The Stack o' Doom: Effective Composition and Use of Offensive Military Stacks

Well thought out article. In the real world, er game, don't you often find it counterproductive to spend a lot of time building up the perfect stack? Getting there with three or four archers while the AI only has one or two can be effective a lot quicker.
 
Well thought out article. In the real world, er game, don't you often find it counterproductive to spend a lot of time building up the perfect stack? Getting there with three or four archers while the AI only has one or two can be effective a lot quicker.
I hope I got across the idea that the stack doesn't have to be "perfect", just appropriate for its era. I've never tried an Archer rush myself, I'm not sure how effective it would be. You'd have to bring a lot of them, I would imagine; then again, they are cheap.
 
I hope I got across the idea that the stack doesn't have to be "perfect", just appropriate for its era. I've never tried an Archer rush myself, I'm not sure how effective it would be. You'd have to bring a lot of them, I would imagine; then again, they are cheap.

Archers get large inherent bonuses when defending cities, and especially cities on hills. Archers also can be promoted with City Garrison. Archers get no offensive bonuses, and cannot receive City Raider. So I doubt it would be effective. Mali might be able to work it but for anyone else the number of archers you'd need per defender would be very prohibitive I'd think.
 
Archers get large inherent bonuses when defending cities, and especially cities on hills. Archers also can be promoted with City Garrison. Archers get no offensive bonuses, and cannot receive City Raider. So I doubt it would be effective. Mali might be able to work it but for anyone else the number of archers you'd need per defender would be very prohibitive I'd think.

I agree with everything said about archers. I was trying to illustrate a point. Sometimes its better to go with what you've got rather than wait several turns to go with something better. Knowing when to do which comes with experience, I guess.

The paper is an excellent analysis of what to do if you have the time.
 
I agree with everything said about archers. I was trying to illustrate a point. Sometimes its better to go with what you've got rather than wait several turns to go with something better. Knowing when to do which comes with experience, I guess.

The paper is an excellent analysis of what to do if you have the time.
I don't think the time involved in building an appropriate collection of units for an offensive is at all prohibitive; I do it in several games and it pays off enormously. If you're finding it takes too long, perhaps you're not leveraging game options such as chopping and slavery, or you may need to reconsider your city placement.

A pillaging stack of 3 units, as the article describes, shouldn't take long at all, assuming you have 3 cities with decent production. That stack can weaken an opponent while you build a slightly-larger city-taking stack.

I'd definitely hesitate to declare war with only Archers at my disposal for another reason: the AI loves to pillage. On flat terrain, even a Chariot (the AI also loves mounted units) is a decent match for an Archer, even a fortified one, and can you really build enough Archers to camp atop every single tile improvement? You could find your tile improvements gone and your economy in ruins in such a situation.
 
I'd definitely hesitate to declare war with only Archers at my disposal for another reason: the AI loves to pillage. On flat terrain, even a Chariot (the AI also loves mounted units) is a decent match for an Archer, even a fortified one, and can you really build enough Archers to camp atop every single tile improvement? You could find your tile improvements gone and your economy in ruins in such a situation.

I actually went and tried out a few archer rushes just for the heck of it. The problem is not that the AI pillages. Pillaging just isn't an issue, period. A couple of scouting warriors might try, but anybody else is going to run smack into the steam of reinforcing archers heading to help your attacking forces and their horses are going to be pillaged or owned by you long before they get a useful number of chariots.

The real problem is that it takes about 3 attacking archers for every defending archer to take an early game city. Getting there with 4 when they've got 2 isn't enough, you need 6. In my games, I generally ended up not having enough to take their capital and having to wander around their territory, pillaging and taking their other cities with my archers. I'd bring in more archers to take their capital and they'd build more. I'd lure a few out and kill them and they'd build even more. I tended to finally crush them, once and for all, shortly after I got the bronze or iron connected and running. CR axe vs archer? No contest.

----



Anyway, some general comments.

