Today I Learned (about Civ3)

Quintillus

Restoring Civ3 Content
Moderator
Supporter
Joined
Mar 17, 2007
Messages
8,421
Location
Ohio
I thought it might be fun to discuss things that we are learning/re-learning about Civ III even after innumerable years of playing it. There's enough variety in the game, including its scenarios, that a truly comprehensive knowledge of every in and out can be challenging to obtain.

I'll start...

Today I learned that Solar Plants are Agricultural improvements, and thus half-price for Agricultural civs. I'd known about Aqueducts and Hospitals, but didn't know about Solar Plants. In part because many of my games had ended/were ending by that point, in part because I almost always build or, failing that, conquer Hoover Dam and thus rely on hydro power. But this was a 20K game, and I was skipping Hoover Dam due to its low culture relative to other options, and noticed that for some reason my Solar Plants were equal in cost to the normal half-as-expensive Coal Plants, and cheaper than Hydro.

What have you learned about Civ III recently?
 
I'd known about Aqueducts and Hospitals
Strangely enough the later are not agricultural, while recycling center are. It being the other way around may make more sense. It would however change the balance even more in favour of agricultural civs.
 
Not so much today but relatively recently. The separate launcher EXE that's exclusive to the gog version fixes the pink text on some of the advisor screens. I always assumed it basically did nothing useful and could be ignored.
 
Been playing Civ3 since 2009, and Conquests since 2013, but it was only a couple of weeks ago that I posted this in "Single Player Tips" thread:
Was surprised to discover a couple of days ago that the Dromon (and presumably also the Man-O-War) can start a GA by sinking a Privateer -- even though Privateers have Hidden Nationality
 
As long as they are not flagged as Barbs, though I never ran into it myself.
 
Not today, but about a week ago. And it's kinda embarrassing, but .... i just found out that you can mass upgrade units much easier via the military screen (F3) and right-clicking on them. There's no need to go into every single city of the map and upgrade the units manually. (actually one of the reasons I always prefered to play on small maps with fewer cities :D )

Only took me about ...well...er... 18 years to figure out? :D
 
Last edited:
Shift-U also allows mass-upgrading -- assuming sufficient funds are available to cover all units of the current type which are currently stationed in a Barracks-town.

And if you're using the Flintlock patch, Ctrl-U will allow stack-upgrading of all currently activated units of a given type, in a single Barracks-town.
 
Not today, but about a week ago. And it's kinda embarrassing, but .... i just found out that you can mass upgrade units much easier via the military screen (F3) and right-clicking on them. There's no need to go into every single city of the map and upgrade the units manually. (actually one of the reasons I always prefered to play on small maps with fewer cities :D )

Only took me about ...well...er... 18 years to figure out? :D
Shift-U also allows mass-upgrading -- assuming sufficient funds are available to cover all units of the current type which are currently stationed in a Barracks-town.

And if you're using the Flintlock patch, Ctrl-U will allow stack-upgrading of all currently activated units of a given type, in a single Barracks-town.
I know how to upgrade single units from the F3 screen, but not en masse. How does that work?
 
You have a list of Civ 3 hotkey commands in your civilopedia during your game, among them the Shift-U command for upgrading all units of a certain type when the upgrading prerequisites are met. The screenshot below shows only a part of those hotkey commands:

Hotkeys.jpg


Here is another very helpful shortkey, that many civers are not aware of. By clicking on the cycle through cities button, this button changes its image and now you can cycle through all your cities in disorder. You don´t have to search them, what can be very helpful, especially when playing on big maps in the later stages of the game:

Cycle through the disorderly cities.jpg
 
You have a list of Civ 3 hotkey commands in your civilopedia during your game, among them the Shift-U command for upgrading all units of a certain type when the upgrading prerequisites are met. The screenshot below shows only a part of those hotkey commands:

View attachment 682493

Here is another very helpful shortkey, that many civers are not aware of. By clicking on the cycle through cities button, this button changes its image and now you can cycle through all your cities in disorder. You don´t have to search them, what can be very helpful, especially when playing on big maps in the later stages of the game:

View attachment 682494
Thanks. Something else I never looked at was the "sentry" action, so I'm not terribly familiar with it.

Then again, I haven't finished a game in probably over a year ... started another one this morning in a bout of insomnia, which hopefully I'll get to experiment a little bit with (Monarch/Continents-80%/Byzantines).
 
Shift-U also allows mass-upgrading -- assuming sufficient funds are available to cover all units of the current type which are currently stationed in a Barracks-town.
Thank you very much. A true life saver!


