A Better AI.

If we have more defending strength than the enemy has attacking strength, then pillage. (longbows in a city on a hill but no counter units)


However I would still question whether it's better to pillage than to just go away and hope the enemy lets their guard down a bit... it could be based on whether it's war for profit vs war because i hate you....


In that situation what about pillage only strategic resources like Iron and Bronze, leave an unit there(for example a longbow, normally its in hills anyway) for slow down the enemy? And maybe pillage any towns in the way of the retreat if its relatively safe?

Or somethnig like that hehe. If you wont take the city anyway, so pillage it a bit.
 
In that situation what about pillage only strategic resources like Iron and Bronze, leave an unit there(for example a longbow, normally its in hills anyway) for slow down the enemy? And maybe pillage any towns in the way of the retreat if its relatively safe?

Yeah, unless you are going to take the City soon, I'd say it's always worth it to Pillage Towns and Strategic Resources. Farms are never worth the Trouble I'd think, but you can make some good money off of burning down towns, and knocking out Iron and Horse in the early game, will make you win the war eventually. Except in the Capitol, archers just can't stand up to CR Axes and swords alone.

Something I was thinking about with War mongers and the Possibility of them getting Domination wins. Is there anyway to scale the AIs chances for war as it aquires more vassals? Monty is insnsane, and always has a 2% war. If an AI gets a vassal, would it be at all possible to code a 1% war increase per vassal aquired. This would give a civ like Mongolia a +5% of picking a target if he aquires 5 vassals, and with is normal 1-2% range, it could push the Warmongers to Vassal Juggernaught, or go down trying.
 
I draft Musketters all over the place and send them to Rheims, so Rheims has 20 Musketters.

Finally, after bringing in a bit more Cavalry and Grenadiers and bombarding my city defences to zero, the AI attacks Rheims and takes it, taking rather heavy casualities.

I wonder what was worse damage to you: taking the city quicky when it had 3 musketeers in it, or taking it after you drafted 20 more and sent them there to die! ;)

Although I do think the AI is not using its large stacks effectively yet, waiting for more strength is probably good under many circumstances.
 
Even though some imperfections have been reported about the 12-21 build, using it, I just lost a Monarch/Epic/Continents game to Wang Kong -- Cultural Victory in 1871. :cry:

W.K. was on another continent with only Brennus. I didn't meet him until several turns before his victory. He wasn't building spaceship parts, so I didn't pay any attention to him. However, a couple turns before the end of the game, I checked the victory screen and I noticed I was doomed!

I know Blake was initially dubious about programming the AI for Cultural Victory, but believe me, the AI can win that way. I was basing my game on getting domination before anybody built the spaceship. That's not good enough now with the Better AI. :goodjob:
 
Not sure if this is completely germain to this discussion, but the way I wage war in Civ4 is to plan, plan, plan, and bring overwhelming force. I start a war with the intention of completely wiping out the enemy civ; rarely will i sue for peace, and only if my plans have gone horribly wrong or I've made a mistake. And even then, a peace treaty is only a tool that enables me to regroup for 10 turns and go at them again.

My games generally follow this pattern: Build 4, sometimes 5 cities and prepare for the Barbarian invasion. Fight off the barbs while building up my nation. I have a flowchart that I follow for tech research, which roughly gets me bronze working/alphabet/civil service as quickly as possible. Once I get Civil Service, (and can sustain a large number of cities), I move towards my first war - by this time I've identified the other civs that I will take out, and in what order. This depends on the map & etc of course, but I'm usually being squeezed, cut off, or need to expand into another civs' territory. Of course If I have room to expand into un-occupied land I will do that first, and pause after building a every few cities so that my economy can recover. But at some point in every game I find that I need to take out other civs in order to stay competitive.

I'll target the weakest civ that I have a common border with first. I'll take out other civs from the weakest to the strongest. My thinking is: I myself am weaker earlier in the game, my combat units don't have a great deal of experience yet, and will gain that through fighting. And by the time I fight the strongest civ, I will be at my strongest.