With regards to pillaging: Although in some cases, the suggested pillaging stack works well, in many cases my preferred pillaging strategy consists of 2-4 groups of 2 (or 3, once in a while) mounted units, with promotions specific to melee or gunpowder (whichever looks most dangerous). The enemy can possibly attack and kill a few, but I operate them close together, so whoever did it then gets killed by the other groups. If some units get damaged, they can quickly be escorted to safety, or even leave on their own. SInce I don't have a lot of different types, casualties mean I can consolidate groups without problem. Prioritizing targets is not much of an issue, I simply remove everything, including the roads. Of course, the drawback is this only works if the enemy doesn't have technology that would be effective against your horses in sufficient numbers (the AI rarely builds enough pikes, for example).

With regards to using a giant stack o' doom, yes, sometimes it will work, but sometimes piling everything together will lose you an army. I often break down into smaller stacks and the best I can say is that it works. It's not always the best choice, but that depends on how the tactical situation looks. Here's an example:
Suppose I show up at a city with an attacking stack of 15 units. The enemy launches an unexpectedly heavy counterattack with 4 siege engines and 4 other attackers. 4 siege engines means that most of my troops are going to take substantial collateral damage. They might even kill a guy or two. Then the other attackers hit me. Most of my best units are pretty beat up, so Most of those attackers will win. I lose 3-5 units and will need 3+ turns to fix my forces up. Many of the enemy attacks were successful, so they have some guys who can try again soon, plus they have time to bring in more siege and more attackers, keeping me from getting fixed up. At best, I probably won't be able to take the city without reinforcements. At worst, they keep on hammering and I lose the stack. It's happened more than once.
Now, suppose I'd broken my forces up into groups of 3. The enemy can probably kill one totally, and leave the next one with a single damaged catapult. But most of their collateral damage will be wasted and I have 3 undamaged stacks. I can quite possibly take the city out from under them the very next turn, allowing followup units to sweep along their roads and mop up.
Actually in most circumstances 3 is kind of light, I'd usually use 4 or 5, but it depends on the situation. Another advantage with smaller groups is that the advance can continue as soon as I've got a few guys who are healed up. I don't need to wait for the whole stack. I can possibly move out the same turn I take the city, rather than waiting for the whole group to be combat ready. I also find it helpful, if non-hostile roads permit, to send a catapult ahead, so that when I arrive, the defense is already toast.
Now, I may take a few more casualties this way (some of the time), but I also get cities taken quicker, which gets them producing for me quicker and the war over with quicker.
If the stack of Doom always works for you, I think it is because you are fighting AIs who are so totally outclassed that pretty much anything you do is going to result in a win. If all they've got is a catapult and 3 archers, there's nothing they can do, except perhaps pull the whole lot back to the next city and gift this one to some other civ to piss you off (neither of which the AI ever does).
 
With regards to pillaging: Although in some cases, the suggested pillaging stack works well, in many cases my preferred pillaging strategy consists of 2-4 groups of 2 (or 3, once in a while) mounted units, with promotions specific to melee or gunpowder (whichever looks most dangerous). The enemy can possibly attack and kill a few, but I operate them close together, so whoever did it then gets killed by the other groups. If some units get damaged, they can quickly be escorted to safety, or even leave on their own. SInce I don't have a lot of different types, casualties mean I can consolidate groups without problem. Prioritizing targets is not much of an issue, I simply remove everything, including the roads. Of course, the drawback is this only works if the enemy doesn't have technology that would be effective against your horses in sufficient numbers (the AI rarely builds enough pikes, for example).