Today I ran into an interesting "AI science issue". Playing on a standard size map with 21 Civs. I'm used to the AI developing rather "slowly but steadily", extensively trading tech with each other. I guess everybody knows that. Usually you end up with most bigger Civs being en par in terms of scientific development unless one Civ is really dominating the scene.

In my latest game China is different. It is at least 3-4 developments ahead of me and everybody else, despite having the same amount of cities as everyone else (well, give or take 1 city maybe). Yet it fields Cavalry (and makes liberal use of them everywhere!) whereas everybody else is still on knights / longbow, some just recently. I found this a remarkable "sprint" by the Chinese, something I haven't seen in other games and only in situations where a single AI Civ completely dominates the scene. This might be more prevalent in higher difficulties, but I still suck at Regent levels. So that was a new experience for me :cool:
 
Thank you very much. A true life saver!


Today I ran into an interesting "AI science issue". Playing on a standard size map with 21 Civs. I'm used to the AI developing rather "slowly but steadily", extensively trading tech with each other. I guess everybody knows that. Usually you end up with most bigger Civs being en par in terms of scientific development unless one Civ is really dominating the scene.

In my latest game China is different. It is at least 3-4 developments ahead of me and everybody else, despite having the same amount of cities as everyone else (well, give or take 1 city maybe). Yet it fields Cavalry (and makes liberal use of them everywhere!) whereas everybody else is still on knights / longbow, some just recently. I found this a remarkable "sprint" by the Chinese, something I haven't seen in other games and only in situations where a single AI Civ completely dominates the scene. This might be more prevalent in higher difficulties, but I still suck at Regent levels. So that was a new experience for me :cool:
I hate to find myself agreeing. My only 3 personal AI observations are:

1) China can overachieve generally, including as you describe by somehow doing better on tech than scientific Civs of a similar empire size and getting a medieval tech sprint that I can't compete with. Even when they haven't had a successful war. I haven't noticed this since I changed their traits, but that has only been a few games.

2) India can often be the opposite. Gets a big empire, avoids wars but somehow falls behind on tech. I often look at Ghandi and wonder what is he actually doing? To the extent that if I need him to be a strong buffer against an aggressive AI I am tossing him a dozen techs just to get him to pikemen and his ruddy elephants.

3) Persia can also overachieve in the late game and catch up in the Industrial era even if coming from a fairly low basis. They seem to consistently pick good fights late in the game and are the most likely to emerge from mid-pack to become a genuine contender. Even though his UU is really early on.

I usually just chalk all of this up to coincidence and/or me being delusional. I have zero opinion on any other AI Civs since the Hittites have gone on a long run with me being normal.
 
Thinking about why that might be...

China's AI is set to build Growth, Production, and Science frequently. So, lots of Granaries, Aqueducts, Libraries, and later, Universities. It makes sense that this could lead to a highly successful Middle Ages, especially with the Rider and their Militaristic trait to deal with any military challenges efficiently. Their other trait, Industrious, will also lead to a better economy.

Persia's AI is set to build Offensive Land, Wealth, and Trade often, while being Industrious and Scientific. The latter two would harmonize well in the Industrial Age, with cheap Libraries/Universities, faster improvements including railroads, and a higher shield base from which to start industrializing. I'm less sure how the Wealth and Trade focuses would aid - maybe more Marketplaces meaning fewer Entertainers meaning more productive cities in those larger-city eras of the game? - but the Offensive Land focus would mean that Persia is more likely to expand, especially if it's running low on things to build in the Industrial Era.

India is Religious and Commercial, and focuses on Growth, Wealth, Trade, and Culture. So, Gandhi doesn't really care about Pikemen, he'd rather be building Temples, Cathedrals, and, likely, Colosseums. His people may be happy, and the Growth focus helps in establishing that big empire, but India would be perfectly fine to lay back, relax, and focus on a Cultural Victory without ever having to build a Pikeman. Perhaps the 1/5 aggressiveness also decreases the incentive to build a military or research military techs.

How much influence do the "Build Often" traits have in practice? I'm not sure, and if there's been a study done on it, I haven't seen it. But assuming they do have a noticeable impact, the flags checked for these civs seem to plausibly create the patterns Fergei describes.
 
Thinking about why that might be...

China's AI is set to build Growth, Production, and Science frequently. So, lots of Granaries, Aqueducts, Libraries, and later, Universities. It makes sense that this could lead to a highly successful Middle Ages, especially with the Rider and their Militaristic trait to deal with any military challenges efficiently. Their other trait, Industrious, will also lead to a better economy.