I'll build several armies and assign each of them 2 to 3 enemy cities as targets. My armies I make up of 4-5 siege units, 15-20 melee/gunpowder units, and assign 1 or 2 stacks of 4-5 cavalry to each. The cavalry does not stack with the army; they perform the traditional cavalry functions of screening, reconaissance, blocking, etc for my armies. Civ4 portrays cavalry very well and they are indispensible. I plan my invasion so that the armies converge towards the final 1 or 2 cities, so that I can merge them - because by the time they have taken several cities they will be weakened, and usually the enemy makes a 'last stand' with all their remaining units in the last remaining city. Usually the enemy will sally forth with an army to attack my units, or go for a border city. i harrass them with cavalry, and am always glad to receive their attack, because they just whittle themselves down and can then be wiped out or ignored at my leisure.

I choose the shortest path to each target and do not stop to pillage, or to attack enemy units that are outside their cities. Never lose your focus and chase targets of opportunity, you'll drag yourself to a halt. I do not pillage - for a couple of reasons - it slows down the invasion which is a game-killer. I do not want a long war, I want to quickly take each city in turn and end the war as soon as possible. Pillaging for a little gold is a waste of precious time. I never fight a war unless I plan on taking out the whole civ. So, if I pillage, i am really just destroying improvements that will be mine in a few turns. In other words, I'm pillaging my own improvements!

I move my army next to a city, with cavalry on the flanks, and far enough away from the city so that the enemy won't send out tropps to attack them. The cavalry gives me eyes and protects the army from attacks to the flanks. I hit the enemy city with a barrage of all the siege units, and then attack. rarely will I bombard a city twice, the trade-off in lost time is generally not worth the advantage of reducing defenses. I'll bombard, then attack on the same turn, take my medicine, and take the city. I'll only raze a city if it's very poorly placed. If I find I've run up against a heavily-defended city, I may set in and bombard it while I move another army over to help. very rarely will i use my cavalry to help take a city - I need them for other roles, and they are too precious.

After I've taken another civ out, I'll regroup, build my armies back up, allow my financial situation to catch up, and hit the next one. the thing I hate the most is when a civ has built a city I can't reach, heh. But it's a small matter, after I've destroyed their main civilization they are eliminated as an effective competitor.

Anyway, the point of all this is to show one way of waging war - and what if the AI fought wars in this way? You'd have your hands full I think, trying to stop such a 'blitzkrieg' attack by multiple large armies. I've not played a game where I didn't wonder, "What the heck is the AI thinking?", because it doesn't react intelligently and is too predictable. I'd like to see several AI types, some that are smart warmongers, some that are so-so, some that try to win via diplomacy, culture, science, etc. and have new civs be assigned an AI randomly at the start of the game. Furthermore, allow the leaders for a nation to be changed, and bring with them a possible AI-type change. Now wouldn't it be cool if your peaceful neighbor suffered a coup and the new leader was a violent warmonger?

I think the AI is where the most improvement could be made, and it's a shame it's not the target of more modders. So much could be done with it that would keep Civ 4 fun and exciting for years to come.
 
Pillaging should probably be a function of AI personality.

If an aggressive leader / warmonger leader can't take a city... it should raze and burn everything within reach. OTOH, a peaceful leader probably shouldn't pillage (other then strategic / towns).

At least, that's my initial thought on pillaging. Or make it random, where the aggressive civs are more likely to pillage as they move, while other civs are less likely.

Personally, I never pillage anything other then strategic resources. And only if they're on the way to the city that I'm assaulting. But if I can't take the city, I'm not adverse to pillaging as I retreat. Unless pausing to pillage will get my troops killed.

...

I'd like to see the AI do more harassing attacks of invader stacks within their borders. Suicidal collateral damage attacks mostly (or attacks by planes). No sense sending single units of other types.
 
I think at least some potential for pillaging is necessary for the AIs. The risk of having your lands pillaged forces players to come out and fight on occasion rather than sitting back in defensive terrain/dug in in cities. In 'stock' Civ4, I never even bother trying to defend cities per se because the AI pillages the hell out of EVERYTHING. You MUST go out and fight the AIs in the field or else your improvements are toast.

