not consider this when deciding when to stop expanding. It makes sense to have the AI aim for 25% more cities than pre-patches.

[Catt]

Originally posted by alexman
Catt, the worldwide slowdown affects the human, just because he can no longer be helped by the AI's research as much.

[. . . ]

I think Theseus' suggestion of increasing the OCN by 25% makes sense. Soren made the AI stop expanding at the OCN mainly for corruption reasons. This behavior has not changed since the game first came out, but what has changed is the amount of corruption in the game. After the patches, police stations increase the OCN by 25%, but the AI does not consider this when deciding when to stop expanding. It makes sense to have the AI aim for 25% more cities than pre-patches.

25% would be a good test as well. I'm all for it.

Just to be clear on my point about tech rates, though -- in one sense it doesn't matter if the world is using infantry or tanks in 1750 AD -- what matters is whether or not the human has a greater chance at a decisive tech lead under the modified or unmodified game. If the human gains no appreciable tech advantage over the AI's as a whole, then I think the OCN experiment might actually go a step or two towards creating a smaller number of "Killer AI's" which pose a real threat to the human (which I hope will be an improvement) rather than a larger number of robust AI's which individually prevent little threat in the late game.

Catt

[alexman]

Originally posted by Catt

If the human gains no appreciable tech advantage over the AI's as a whole, then I think the OCN experiment might actually go a step or two towards creating a smaller number of "Killer AI's"

As you say, further tests in a number of different map settings is required to say for sure. But my guess is that at higher difficulty levels (where the human relies on the AI to at least help him out doing the research) the human research will slow down with the AI in the modded OCN. At easier levels (where the human does most of the research), the AI wars will increase the gap.

Here are two examples of how I imagine the effects of the increased OCN:

Scenario 1: The human has been playing catch-up all game, and finally is at the position to attack. Catching up will actually be easier with the increased OCN (assuming he manages to stay out of the wars), but it will be tougher to defeat the surviving AIs. So this scenario is probably tougher with the increased OCN.

Scenario 2: The human has been among the strongest civilizations, and is ready to pull ahead for good. With the higher OCN, assuming that he survived the increased AI aggression, it will be easier to win the game, because the AI will be further back in tech. But it will be less likely that the human will be left alone to build and do his research unmolested. So it's hard to say if it would be more of a challenge.

[Catt]

I was imagining the same scenarios. If the human can avoid warfare (or at least allow it to only slighty impinge on other growth needs), the increased OCN may in fact make it easier. Certainly worth testing though

Catt

[joncnunn]

I think the increasing OCN would also increase the number of cities that needed built before the FP could be built.

In a recent game on a tiny map on Emperor level, I was running a very well empire with 10 cities (5 adjoining rivers, 8 coastal). There was hardly any corrpution / waste at all when the FP was completed. (Only mod was double hit point.)

Now if the OCN had been increased to 21 I wouldn't have been able to build the FP, and so would have had more corruption than current current rules.

[alexman]

You're right that the FP can be built as soon as you get half the OCN, but most players build it much later than that anyway.

As for your example, on a tiny map (OCN=12) you can build your FP as soon as you get 6 cities. Increasing the OCN by 25% to 15, would mean you could build the FP at 7 or 8 cities. Not such a big deal.

[Jaybe]

alexman, instead of increasing OCN by 25%, how about changing it so you can get a better handle on the affects and then tone it down as appropriate.

Go overboard, I say. One Hundred Percent increase.

(I have been using 50% increase for several months in my own (non-AU) mods, but with the purpsoe of corruption modification).

[alexman]

Jaybe, thanks for the suggestion, but I think I already did go overboard with the 50% increase. See the OCC experiment in the previous page. The AI was at war for almost the entire game.

[Soren Johnson]

Originally posted by alexman
Jaybe, thanks for the suggestion, but I think I already did go overboard with the 50% increase. See the OCC experiment in the previous page. The AI was at war for almost the entire game.

I would highly recommend that you run the experiment a few more times. Just off the top of my head, I wouldn't assume that a higher OCN would lead to a more aggressive AI. Perhaps the opposite, in fact.

[Catt]

Originally posted by Soren Johnson Firaxis


I would highly recommend that you run the experiment a few more times. Just off the top of my head, I wouldn't assume that a higher OCN would lead to a more aggressive AI. Perhaps the opposite, in fact.

Why is that?

On a larger point, care to share anything off the top of your head that might contribute to AI civs being more aggressive about exploiting a relative advantage? Just pushing the aggressiveness sliders up makes every civ a Bismarck / Shaka psycho that attacks all the time - with or without advantage. We'd sure like to stimulate a slightly more aggressive AI, but one that employs a "smart aggressiveness." 'Course, I guess making every aspect of the AI "smarter" is the whole trick, isn't it?

Thanks for continuing to check in on us!

Catt

[Dominae]

Originally posted by Soren Johnson Firaxis
I would highly recommend that you run the experiment a few more times. Just off the top of my head, I wouldn't assume that a higher OCN would lead to a more aggressive AI. Perhaps the opposite, in fact.

Why?

Perhaps because the AI remains in its peaceful "REX" phase (or something like it) until the OCN is reached. Once its empire is fully "fleshed out", it enters "production" or "military" phase, and looks for reasons to pick fights. Thus, the OCN and AI aggressiveness may be inversely related (at least indirectly). Just a guess, though.


Dominae

[alexman]

This for sure is true if the AI is under the OCN and has room to expand. But is this the case if it runs out of room too? It certainly didn't look like it, but I'll try again on a different map.

PS. Any help to get more data would be appreciated.