Persia's AI is set to build Offensive Land, Wealth, and Trade often, while being Industrious and Scientific. The latter two would harmonize well in the Industrial Age, with cheap Libraries/Universities, faster improvements including railroads, and a higher shield base from which to start industrializing. I'm less sure how the Wealth and Trade focuses would aid - maybe more Marketplaces meaning fewer Entertainers meaning more productive cities in those larger-city eras of the game? - but the Offensive Land focus would mean that Persia is more likely to expand, especially if it's running low on things to build in the Industrial Era.

India is Religious and Commercial, and focuses on Growth, Wealth, Trade, and Culture. So, Gandhi doesn't really care about Pikemen, he'd rather be building Temples, Cathedrals, and, likely, Colosseums. His people may be happy, and the Growth focus helps in establishing that big empire, but India would be perfectly fine to lay back, relax, and focus on a Cultural Victory without ever having to build a Pikeman. Perhaps the 1/5 aggressiveness also decreases the incentive to build a military or research military techs.

How much influence do the "Build Often" traits have in practice? I'm not sure, and if there's been a study done on it, I haven't seen it. But assuming they do have a noticeable impact, the flags checked for these civs seem to plausibly create the patterns Fergei describes.
An interesting analysis of my hallucinations! I would have thought playing at Emperor difficulty the AI would be buildinf pretty much every building (in peacetime) pretty much when it becomes available?

Build often for units is meant to be pretty negligible.


I assume it is the same for buildings, hence why I think I may be seeing patterns that don't exist. As humans are wont to do. It was just funny to read someone else having the same experience with Mao. The vast majority of times in my experience China are fairly normal (albeit consistently good), but India lagging behind average on tech (relative to empire size) and Persia coming to life in the Industrial era (geography permits) seem staples of my games. Both human and AI only test games.
 
I would have thought playing at Emperor difficulty the AI would be building pretty much every building (in peacetime) pretty much when it becomes available?
This idea might be somewhat true, but to a significant lesser degree than you imply. Emperor gives only a 20% discount for AI, this tends to be more than offset by its incompetence. Since (good) human player will often not have the means to build everything once it is available, same is true for AI. Still the analysis by Quintillus might be overestimating systematic effects compared to pure chance. Chance (in a wider sense) plays a major role.
 
This idea might be somewhat true, but to a significant lesser degree than you imply. Emperor gives only a 20% discount for AI, this tends to be more than offset by its incompetence. Since (good) human player will often not have the means to build everything once it is available, same is true for AI. Still the analysis by Quintillus might be overestimating systematic effects compared to pure chance. Chance (in a wider sense) plays a major role.
Sorry, I keep forgetting to reference modifications. My Emperor difficulty has a cost factor of 6 (the old Deity or Demigod?). Plus I was also wrong to assume that just because 'build often' doesn't vary unit production preferences much that it wouldn't impact on city improvement build preferences. Especially when traits make different buildings much cheaper (whereas there is no such unit cost variations caused by traits).

Anyway, today I learnt after Quintillus's response to my paranoia that there is an impressive variation in 'build often' preferences for the various Civs, which must explain why they can sometimes feel different, beyond just their aggression levels.
 
So after my post on the topic I thought I'd have fun and set up a quick game on a small map with India, China and Persia as some of the Civs to test my prejudices about these three Civs. Unfortunately I was on another continent to India and China so couldn't see what they were doing. But it appears in this game, they didn't really fall into their respective habits of tech slump and tech sprint (although India may have squandered a strong start according to the graph). Persia, however, presented an absolutely beautiful example of what Persia does in about 50-60% of my games. I was his unfortunate neighbour Sumeria, and this is a brief history of an example of my experience of the crazy Persian Industrial era power spike. Here is the map at the end of the turn after Xerxes had declared war on me and frankly beaten me (taking Akshak to the north and a city beginning with Z to the south in the first of what would likely be many of my cities before he would ever speak to me).
1706285411140.png


And here is the power graph showing Persia heading for lift off in the Industrial era (you can just see the Persian graph starting to jutt out to the left as mine falls by the equivalent amount having lost two cities). Note that his score is completely ordinary. A thoroughly middle of the pack Civ for the ancient and medieval eras and before his Industrial era aggression his score was only 4th out of the 6 Civs on our continent. Any other Civ in the game and I would not be concerned about them as we approach the Industrial. But for this one I had my eyes on him given my recent posts on the subject.