This is good and bad IMO. Its bad when the AI is obviously going to take the city it is pillaging, but its very good about drawing a response from the defender. In many cases it can forces the defender to actually have to attack units at a disadvantage just to keep them from pillaging.

Another thing that would interesting is if the AI would actually 'blockade' coastal cities. Many time those cities are heavily dependent on coastal tiles for food. If the AI just parked some ships off the coast of some of these cities in a 'net' it could do some SERIOUS economic harm in many situations, especially later in the game when there is less likely to be Slavery in use or alternate in-land food sources.

Looking forward to the next build. Thanks again for your effort!
 
Even though some imperfections have been reported about the 12-21 build, using it, I just lost a Monarch/Epic/Continents game to Wang Kong -- Cultural Victory in 1871. :cry:

W.K. was on another continent with only Brennus. I didn't meet him until several turns before his victory. He wasn't building spaceship parts, so I didn't pay any attention to him. However, a couple turns before the end of the game, I checked the victory screen and I noticed I was doomed!

I know Blake was initially dubious about programming the AI for Cultural Victory, but believe me, the AI can win that way. I was basing my game on getting domination before anybody built the spaceship. That's not good enough now with the Better AI. :goodjob:
No, it just means you have to make a minimal effort to stay on top of things. I daresay you could have simply sent a caravel over to make sure you have met all the AIs, way before the 1800's, surely. Since you were already building for a domination win, it should have been fairly straightforward to invade him way earlier.

Wodan
 
I see the pillaging issue like this:

The AI should round up all it's units on 'city attack' mission targeting the given city and compare it to the sum of known defender unit strenghts in the city area (not just within the city to make luring the AI into a trap more difficult).

If it seems the city could be taken the AI always should go for the city, without pillaging. That's sure.

If it seems the city cannot be taken but defender is weak to attack the intruders then go pillage with the city attack stacks (not with individual units it's too risky, maybe splitting up the huge attack stacks to smaller well-balanced ones but it might be too difficult to implement) until:
a.) balance of attacker/defender changes so now it seems the city could be taken. Then regroup if necessary and attack the city
b.) defender becomes dangerous to the assault stacks (got relieve forces or anything). Then run away and regroup in friendly territory. Maybe consider changing to defensive war.
c.) nothing else to pillage. Choose another city for pillaging and change the attack focus to that city too. Regrouping might be necessary to keep stacks together while pillaging.

To the healing issue:
Attacker AI should consider pulling out damaged units to friendly territory to heal from a stack instead of the entire stack waiting for healing. Later these units should rejoin the stack.

To the builder AI defense issue:
Maybe it would be better to make builder AI's not to train fewer troops but defensive kind of troops instead. Archers, city garrison riflemen etc. An army adept at defense but hardly usable at offense well. Very few siege engines just for anti-stack.
Builder AI's shouldn't be just prey. They should defend what they build after all. But you would know you don't have to fear them attacking you with dozens of trebs :)
 
If you picked 20 different real Civ4 players, identified their playstyles, and coded it into the AI, you'd really have something.

After all, the best AI would be random and unpredictable, including making mistakes, but have an overall 'style' or major goal in mind, and emulate a human as closely as possible.
 
However I would still question whether it's better to pillage than to just go away and hope the enemy lets their guard down a bit...

Do you mean with the enemy the player or an other AI?

Because i think a human player will never let his guard down if an evil stack of death is wandering through the land. Furthermore, if I know that the AI restrains from pillaging, I will focus on the defence of my cities.

I understand, that the AI is hampered in wars among each other by pillage. But it is fun and gives against a human player the sense of urgency. In a Vanilla game lately, the Egyptians declared an early War on me, but couldn't conquer with chariots my cities. At the same time I hadn't enough Spearman to take out the chariots, so they burned down my villages and farms.

Pillaging should probably be a function of AI personality.