With regards to using a giant stack o' doom, yes, sometimes it will work, but sometimes piling everything together will lose you an army. I often break down into smaller stacks and the best I can say is that it works. It's not always the best choice, but that depends on how the tactical situation looks. Here's an example:
Suppose I show up at a city with an attacking stack of 15 units. The enemy launches an unexpectedly heavy counterattack with 4 siege engines and 4 other attackers. 4 siege engines means that most of my troops are going to take substantial collateral damage. They might even kill a guy or two. Then the other attackers hit me. Most of my best units are pretty beat up, so Most of those attackers will win. I lose 3-5 units and will need 3+ turns to fix my forces up. Many of the enemy attacks were successful, so they have some guys who can try again soon, plus they have time to bring in more siege and more attackers, keeping me from getting fixed up. At best, I probably won't be able to take the city without reinforcements. At worst, they keep on hammering and I lose the stack. It's happened more than once.
Now, suppose I'd broken my forces up into groups of 3. The enemy can probably kill one totally, and leave the next one with a single damaged catapult. But most of their collateral damage will be wasted and I have 3 undamaged stacks. I can quite possibly take the city out from under them the very next turn, allowing followup units to sweep along their roads and mop up.
Actually in most circumstances 3 is kind of light, I'd usually use 4 or 5, but it depends on the situation. Another advantage with smaller groups is that the advance can continue as soon as I've got a few guys who are healed up. I don't need to wait for the whole stack. I can possibly move out the same turn I take the city, rather than waiting for the whole group to be combat ready. I also find it helpful, if non-hostile roads permit, to send a catapult ahead, so that when I arrive, the defense is already toast.
Now, I may take a few more casualties this way (some of the time), but I also get cities taken quicker, which gets them producing for me quicker and the war over with quicker.
If the stack of Doom always works for you, I think it is because you are fighting AIs who are so totally outclassed that pretty much anything you do is going to result in a win. If all they've got is a catapult and 3 archers, there's nothing they can do, except perhaps pull the whole lot back to the next city and gift this one to some other civ to piss you off (neither of which the AI ever does).

This is pretty much negated by a Medic III unit if you're playing Warlords. The difficult choice is therefore only when you face a huge number of siege units (eg. a dozen) and don't have loads of units in your stack to absorb the collateral. You would almost certainly lose a few groups if you break them up, but you reduce the risk of losing your entire stack. However, I find that the AI is more willing to attack a smaller stack, so it's really difficult to say for sure which way is better.
 
I've never been hit by more than 2 or 3 catapults, when I used a large unique stack. Meaning that the attacking catapults died, and I had to heal, which I did while bombing the city defenses to 0 = total loss for the AI, total win for me (free xps).
Later, in artillery era, it's different.
Usually there are railroads allowing the AI to send all his artilleries at you at once. This hurts :(...
 
Later, in artillery era, it's different.
Usually there are railroads allowing the AI to send all his artilleries at you at once. This hurts :(...

Yes, this is a key point. The AI loves Artillery; it's surprising it isn't programmed to love the earlier siege units nearly as much, since they make great counter-attacking units when some uppity human has the temerity to take one of their cities.

What I prefer to do when the AI has artillery is to try to find out where they are. The AI usually bulks them up in one city on a border with a neighbour. Then I declare war and try to attack that city on the first turn of war. This is, obviously, not always possible (though it's ideal for Marines, with their bonus attacking Artillery, especially if the city in question is coastal). Nevertheless, it's better to take out Artillery in a direct attack than to absorb their collateral damage. Even if you can't attack on the first turn, the AI often prefers to leave siege units in its cities to defend (go figure), so provided your stack sticks to good defensive terrain, they might be safer than you'd think. Bombers will damage Artillery, as I recall, so that's another option.

If you can't find or take out the AI's Artillery stack, you'll need to bring along about a dozen or so city garrison-promoted Infantry or, better still, Mech Infantry. Earlier in the game, you can also build city garrison Grenadiers and then upgrade them to Machine Guns, who otherwise can't get CG promotions.
 
Yes, this is a key point. The AI loves Artillery; it's surprising it isn't programmed to love the earlier siege units nearly as much, since they make great counter-attacking units when some uppity human has the temerity to take one of their cities.

What I prefer to do when the AI has artillery is to try to find out where they are. The AI usually bulks them up in one city on a border with a neighbour. Then I declare war and try to attack that city on the first turn of war. This is, obviously, not always possible (though it's ideal for Marines, with their bonus attacking Artillery, especially if the city in question is coastal). Nevertheless, it's better to take out Artillery in a direct attack than to absorb their collateral damage. Even if you can't attack on the first turn, the AI often prefers to leave siege units in its cities to defend (go figure), so provided your stack sticks to good defensive terrain, they might be safer than you'd think. Bombers will damage Artillery, as I recall, so that's another option.