1706285659658.png


Overview of a typical Persian AI's game

Ancient era - Persia borders me and a somehow puny Scandinavia that botches its expansion phase (maybe there were barbarians in his neighbourhood?). Persia is stuck in the north of the eastern continent, with my mighty self as Sumeria, keeping him bottled in and on good terms. Persia behaves like a typical stupid Civ 3 AI in this era. He doesn't attack the puny Scandinavians that have only 4 cities. No, instead he sends his Immortals to fight the advanced Byzantines in the south east of the continent. Byzantine is my preferred tech trading partner as the French and Mayans scare me a little. Xerxes has left it too late , Byzantine has pikemen and the war quickly fizzles out into a stalemate.
Summary: Persia has squandered a really promising opening with a weak neighbour and sank to the lower middle of the pack amongst the AIs by picking a stupid long distance war and effectively wasting his insanely powerful UU (another trait of his).

Medieval era - Byzantines and myself are firm friends adn MPP after France attacks me. It delays our teching a little bit, but we are preferred trading partners and I trade with them every tech I get. We are both scientific and trading techs at pretty much full tilt. I am absolutely prioritising Education and Universities for culture effect as I have some questionable borders that are at risk of flipping. However, Persia, despite still being limited to up in the north of the continent is notably out-teching our combiined efforts throughout the Medieval era. Byzantine drops out of the war with France so I MPP with Maya to rebuff the mighty French (and end up picking off a couple of French cities, one of which was originally Byzantine). The point here is Persia attacks the Byzantines again (again ignoring the low hanging fruit of Scandinavia) and his knights have no greater luck against the Byzantine pikemen and musketeers. It is a fairly short war and Persia manages to maintain his tech lead despite again being a stupid Civ3 AI making poor military decisions. With the French aggression subdued, I dissuade further aggression from the Joan of Arc (and culture flips) by garrisoning the bulk of my military in my new French conquests. I am aware though that my game has become all about racing to Nationalism (with the Byzantines) to get Riflemen and draft, so that we can jointly repel the Persians, hopefully before he gets Cavalry.
Summary: Again, sub-optimum by Persia. He could have mauled the Scandinavians or even had a go at me as Sumeria whilst I am stuck in a war with France. He looks geographically locked in and contained. He is however, teching like a god despite being caught in a war and is outteching the combined efforts of 2x Scientific Civs of broadly comparable power and wealth.

Industrial era - the Byzantines and I combined cannot keep up with the Persian tech and he gets Cavalry about 15 turns before I could possibly get Riflemen. Nobody else in the game has cavalry. It is clear he can do me serious pain at this point of the game (due to the Deity level cost factor I have for the AI and the fact my military is outdated due to my focus on teching above all else). I am still trying to keep him sweet so have given him ROP for much of the game and thrown in gold per turn bonuses and the odd luxury. Unforunately he uses the ROP to sweep his cavalry down to the Byzantines and quickly chomps off half of her territory. Third time has proven a charm for Xerxes. I manage to let the ROP with him lapse to reduce his military effort against Byzantine and he sues for peace with her and then boats over some possible reinforcements to his new southern colonies. I get no luck and my free Industrial tech is Steam Power, not Nationalism. Riflemen are about 12 turns away and I cannot trade for them. Frankly, I've neglected Xerxes a little as with my custom domination rules (30% territory and population) I am focussed on curtailing Japan from a domination victory on the other continent and France still hates me and I need to avoid culture flipping my French conquests. Xerxes uses his successful doomstack in conquered Byzantine (reinforced by naval transports), declares war on me and causes sufficient mayhem with his pincer attack into my north and south flanks that I quit the game with riflemen, and hope, both well out of sight. It is not my finest hour but it has been a fun game to again test MPPs replacing MAs and to watch Xerxes do Xerxes things.

Summary: Xerxes turns into his usual Deep Blue-esque, five-headed AI god mode as he approaches the Industrial era and goes from pretty much 7th ranked Civ out of 10 to being well on course for a domination victory following his perfect timing to Cavalry rush the Byzantines and myself. I've played a lot of Civ 3 in recent years and I never see any other AI Civ go from middle of the pack (or in this case, bottom half of the table) to threatening for the win with anything like the consistency Xerxes manages. The other AIs (excepting maybe China) lose their heads in the Industrial era, go for stupid wars of attrition and let the human player catch up and overtake. Whereas Xerxes actually consistently close the gap to the other Civs (human and AI) in the late game from a weaker powerbase and, can often leapfrog these rival Civs if circumstances can permit it. It was almost a pleasure to be spanked by his pincer movement in this game. It didn't feel like losing to the goofy Civ3 AI and I can only applaud my AI overlord.

Anyway, this is a longwinded way of saying if you want a fun late game AI challenge in your customised games then replicating the Xerxes's traits and 'build often' attributes (as laid out by Quintillus) is definitely worth a try.
 
Last edited:
Top Bottom