Defending your opponent to death....

[Velociryx]

Here's a little something to mull over:

Attacking is almost a necessity in this game. The AI ruthlessly attempts to hem you in, stick you with a bunch of marginal terrain, and then mass troops to finish you off. To counter this, we humans REX, and hit them hard, before they have a chance to set up.

But perhaps hitting them hard is not always the best (most efficient) approach to take.

Now, ultimately, the goal of warfare in Civ3 is two-fold: First, to gain GL's, and second, to capture enemy cities, so "defending your opponent to death" will only work to a point....in the end, you will need to take to the field actively and go take those cities.

When to use the strategy
*Note - this would never in a million years work against a human player and is meant only for the AI

Some of this has already been discussed to a degree. The best time to begin using this strategy would be at the outset of a war. Goad your opponent to declare against you and station your troops inside your borders. Ideally, you will have a road network constructed through some rough terrain (hills/mountains) that lies directly in your enemy's path, preferably with fortresses built in at least a few of those areas, and fortifiy your troops there (this then, should spell out pretty clearly when/where to make use of this strategy, as such a situation will not be present in every game).

Immediate advantages: your best defenders are invariably cheaper than your enemy's best attackers, and yet, the combined multiplicative effects of the terrain/fortresses will "boost" the defensive value of your unit to at least on par with the attackers "attack value." This means more bang for your buck. Your cheap unit now has at least as much stopping power as the attacker, who spent more shields than you. Good way to even out those production advantages the AI gets.

Holding the line: Having one fortified troop per tile will draw the AI's attention and fire, but if/when the AI masses for an assault, that lone defender may well be overwhelmed, so it is in your best interest to have some backups...2-3 per tile works best (adjusted upwards depending on the size of the approaching force), which implies that this strategy is best used along choke points, to spare you from having to build silly amounts of units, and further, those forces should be backed up by fast moving attackers to polish off any weakened units (those that manage to kill off some of your defenders, or the AI's own fast attackers who retreat before being killed).

Setting yourself up thusly at the outset of the war (and if you goad your AI neighbor into the fight, you can more or less control the timeing OF the war, and thus give yourself the time needed to arrange your defenses) will see you absolutely CRUSH the AI's attack wave in short order, whereupon you can stroll casually into his turf and repeat the process.

Smother and siege: When you defeat the AI's incoming attack force, that's the time to move out, but you don't want to make straight for the first city. Oh sure, you'll be sending a strike force there, but remember, Civ3 is not a classic form war game, so you don't have to trouble yourself with such things as supply lines....meaning that your forces can penetrate as deeply into the enemy's territory as you wish.

Specifically, what you're looking for is terrain similar to that found in your homeland. Mountains or hills (nevermind the roads, you're now in your opponent's territory, and they won't do you any good, however, if you're playing an industrious civ, it might pay to bring along a couple workers to help construct fortresses on the fly). Your next move is to advance your defensive line to a point that lies beyond the first city you're targeting to hit, posting them in high terrain (hills/mountains), and build a fort for them if possible/applicable.

This gives you the same advantages it gave you when you were defending your home turf, plus a few. First, the AI will almost always make for your defensive units, rather than attempting to save the city that your strike team is moving toward. Second, it gives you expanded LOS. Thanks to controlling the high ground, you can see your enemy approaching, see what he's bringing, and from which direction they're approaching. This facilitates your planning, and you can adjust your at home troop production accordingly.

Finally, when you are ready to assault the city, your strike force moves up to the (now isolated and cut off) enemy gates, and blows them down. This gives you a place to pull back wounded troops to heal, and gives you another place to build reinforcements from (in addition to likely giving you a few new workers to help out with terrain improvements and such).

In the ancient era, no one can do this better than the Greek. Using their Hoplites as "defensive attackers," they can declare war on their enemies and invade, parking themselves on critical resources, hills, and mountains, and laughing all the way to the bank as the hapless AI sends out its archers (the peer of the Hoplite) to attempt to kill off a base 3 defender sitting on favorable terrain. Essentially, this gives the Greek "swordsmen" long before anybody else has access to them.

