Return to Civ 3

[Divaythsarmour]

I usually play Civ 4 BTS, Monarch with a preference for peaceful strategy. But I spent this last week at a place up in New Hampshire where the computer has a crappy Celeron with 256 Ram. So I spent the whole week happily relearning Civ 3 (Gold editon). Here's a brief synopisis of what I learned:

1) Peaceful strategy doesn't work even at easiest level. I was able to rex out to a decent size landmass before being blocked off, but wasn't lucky enough to get oil. On one occasion I did get oil, but then the source dried up. It's been my experience that without oil, you can't survive the modern age. But perhaps it would still be worth putting to a real test. It would take longer than a week.

2) Rome's Leagonarys great UU for early momentum toward conquest/domination. Having given up on the peaceful strategy, I decided to try the other way. I remembered that the legionarys where very good. I expanded out to 5 cities, building barracks and granaries. Once I secured iron, I started building nothing but legionarys. It wasn't long before I had another five cities and two great leaders. I was tempted to use one to build the pyramids but I built another army instead. I figured I would just take the pyramids later.

3) Hannibal's spearmen are best early defensive UU. I always thought that hoplites were good. But those spearmen of Carthage (sorry I can't remember their real names) were very tough defenders. On one occasion I attacked one out on a hill top with a single leagonary. Obviously the leagonary died. so I hit him with another and he died. So out of frustration, I tried one of my early leagonary armies and lost it, all to the same single Carthaginian spearman. Now that's tough.

4) The AI in Civ 3 is very sneaky and cunning. For all the hype about how great the AI is in BTS (and it probably smart on some level) there certainly was no room for mistakes with the AI in civ 3. Perhaps it's the elegant simplicity of the game design that makes the AI seem so good? Maybe it's because the AI in Civ 3 does things that a human wouldn't necessarily do? For instance, it would send units to steal workers and harrass my cities, while I'm bringing overwhelming force to bear on it's cities. A human would most likely pull all units in to either defend the cities or harrass the attackers. Ultimately, the effect psychologically (and in some cases practically) would slow down my progress substantially. I was impresed.


I eventually managed a domination victory playing as the Romans on Regent level. The game took over 7 hours. It felt a little bit tedious toward the end.

 

[Snarkhunter]

I eventually managed a domination victory playing as the Romans on Regent level. The game took over 7 hours. It felt a little bit tedious toward the end.

What size map, out of curiosity?

kk

 

[Hasdrubal Barca]

Civ3 is far better then Civ4

 

[Jokeslayer]

1) Peaceful strategy doesn't work even at easiest level.

I think this would depend what you mean by "peaceful strategy". Somewhere in the Stories and Tales section is a game where someone won without any military. It's certainly possible to win without launching any wars or conquering foreign cities, although it's probably more difficult than a normal mix of war and other stuff. VC is also a factor. 20K would be far easier than, say, domination.

 

[Divaythsarmour]

I think this would depend what you mean by "peaceful strategy".

That's a very good point.

My preferred strategy would be to expand until it's no longer possible. Then begin building wonders, the economy and the military for defensive purposes. I would want to reach techs for new wonders ahead of the other civs and have plenty of money, technology and updated military units. If someone attacks, I can obviously defend myself. I just wouldn't go on the offensive.

If the residents of a particular neighboring city wanted to overthrow their governors and come over to my side, then I would welcome them. So I guess a culture war would be OK.

It's not like I'm some sort of "peacenick" or anything. I can enjoy conquering the world as much as anyone else. I just think it would add immensely to the attraction of the game if it were possible to adjust the scoring or add more victory/game conditions to make a "peaceful strategy" a possible style of play.

I think they've accomplished it to some degree in Civ 4. I've been able to do it (more or less) up to the level of Noble (where if the scales were equal, might be between regent and monarch on civ 3). I doubt if those scales are equal. According to the manual, the Noble level in Civ 4 is where the settings for human and AI are equal costs, penalties etc. It's also the defualt setting for the AI. But I've read members of this website debate that point, so I'm not really sure. I would be curious to know where that "equal" setting is in Civ 3?

Sorry, I can see that I'm al over the place in this reply.

 

[Jokeslayer]

Regent is supposed to be the equal setting. Not sure if everything is equal, but it's not completely a blank slate. Builds cost the same though.

If you don't build any/very much military, I can see where you might run into early wars from more aggressive neighbours, but it sounds like that isn't your real problem. It's the resources that are a big issue. I'm not sure when I last found myself lacking a critical resource and not being able to take it, but you should be able to trade for it.