If you can't find or take out the AI's Artillery stack, you'll need to bring along about a dozen or so city garrison-promoted Infantry or, better still, Mech Infantry. Earlier in the game, you can also build city garrison Grenadiers and then upgrade them to Machine Guns, who otherwise can't get CG promotions.

good call on attacking the artillery city first.

The AI builds catapults and trebuchets too, but there are 2 factors making the artillery era a lot worse :
- the AI upgrades all it's units. So it's a whole game worth of siege that can be thrown at you at this point
- railroads make it very hard to move out of range
 
Technically, you should even only be considering using an SoD if your troops aren't up to the challenge of taking enemy cities on their own.

For instance, a plains or grasslands enemy city with no culture and 2 basic archers can be easily overtaken with about 4 Combat 1 Cover promoted Swordsmen, a situation that's easy for an Aggressive Civ to engineer.

Once the Swordsmen heal up (ideally with a Medic 1+ unit), you can move on to the next "weak" city. This is classic rush strategy, of course, but it can be worthwhile to break off a few units from a stack for just such a purpose if the opportunity presents itself. In the Medieval and Renaissance ages, your Frigates can make weak coastal cities easy targets for such breakaway secondary stacks.

At the lower difficulties, it's not actually all that advisable to make a civ-conquering stack if you only mean to take three or four cities from a large civ. Getting all those units up later might be less advisable than just getting what you want quickly, then suing for peace.

This is particularly true if you're only interested in a particular city for a key resource. While the "optimal" SoD is useful as a key reference, it bears mentioning that it's always heavily subject to situational modifiers.

In the post-Renaissance age, you can split up your stacks into smaller stacks if the enemy is weak enough that you don't fear counterattack and feel that making multiple stacks will allow you a faster advance.

Aside from this commentary, I also have a question.

As everyone can surmise, I don't prefer to use City Raider promotions, except on Trebs which I ideally only use to attack cities anyways. I prefer to use Cover or Shock promotions on my Swordsmen because that makes them less susceptible to counterattack in case I decide to break them off to go after a weak city. Moreover, such promotions are also more useful in case I go against an enemy stack in the field or while in a captured city.
 
As everyone can surmise, I don't prefer to use City Raider promotions, except on Trebs which I ideally only use to attack cities anyways. I prefer to use Cover or Shock promotions on my Swordsmen because that makes them less susceptible to counterattack in case I decide to break them off to go after a weak city. Moreover, such promotions are also more useful in case I go against an enemy stack in the field or while in a captured city.

If you're fighting archers with swordsmen, then Combat promotions are generally almost as good as Cover, and a lot more flexible.
 
That's not true. I know because I actually did try Combat 2 Swordsmen against Combat 1 - Cover Swordsmen. The difference is significant, particularly when the archer in question is fortified in a city or a hill, or god forbid, in a city on a hill. The difference, particularly in hp loss, is even more significant when you're on a time table. You generally want to capture the cities you want ASAP, then sue for peace.

While it's true that Combat 2 Swordsmen are better in singles and doubles generally speaking, when you're talking about grouping more than two units in a single bunch, it pays to "specialize."

Hence City Raider or City Raider 2 Swordsmen rather than Combat 2 Swordsmen, especially in stacks 10 units large.

Shock Swordsmen are something I've found useful as well. Ditto for Cover Axemen. The logic behind these promotions is direct stack to stack fighting, particularly when you don't have the benefit of cata support (you don't want to sacrifice them on field combat, or you lost your catas taking a city).

Attacking with a Shock Axeman is no good since the best defender against him is an Archer, in which case his Shock promotion is useless. Likewise, attacking in the field with Swordsmen usually results in a defending Axeman, so your Cover Swordsman would not be the best for that.

For defense purposes, Shock Axeman works against attacking Axemen, as well as swordsmen. Attacking Archers aren't usually a problem for any unit. These units will serve just as well as strike defenders within your culture borders once your conquest is complete, whether from barbs in the early game or against encroaching enemy stacks.
 
That's not true. I know because I actually did try Combat 2 Swordsmen against Combat 1 - Cover Swordsmen. The difference is significant, particularly when the archer in question is fortified in a city or a hill, or god forbid, in a city on a hill.