-=Vel=-

[Velociryx]

Final notes:
It should be noted that this is NOT a fast strategy. If speed is required, this is not the way to go. However, if you're playing for maximum efficiency and don't mind sacrificing speed, this will serve you well.

-=Vel=-

[Arrian]

I dunno... I'd rather just whack 'em with a horde of horsemen and be done with it. The AI's counterattack will be a once-and-out thing, taking place near a city of theirs that I've taken away.

The closest I've been to this strat was actually in a game I started last night. I was in the middle of crushing India when the Zulu attacked me. They attacked a former Chinese city, built on a hill. They brought a swordsman. Later, impis starting showing up, plus more swords. I had an elite sword (who had already generated a GL), a pikeman, a catapult, and a couple of horsies in the city. The city had walls & a barracks.

I didn't have troops to spare, so I fought on the defensive. The 2 horsemen became Samurai pretty quickly, and I was buying extra pikes as cashflow permitted. I can recall suffering exactly one casualty during this stalemate time: an elite samurai which left the line to hunt down an impi (impi 5, esam 0). I generated several elite samurai as troops began trickeling in from the Indian front, and eventually moved to the offensive, taking two cities. But again, I gotta tell ya, I'd rather I'd had the firepower to just wipe 'em out. But after killing China, Persia and India, I kinda ran out of gas.

-Arrian

[King of Rasslin]

The AI is so dumb when you make a little fort in his territory. The strat will work, but it won't win you any cities. I tried a variation with this as Japan. I would make a ton of samurai to attack the French. I would wait on the border as they hopelessly attacked my fortified samurai. Then I would move in and go for the kill. Unfortunately, the French make a lot more musketeers than the other AIs made musketmen. They had around 4 musketeers in each city. I captured a single city and lost over a dozen samurai. War weariness set in and I canceled the attack. I think this strat would be great in the ancient ages, but its usefulness will gradually decline. The blunt but effective city grabbing with cavalry is more effective after they are available. You don't need this strat when infantry are available because they are so tough anyway. Tanks and mech infantry can crush cities without this.

My failure was waiting too long. I just had to wait until I had my precious samurai. The result was an offensive failure, even if I killed more of the French than I lost. I would try it with spearmen, but very soon. I would use hoplites but I don't like the early golden age.

[Velociryx]

Quite so, Brother Arrian....it's usually my preference to deliver a knockout punch as well, but magnifying the power of cheap defenders has its uses (your latest game being a classic example).

Another would be, if you're neck deep in a war with an AI civ, and another decides to declare against you. Your army is out of position to deal with the threat, but you've likely got city garrisons ready and waiting. Would be easy enough to shuffle those onto favorable terrain at some forward point to meet your rival in the field, and replace the garrisons while they were on the march.

Also, once you have locked down your continent, using the defend to death strategy is a good way to help drag a war out, if you're looking to milk a few more GL opportunities out of the equation. At that point, you already know you can whack your enemy, but you may not want to kill him outright (as it would take him out of the game, and you'd rather leave him alive and furious as a punching bag).

-=Vel=-

[Velociryx]

King, it is true that the strat by itself will not win you any cities. The point of it though, is to provide the AI a focal point that exists somewhere besides your attack force set on taking AI cities. Give them stout defenders set up in terrain favorable to you and let them wear themselves out on that, while your strike force slips in and takes the city in question.

Vel=-

[King of Rasslin]

It's just as effective to do the same thing with offencive units. Just set a few swordsmen in your border cities. Because you have roads and he doesn't, you can wait for him to come near your cities. Then butcher him with the swordsmen. I think defending the border with 2-3 units per tile is a little ridiculous. I think a dozen horsemen in a city is more effective; attack, retreat, repeat. The fort in enemy territory made a ton of war weariness in my unfortunate Japan experience. You should just get a war over with as quickly as possible. Avoiding war weariness is why fast units are so popular. I later killed the French with cavalry. 8 cities captured opposed to the samurai only taking 1.