 

[Divaythsarmour]

What size map, out of curiosity?

kk

It was the smallest map. I chose one of those more dense looking pangea maps with the oldest earth, temperate climate, and temperate forest. It ended up being this more or less trapezoidal continent with a couple of good size islands 6 or 7 tiles from the mainland.

 

[PaperBeetle]

There are essentially six VCs in Civ3. Three of them are fundamentally military in nature, so no surprise that you can't win that way by rexing and then turtling (except 100k cultural, which would be possible but obtuse). The three scientific VCs are 20k, Space and Diplo, and all of them are eminently possible from a turtled position. After all, people have won them all with just one city.
Of course, choosing not to expand through conquest pretty much equates to letting your rivals be the same size as you throughout the game, and so you might expect them to be keeping up with you in tech and productive capacity too (I am assuming that as a Civ4 player you aren't leaning on the micro, which is how in Civ3 you could get better results than the AI from the same amount of land / population). Similarly, you would be offering an advantage over yourself to any civ that did go on a successful early conquest and ended up with more than its fair share of land or significant wonders.
My counter-question is: if in Civ4 it is practical to stop expanding when the rex phase is over, but still expect a better chance at victory than the AI, does that mean the Civ4 AI is particularly worse than the player at some important part of empire management?

 

[MAS]

1) Peaceful strategy doesn't work even at easiest level. I was able to rex out to a decent size landmass before being blocked off, but wasn't lucky enough to get oil. On one occasion I did get oil, but then the source dried up. It's been my experience that without oil, you can't survive the modern age. But perhaps it would still be worth putting to a real test. It would take longer than a week.

You can trade for oil.
I'm sure its possible to remain completely peaceful (safe for a couple of defensive wars) and win the space race or something, even at high difficulty levels. Its a different type of game though.

2) Rome's Leagonarys great UU for early momentum toward conquest/domination. Having given up on the peaceful strategy, I decided to try the other way. I remembered that the legionarys where very good. I expanded out to 5 cities, building barracks and granaries. Once I secured iron, I started building nothing but legionarys. It wasn't long before I had another five cities and two great leaders. I was tempted to use one to build the pyramids but I built another army instead. I figured I would just take the pyramids later.

This UU gets +1 defense rating. Making them equal to a hoplite when it is being attacked. I don't see how they are better than a normal sword when it comes to conquest. Their ability to shrug off counter attacks more easily is marginal at best. If you play as Greek, you can also build a stack of swords and include 1 hoplite in the stack and you are about equal.

3) Hannibal's spearmen are best early defensive UU. I always thought that hoplites were good. But those spearmen of Carthage (sorry I can't remember their real names) were very tough defenders. On one occasion I attacked one out on a hill top with a single leagonary. Obviously the leagonary died. so I hit him with another and he died. So out of frustration, I tried one of my early leagonary armies and lost it, all to the same single Carthaginian spearman. Now that's tough.

Just a bit of random luck here, both hoplite and the Carthaginian spearman have 3 def rating, but the Carthaginian spearman has one extra att rating at the cost of 10 extra shields. They can attack like an archer, but they are equal to both a hoplite and a Leagonary when being attacked.

Really, its all numbers and chances, equal numbers means equal chances,
a defense value of 3 means a defense value of 3! Name and graphics don't matter.

4) The AI in Civ 3 is very sneaky and cunning. For all the hype about how great the AI is in BTS (and it probably smart on some level) there certainly was no room for mistakes with the AI in civ 3. Perhaps it's the elegant simplicity of the game design that makes the AI seem so good? Maybe it's because the AI in Civ 3 does things that a human wouldn't necessarily do? For instance, it would send units to steal workers and harrass my cities, while I'm bringing overwhelming force to bear on it's cities. A human would most likely pull all units in to either defend the cities or harrass the attackers. Ultimately, the effect psychologically (and in some cases practically) would slow down my progress substantially. I was impresed.

Once you get the hang of it, this will no longer happen to you.
If the AI does something the human doesn't it is because the human is smarter than the AI. And this AI behavior can be exploited. Just leave some worker unprotected and the AI will move its units out of its cities, allowing you to take the city more easily.

 

[Divaythsarmour]

My counter-question is: if in Civ4 it is practical to stop expanding when the rex phase is over, but still expect a better chance at victory than the AI, does that mean the Civ4 AI is particularly worse than the player at some important part of empire management?