Combat 2 Swordsman vs fortified unpromoted Archer in 20% culture city on hill is 7.20 vs 6.60 = 1.09.

Combat 1 Swordsman with Cover vs fortified unpromoted Archer in 20% culture city on hill is 6.60 vs 5.85 = 1.13.

Not a big difference.

The stronger the defending archer, the less the difference.

Combat 2 Swordsman vs fortified City Garrison 1 Archer in 60% culture city on hill is 7.20 vs 8.40 = 0.857.

Combat 1 Swordsman with Cover vs fortified City Garrison 1 Archer in 60% culture city on hill is 6.60 vs 7.65 = 0.863.
 
Again, I disagree. The odds ratio is one thing, but the hitpoint on survival is not indicated in the calculation. IMX, the Cover Swordsmen are both more successful and also require less healing.

Even a 10% difference in the odds ratio is quite significant, IMO.
 
The odds ratio is one thing, but the hitpoint on survival is not indicated in the calculation.

That's not true. The distribution of hitpoints when you survive the battle depends only on the odds ratio.

Even a 10% difference in the odds ratio is quite significant, IMO.

I would agree with that. But the difference in my examples is much smaller: 4% in the first example and 1% in the second example.
 
Shrug. If there's little point to specialized promotions, that in itself is a separate issue from stack use. I seem to get the impression that you think that the Cover promotion is useless?
 
Shrug. If there's little point to specialized promotions, that in itself is a separate issue from stack use. I seem to get the impression that you think that the Cover promotion is useless?

No. Cover is definitely useful if you might be defending against archery units. It's also useful if you're attacking archery units in situations where they have few bonuses (e.g., an enemy SOD in the field with a bunch of archery units in it). But often the situation is that the defending archery units have lots of bonuses, and in that case Combat is about as good, and more flexible.
 
But that's exactly the point I was making. These Swordsmen are more flexible when they need to be: they can attack cities with Archers, defend against archers or attack SoDs with mainly archer defenses, or simply take out archers en route elsewhere.

A Combat Bonus is arguably better when attacking a city (the most common situation in which an archer has plenty of defenses) but in that case, you would do even better with City Raider Swordsmen or City Raider Axemen.

On top of that, the added benefit of a Cover Swordsman being slightly better than a Combat 2 Swordsman, IMO, firmly weighs in favor of the Cover Swordsman, especially in situation where you can deny the enemy the use of Copper or Iron.
 
Several comments:
First of all, thank you DaviddesJ for reminding me about how combat promotions work differently from all other promotions and giving me an example that allowed me to understand what this means.

The thing is, the attacker's combat promotions are applied to the attacker's combat strength, while all other promotions are added up and then applied to the defender's strength.

So, in the above example, the 10% combat promotion would be increasing the attacking swordsman's strength by .6, while the cover promotion would lower the defender's strength by .75, which isn't a big difference. On the other hand, if the archer were attacking, the defending swordsman would still get a .6 from the combat promotion, but would get 1.5 from the cover promotion, which is a much bigger difference.

I guess the lesson here is that combat promotions are more valuable if you are planning on being the attacker and have a much higher base combat strength.

I would also like to note that the way the combat system works, even a 1% increase in combat strength can make a massive difference in certain circumstances. Here's why: The amount of damage per hit is calculated at the start of combat and is based on the relative firepower of the two sides. If the odds ratio is exactly 1:1, each side will inflict 20 (out of 100) points of damage per hit. However, if the odds ratio swings even a tiny bit in favor of one side or another, the way this implemented means that the favored side will still be doing 20 per hit, but the ever so slightly weaker side will be doing 19 per hit. The stronger side still needs 5 hits to kill their foe, but the weaker side suddenly needs 6 hits to get the target above the 100 damage needed to kill. Alone, the 1% increase in chance of hitting may not be a big deal, but combined with needing 20% more hits, it swings the odds drastically in favor of the stronger side. Note also that there's a similar gap (increase/decrease in number of hits needed) at 5:3 or 3:5 odds (if I remember correctly). So, if a small change in power gets you past the 1:1 point or either of those others, it's well worth even that small change in combat power.
 
Top Bottom