[Arrian]

Vel,

I think the strat you descibe actually may work best when dealing with second front, when you're short on troops. Excellent in the middle ages, actually. By then you actually can afford to spend some time building forts and such. Plunk a line 'o pikemen down on some hills, back them up with a few offensive units and any catapults you may have lying around, and await the AI's longbowmen (ideal situation, clearly. The AI does seem to avoid chivalry, though, so most likely you will not face knights). Add an attacker here and there as you finish off the 1st front. By the time the bulk of your army can work its way over to the 2nd front, you should have some elite units over there, and the AI attack force should be decimated.

-Arrian

[King of Rasslin]

The strike force "slipping in" is a good idea. Although the French were aggressive with the musketeers, they left a lot in their cities too. "Slipping in" cost me a dozen samurai. I retreated with the others, upgraded to cavalry, and finished them. It might have been so easy for the cavalry because of the samurai killing so much. I guess I did use combined arms, although not at the same time.

[Velociryx]

It all depends, bro. In the early game there's no war weariness to worry about, so ending the fight quickly may or may not be in your best interest. At that point, you may find that by extending the war for a few turns will net you another GL...surely that's worth the price of admission.

As to stacking defenders along the border....note specifically that the strat calls for the presence of a natural choke point....not present in every game, when when and where you find them...yes! I'm a firm believer in shutting the AI down at choke points. Easy thing to slap a fort down and staff it with a few spearmen....simplifies the whole defensive equation.

You're right in that it IS just as easy to use swordsmen, but swordsmen are 10shields more expensive than spearmen. If you're looking at cost effectiveness, the ability to fortify a "front" with cheap units while your expensive attackers are on some other front causing trouble is valuable and cost effective. At this point, and where the strategy finds its greatest utility, is when you are in the position of needing to make every shield count.

-=Vel=-

[Velociryx]

Cross posted with ya both....and agreed, Middle Ages is a splendid opportunity for use of the strat (as would be any time you're more interested in gaining GL's from a protracted war than simply polishing an AI off).

As a "hold the line" strat for a second front, it's tough to match, you're right...cheaper troops than the attacker, and it spares having to fight in and around your cities.

-=Vel=-

EDIT: And, as the Greek, if you don't mind blowing your GA early, it's a good way to teach the Romans a thing or two...

[King of Rasslin]

Vel, where did you learn to type so quickly?

Anyway, although war weariness isn't an issue with the ancient era, I wouldn't find the strat better than grabbing the cities as quickly as possible. What I overlooked is the fact that I was trying to take a size 12 city from France. In the ancient era, cities don't give that hefty 50% defence bonus unless they have walls. And the comps don't make walls often.

It is more effective to snatch the cities as quickly as possible than go for a cost effective, long war. It is better to make the more expensive swordsman and take the cities before they make aquaducts. After that point, you are going to lose a lot of units over a single city. The middle ages would be an ideal time for the strat: before musketmen but after spearmen. It would also work on an AI unfortunate enough not to have saltpeter

But the AIs share their resources, so that isn't a likely scenario...

[Arrian]

Yeah, this is definitely not a strategy you plan on from the game start... it's something you can use if you find yourself in a specific situation. In that situation, however, it can be gold. First, don't despair of a second front (the AI fears it like the plague, but we're smarter). You might be surprised how easy it is to destroy a large attack force with relatively small numbers of troops, if the terrain helps out a bit (i.e. you have defenders covering a few attackers on hills or moutains, they march up on grassland). And like Vel said, how much is a GL worth? Well, I'd have to say 400-600 shields, but who's counting?