If an aggressive leader / warmonger leader can't take a city... it should raze and burn everything within reach. OTOH, a peaceful leader probably shouldn't pillage (other then strategic / towns).

I second that. For Ghandi it should be unlikely, for peaceful leaders it should be somewhat random and the mongols should burn anything they can.
 
No, it just means you have to make a minimal effort to stay on top of things. I daresay you could have simply sent a caravel over to make sure you have met all the AIs, way before the 1800's, surely.

Surely, yes. The point I didn't quite make is that the clock for Cultural Victory can tick faster than the one for Spaceship. But, I suppose any human that's won a C.V. already knows that.
 
I second that. For Ghandi it should be unlikely, for peaceful leaders it should be somewhat random and the mongols should burn anything they can.

That would hurt the Mongols chances of winning if they pillaged areas of cities they can capture.
Actually 1 pillager is waste if the city could be taken.

I know in reality Mongols fought much like pillaging and razing and massmurdering and then subjugate other peoples but in Civ game it's just suicidal tactic to not take a city with all it's improvements intact if you can

btw Mongols were steppe people not really building cities but rather nomadic AFAIK
 
Hi guys,

Here's another. Compare what tiles are being worked (forest + grass mine) to what it could be working (farm + plain mine on river). Same food count and same hammer count, but we're missing out on 2 commerce....

Wodan
 

Attachments

  • hammer emphasis0000.JPG
    hammer emphasis0000.JPG
    161.4 KB · Views: 123
Furthermore, if I know that the AI restrains from pillaging, I will focus on the defence of my cities.

Yep, that sums it up nicely. The player should never be feeling secure that the enemy wont wreck his improvements. The AIs should pillage to keep players 'honest' and trying to clear their land of enemies, not just digging in on the most defensible terrain and forcing the enemy to attack them for results.

Pillaging draws a response and is a valid tactic and should be encouraged for the AI in almost any situation where it's not likely to take the nearby city in relatively short order.
 
Pillage or not pillage is a tough question for the AI. (Pity poor Blake!)

And as others have said. If the AI doesn't pillage, then the human doesn't have to leave the safety of his cities to go attack in the field. The AI (or at least some of the AIs) need to be scary sociopath monsters.

When a human goes to war, we have numerous, sometimes conflicting goals:

a) weaken (equal or greater power then me) - do as much damage as possible without getting caught, mass pillaging and city razing are the primary goal - this also includes "war through other means" such as gifting luxury resources, then removing them at a later date - spy missions - or bribing other civs to attack my target - camping ships on top of food resources - or just sitting on a tile to prevent the city from working it

b) conquer territory and wipe them out (equal power or a really minor power) - take cities / land away from the enemy - not much pillaging here other then strategic - I don't plan on stopping until I run out of troops or the enemy cries uncle (and maybe not even then) - or I'll stop out after taking all cities near my lands - after I'm done, the enemy will *never* be a factor in global politics again

c) get someone to capitulate (lesser power that I'm friendly with otherwise) - take a city or two and sue for capitulation

d) stop something from occurring - such as making a deep strike into enemy territory to take a city that is trying to build a wonder

I'd love to see the AI learn "war through other means" where they gift resources to an enemy, then pull the rug out later.
 
Yep, that sums it up nicely. The player should never be feeling secure that the enemy wont wreck his improvements. The AIs should pillage to keep players 'honest' and trying to clear their land of enemies, not just digging in on the most defensible terrain and forcing the enemy to attack them for results.

The invaders should also try to move in better defendable tiles if possible.
 
Good pillage strategy involves using few units (no stacks) and going for valuable resources.

So the AI personalities that are made pillage-inclined should create a few mounted (chariot, horse archer, knights, cavalry) units and send them in as singles or duos to pillage cottages and strategic resources of any kind. Cottages = money for invader, losing any sort of happiness, health or strategic resource usually hurts the defender quite hard.

Any chance of a new release with the bugfixes soon?
 
Come to think of it, the same pillage strategy (excepting cottages) could be used for spies. Would be wonderful to see the AI use spies well, that would make the end game MUCH harder.
 
Back
Top Bottom