I wouldn't want to say that it's practical to stop expanding when the rex phase is over in Civ 4. My guess would be that hardly anyone would play that way. It's just a strategy that I've been experimenting with lately.

If you do stop expanding, many of the same problems that you would expect in Civ 3 occure in Civ 4 as well. One or more AI civs get a lot more land (through conquest) and eventually out-tech you at an incredible rate.

I would say there are more options in Civ 4 that give you as a human player a chance to keep up. The best one IMO is the super specialist economy. It was initially an approach championed by a member on this site named Obsolete. He's better at it than I will ever be, but I understand some of how he leverages certain game mechanics. For instance:

There are better benifits in Civ 4 for an industrious leader than there are in Civ 3, 50% fewer hammers (same as a shield in C3) required for wonder production. If I remember correctly in Civ 3 an industrious leader only gets one additional shield per city. I could be wrong about that.

There are also huge benifits for various resources, i.e. stone, marble etc..toward the production of wonders. So in the case of the pyramids, if you have stone it requires -50% hammers. Anyone can get that benifit, but if you're industrious you're stacking the two benifits.

There are 5 different sets of civics that you can choose from that have a variety of effects on aspects of your system of government. There are two very good civics for wonder building. One is Organized religion which allows you a +25% production on all buildings and beauraucracy which gives your capitol city a 50% increase in production of everything.

In the case of the two civics above, you can run them simultaneously. So one could leverage all of those as well as other means toward a very good production economy.

All of the wonders give you a variety of individual benefits + points for culture as in C3 + points toward "great people", who are very much like the hero's that get created in C3 when an elite unit does something amazing on the battlefield. There are many ways of using those great people, one of which is to make them "super specialists."

What super specialists do is they give a certain per turn benifit to a city in the form of hammers toward production, gold, food and culture. I don't have them all memorized so I won't bother giving an example. But one could view them as being the equivalent of adding another nicely developed tile into the city plus a culture benifit as well as more points toward further great persons.

One huge catch is that it always requires another 50% GPP (great person points) to create the next GP. But I've seen in some of the posts by Obsolete and others, where they've been able to create as many at 20 or more great people per game by leveraging variables. That is one really huge counter-balance to the AI.

But it's still not enough counter balance at higher levels of the game to pull off what I'm calling a "peaceful strategy." Players like Obsolete will use the super specialist economy to get to a certain tipping point in the game where they have a substantial military advantage. Then they take that advantage and conquer the world with it. He does it at the highest levels. I can barely pull off what he does at monarch level.

So it's not that the AI in Civ 4 is worse. In fact there's no possible comparison. Perhaps it's the difference in the complexity of the game that causes it to feel different. And as you know, feelings aren's facts. Feelings change. Facts don't.

 

[Divaythsarmour]

You can trade for oil.
I'm sure its possible to remain completely peaceful (safe for a couple of defensive wars) and win the space race or something, even at high difficulty levels. Its a different type of game though.

This UU gets +1 defense rating. Making them equal to a hoplite when it is being attacked. I don't see how they are better than a normal sword when it comes to conquest. Their ability to shrug off counter attacks more easily is marginal at best. If you play as Greek, you can also build a stack of swords and include 1 hoplite in the stack and you are about equal.

Just a bit of random luck here, both hoplite and the Carthaginian spearman have 3 def rating, but the Carthaginian spearman has one extra att rating at the cost of 10 extra shields. They can attack like an archer, but they are equal to both a hoplite and a Leagonary when being attacked.

Really, its all numbers and chances, equal numbers means equal chances,
a defense value of 3 means a defense value of 3! Name and graphics don't matter.

Once you get the hang of it, this will no longer happen to you.
If the AI does something the human doesn't it is because the human is smarter than the AI. And this AI behavior can be exploited. Just leave some worker unprotected and the AI will move its units out of its cities, allowing you to take the city more easily.

Mas,
Your points are all concise and very well taken. Thanks.

 

[PaperBeetle]

Post #10

Actually, I figured you would say something like that... I was just testing really. I was aware that the Civ4 domestic economy is sufficiently complex that the AI could never formulate strategies around it as coherently as a human player. This is also true of the Civ3 economy, of course, but to a much lesser extent. OTOH, it is reassuring to know that conquest is still often advantageous in Civ4. I could never get my head around the defender-has-an-answer-to-every-attacker combat model in Civ4, so I never could make a decent conquest, which might explain why I found it frustratingly difficult to actually win.

 

[Divaythsarmour]

I could never get my head around the defender-has-an-answer-to-every-attacker combat model in Civ4, so I never could make a decent conquest, which might explain why I found it frustratingly difficult to actually win.