-Arrian

[Harovan]

I have seen the AI agreeing to very humiliating (for it) peace treaties, after their assault faltered due to well-placed defenders, even though I did not even cross their borders. A recent example was Mister B, after he made me a tough time with his Archers, I let him run against 2 vet spearmen fortified on a chokepoint of 2 mountains. He lost all of his attackers, and gave me 3 techs for peace.

The point is not (only) to take AI cities. They are corrupt anyway. The plan shall be to weaken the AI at all cost. Since the loss of all of it's attackers weakens the AI significantly and it will waste shields on new units instead of settlers, I can imagine, that this approach works. But I prefer a more "active" strategy anyway.

[King of Rasslin]

Arrian, using defensive units to get leaders rarely works in that situation. After taking 1 damage, a spearman is replaced by a healthier one in the stack. And you can't heal in enemy territory until battlefield medicine. And then you have to safely get the leader out of there and back to your land. Defending at the border doesn't work well either because the AI will focus all of its power on a single stack. This usually hurts the AI a lot but it almost always overwhelms the stack. I would say infantry would do well defending the border, but not running through enemy territory. Zulu impi would best be suited for that, but I hate their traits.

[Arrian]

No, not using defensive units to get leaders... using good defenders on good terrain to force the AI stack out in the open ground where your limited no. of attackers can waste them at will, getting promoted in the process. Having catapults makes this even easier, as you will often knock of a hit point. If the AI puts together a good stack, you can whittle away at it and still be secure behind your line of defenders, despite being nearly fully engaged on another front.

-Arrian

[King of Rasslin]

Ok, I see now. I would make the mistake of totally surronding them and losing a lot of units in the process. I think "leading" them into open land would work better, but...

The defensive strat is mainly for people that are at war in 2 fronts. I doubt you have any offensive units available to "whittle away" at the stack. Although it work work very well if you planned it from the start, this isn't really a strat you plan on doing.

[Lawrence of Arabia]

So I tried this theory out 3 times.
The first time, I did it too early and I ended up falling behind tech and city wise. It seems you need to wait until after you have finished expanding.
The second time it worked out perfectly. The AI kept hammering at the 2 fortresses guarding the chocke point.
The third time it worked for a short amount of time, but then the AI got smart, built boats and went around it.

I felt that in general, it helped. But as a couple of posters said, the amount of resources/time you spend building fortresses might have better been spent attacking the enemy. Of course, if you are short in troops, you may consider putting what little resources you have, into these defensive shields.
i just prefer rolling in with Knights and terrorizing everyone

[King of Rasslin]

The nice thing about samurai is that they can "roll in" like knights and defend like musketmen. Their movement combined with defence makes them like the mech infantry of the middle ages. I bet defending the enemy to death would work with others civs than France. They make probably 3x more musketeers than other civs make musketmen. They didn't have many culture buildings, though. In that way, the samurai succeeded in making them waste production, but...

It also slowed me down. I only had 1 luxury where I started, and my luxury meter was at 20% before I finally stopped the war. Although I could have won if I sent in enough samurai, the war was allowing the 6 other civs to jump ahead of me. War is really against production, especially if you are only gaining territory. All I gained from finally defeating France was some incense. Now I have 2 luxuries

[Cookie Monster]

I've been thinking about this one for a while now and here's what I think.

What Velociryx has outlined in his strategy above is classic defense in depth. He tries to draw the fire of the main AI assault into a controlled region and then use his offense to grind down the AI's current offensive power.

What I have observed in most of my games is if one doesn't eliminate the giant stack of AI attackers then you will lose any cities you capture from the AI unless you have massive defenders installed in captured city. By grinding down the AI in a protracted brute force defense you eliminate his potential to launch counteroffensives to your own attacks. Sure the AI may be able to quickly replace his loses if you don't take any of his cities but it will take time to assemble his forces into another monster stack. If you play your cards right and have a reserve force available you can pounce on AI cities thereby diluting his replacements.