You must be referring to how the best defender in a stack automatically responds to each attacker. It clearly gives the defender the tactical advantage.

It's funny, because when I was playing Civ 3, I did notice that each city had the same type units defending a city. A possible difference in those game mechanics never occured to me.

Are there any general rules in Civ 3 for which unit in a stack responds to a given attacker?

One thing I noticed (and liked very much) was how if you had an artillery piece or an archer in a stack of city defenders, it would fire once at each attacker, weakening it just a little. That can really make a difference in the outcome.

In Civ 4, the trick to succesful conquest is to vary the type of units in your offensive stacks (mostly to survive counter-offensives). And bring plenty of bombardment to first weaken the city defenses, and then weaken the defenders themselves, before committing the assault units.

In the early game, before bombarding weapons, the city defenses aren't as tough, but you still have to factor in for heavy losses.

 

[MAS]

In Civ3: The unit with the highest chance of winning will be selected to defend.

This can be annoying, if you have a vet sword stacked with a regular spear, then you would like the spear to be attacked so you can attack your next turn with a full health vet sword. But the sword has a higher overall chance of winning because it has the same def rating but an extra HP, so the vet sword will be defending and ends up damaged or even killed and then you can't attack next turn.

If all units have an equal chance of winning (same def rating and same HP) then the cheapest unit is selected. So a stack of pike and knight, the pike will defend.

Of course, if two units attack you, then the pike will defend first, lose a HP, and thus the next battle, it no longer has the highest chance, so then the knight defends.

In any event, if a unit with the defensive bombard ability is in the stack, it will fire a shot, regardless of whether it is selected to do the rest of the defending.

This mechanism causes a stack of defenders to all get a turn. Only if a defender wins a fight with full health, it will defend a second time. And winning twice in one turn means auto-promote, so this will cause that unit to gain a HP, and increase the chance it wil end up having to battle the next unit as well. (This causes many elites to end up being killed the same turn they where gained, annoyingly)

If you attack a big stack of defenders, and you fail to kill each and every one of them in that attack, then they will just heal in the inter-turn and your HP damage is undone. As every defender that loses a HP will just go to the bottom of the stack. This is why you should always try to make sure you can kill an entire stack, instead of just damage them. Otherwise, any units that where sacrificed during the battle where lost in vain.

 

[PaperBeetle]

Correct. It is also worth noting at this point that a turn comprises first your unit moves, during which you may attack, and later the AI's unit moves, during which you may get attacked. So if you attack and a kill an AI unit, and the AI then unsuccessfully counterattacks, your unit will auto-promote.

 

[MAS]

Units with blitz can become elites very easily, as long as they win 2 fights in a row you are guaranteed a promotion.

Did it never happened to you that you pop a goody hut with your first (and regular) warrior and 3 barbarians pop out, they all attack and your warrior ends up elite?

 

[The Professor]

Did it never happened to you that you pop a goody hut with your first (and regular) warrior and 3 barbarians pop out, they all attack and your warrior ends up elite?

The way my luck runs, I usually wind up with a dead warrior. However, I've seen this with tanks all time. Gotta love the blitz ability.

 

[Quintillus]

Correct. It is also worth noting at this point that a turn comprises first your unit moves, during which you may attack, and later the AI's unit moves, during which you may get attacked. So if you attack and a kill an AI unit, and the AI then unsuccessfully counterattacks, your unit will auto-promote.

Never knew that about auto-promote including the AI turn as well, though it makes sense.

Units with blitz can become elites very easily, as long as they win 2 fights in a row you are guaranteed a promotion.

Did it never happened to you that you pop a goody hut with your first (and regular) warrior and 3 barbarians pop out, they all attack and your warrior ends up elite?

I've had that happen, though always thought it was simply due to good luck (and at least one of them was...right? Three wins isn't auto-double-promote, is it?).

The requirement to attack with catapults (and the ability of catapults to win with no other attackers) always irked me about Civ4. It simply seemed way too defence-oriented without gazillions of catapults - and I consider myself a more defensive player than average. Plus not being able to destroy roads with ships or artillery. Though I probably will re-install Civ4 over the summer at some point and see how Huge maps run with 2 GB of RAM (0.5 just isn't enough for Civ4).

 

[MAS]

I've had that happen, though always thought it was simply due to good luck (and at least one of them was...right? Three wins isn't auto-double-promote, is it?).

Well, its good luck in the sense that he won all 3 fights!
Yes, the way it appears to work, is that a flag is set after a victory, and this flag is removed the next turn.

It happens less often in C3C because of a change in how barb AI works.

 

[dexters]

4) The AI in Civ 3 is very sneaky and cunning. For all the hype about how great the AI is in BTS (and it probably smart on some level) there certainly was no room for mistakes with the AI in civ 3. Perhaps it's the elegant simplicity of the game design that makes the AI seem so good? Maybe it's because the AI in Civ 3 does things that a human wouldn't necessarily do? For instance, it would send units to steal workers and harrass my cities, while I'm bringing overwhelming force to bear on it's cities. A human would most likely pull all units in to either defend the cities or harrass the attackers. Ultimately, the effect psychologically (and in some cases practically) would slow down my progress substantially. I was impressed.


I eventually managed a domination victory playing as the Romans on Regent level. The game took over 7 hours. It felt a little bit tedious toward the end.

I got into an MMO the year Civ4 released and never truly got into it and perhaps having helped test and give feedback to the developers prior to the games release took some thunder away. I Still have to get BTS but the removal of the Civ3 style trade table really really disappointment me. By the time I was brought onboard, the trading system was already there so I decided not to fight it. The Vassal system in Warlords was great. I don't know what they added in BTS, but when I get around to buying it, I'll find out.

In anycase, the AI in Civ3 is good. Many expert players have figured it out, or they have figured out that one weakness that they can hit on everytime to cause it to unravel and so that one critical flaw makes the AI broken to those players. It's not unreasonable to say it, but I think if a player doesn't know the flaw or choose not to exploit it can still enjoy a fairly robust AI and a very well taught AI who has a frew tricks of its own. The biggest achievement in Civ3 was probably the ability for the AI governors to have a non-linear build oder for their cities based on the empire's circumstances and terrain. As opposed to the fairly predictable build order of AI cities in Civ2.

I want to say I really enjoy Civ3, I have returned (or is trying to return) to the Civ community and I plan to fire up my old C3C disc.

First let me address what I consider to be the glaring weaknesses of the AI

• For my the most glaring problems with the AI are the MMP entanglements and war declaration cascades that sometimes don't make sense. I've seen an AI repeatedly declare war and make peace with another Civ, and what was causing this was the AI's MMP ally was at war with a Civ the AI obviously has decided to make peace with. This problem surfaces sometimes but its not common.
• The AI can be baited fairly easily with an 'undefended' target.
• AI garrison/target selection at wartime is not the best. This can cause it to lose momentum. It will always peak the weakest target to attack, as opposed to the most strategically significant. So it can take a city, stop, go all the way around to attack another target, meanwhile its newly conquered territory can be retaken as the AI's attacking forces have moved so far away it cannot be used by the AI to counter an advancing counterattack.
• -When the AI gets beaten down to OCC or a few cities, it really flies off the rails. The AI leaders behave as it they have an empire when infact they have nearly nothing. It would have required a programmer of two a few months to shoehorn a specific set of AI routines for AI civs that are that size, and a writer maybe a few weeks to write up alt. dialogues for the AI leaders. That to me would have enriched the Civ3 play experience the most. But alas, what we have are Civs who remain cocky and think they are equals among nations when they have no power to back them up.

That said, I am really in awe of what Soren Johnson and his team, but Soren specifically, achieved through the 2+ years Civ3 was being patched, and expansions released for it. He used to read the Civ3 strategy forums to look for AI problems and fix them. The cunnining AI was essentially built on Soren's intuition and the years of feedback he got form the community. The sad part was that we never for a 1.22f for Conquests. Several bugs remain unfixed and I sometimes wonder what fixes were in the can but ultimately not released because the whole 1.22f patch was cancelled by Atari cutting funding to further support.

In anycase Here are some AI changes/additions added from player behavior/feedback that I noticed through the years.

• ROP Rape
• Sneak Pillage (AI will sneak a high defense unit usually far away from where your troops are massed past your border cities to pillage resources deep in your territory and hit targets of opportunity such as an undefended city or workers) ~added with the PTW expansion iirc.
• AI SoD behavior changes (AI will split up its SoD) ~PTW era change
• AI tech choices changes (several tweaks implemented throughout several patches)
• AI navy behavior changes (AI will defend its transforts with naval SoD) ~C3C era change he WW2 conquest scenario
• AI air power changes (AI will build more air units and use them) ~C3C era change to accompany the WW2 conquest scenario

There are many more. I wish someone can add to this list.