All of this takes carefull planning, but is that not what happens in real life?

[Cookie Monster]

Sorry forgot to mention that all of this talk reminds me of the colossal struggle between Germany and the USSR in WW2. The Russians gradually learned to use there manpower to create the offensive defense that Vel talks about. What you can do in Civ3 is elicit the help of other AI's in your war against the current adversary.

[Travathian]

I've used this strategy more than a few times, but primarily it has been in the Industrial or Modern ages. By then the AI usually has massive numbers of troops, varying from ancient era bowman to infantry.

By now the AI will have railroaded most of its usable squares, but not squares in its territory but outside of city squares. So all I do is look for a mountain range square in it's territory without an adjacent railroad, with more mountains adjacent so I know I have at least a 1 turn barrier. Then my transports plop down 16 infantry or mech infatry. On the computer's turn they tell me to leave, give them the finger and declare war, computer moves al of its troops onto the mountain next to me, on my turn I fortify for the extra %50 bonus and if I brought workers, build a fortress (either bring enough to do it in one turn or dont bother). Then I hit end turn and go watch tv for a while. If said mountain range is particularly long and winds thru the AI's territory, consider bringing a few artilery to just blow up roads and improvements, while moving the stack along the mountain range.
My best game so far I mowed down 200+ units in this fashion, granted half of them were inferior units, but those are one less thing I gotta kill to take a new city. Plus, this puts the AI into a 'war' mode, where it'll switch production to military units to throw at my virtually indestructable defenders. After the second wave I usually get all the other AI's to declare war on my target to attack, while I sit back on my home continent continuing to build infrastructure and gain tech. Even in a Democracy, with decent luxaries and infrastructure, I can last the 20 turns for the war pacts to end and then sue for peace, without going into nationwide riots, plus I am usually a few techs ahead by now.

[Tim_Hobbit]

Related question: Bringing along the workers to build that fortress is nice at the begining, but there have been times that I didn't want to leave that fort behind undefended, but didn't have the forces to leave a defense.

How do you remove fortresses?? I have checked the manual, the civilpedia and the forums, don't find a way to remove them!

Tim

[Cookie Monster]

Originally posted by Tim_Hobbit
Related question: Bringing along the workers to build that fortress is nice at the begining, but there have been times that I didn't want to leave that fort behind undefended, but didn't have the forces to leave a defense.

How do you remove fortresses?? I have checked the manual, the civilpedia and the forums, don't find a way to remove them!

Tim

I'm pretty sure that pillaging a tile that contains a fortress will remove it. If there are other improvements in the tile in addition to the fort however, you must pillage them first.

I've had the computer pillage my fortresses so I know that it is possible.

Give it a try. Good luck.

[Tim_Hobbit]

Thanks! Will have to be more persistant, I probably only did it once or maybe twice with a railroad on square and nothing happend to the fortress.

[ShredZ]

Does anyone find that the AI simple move around your fortifications? I build one on the road toawrds my empire, but they just move around it and march straight towards me preparing to pillage everything on their way... =\

[Arrian]

That's why this requires a chokepoint in order to work. The AI will definitely go around a fortified area if it can get there by land. It will sometimes use ships to go around a narrow fortified area, but less often.

Basically, use that to your advantage. If you build forts on the highground, the AI will be forced to march in the open land. And that's where you can kill them. Ask Theseus about "kill zones" sometime.

-Arrian

[FrantzX]

You know, I fear the day Soren teaches the AI to use artillery like we can.

[Harovan]

Yea, that will be awesome, to lose all improvements I built in 1000 years in just one turn. Btw: This day is close, although not with the AI: In PTW, in MP .

[Theseus]

I've been meaning to post about my Egyptian epic, but I keep getting distracted (as in, another turn another turn).

I thought people would enjoy this though...

Be afraid, be very afraid.

(But no arty, thank god... or that's what I thought until the Bombers showed up)