A question of tactics

[mbbcam]

I would like to ask forum members for some thoughts about the question of tactics in warfare. It seems to me that the small details of how to fight and how to capture cities have generated surprisingly little discussion in the past. Perhaps this is because others find it easy. I certainly do not! A common experience for me is facing a constant stream of AI generated melee units that pop out of the cities, requiring me to keep ranged units at a "polite" distance to avoid being minced. I read reports of others capturing cities in five turns, and wonder how they do it. They are obviously doing something that I am not. So I would like to request others to share what techniques they use. It would be good if experienced fighters could give a little detail. In the past I have read advice on the level of "wait till the right moment, then move in", which, though well meant, was not too illuminating!

Cheers!

 

[so_what]

There are a few ways to take cities quickly.

1. Requires good diplomacy skills - Bribe a warmonger to attack this civ, let the warmonger's units tank the city defense and you try to snipe the last hit.

2. Requires science lead - Science well until you hit key techs well ahead of AIs. After that you can blitz the whole map in 20 turns.

3. Requires some time (actually this is the most stable tactic for me in early game during SV, CV, Dip) - you can worker bait out the defensive troops of out the city to kill them (in terms of how long this takes can vary by the strength of the AI.) Once the defensive troops are gone, there is usually a window of 3-4 turns. During which you can use a melee troop to attract the city's fire and use range troops to finish the city defense.

4. Requires decent science and some casualty. This happens when doing DomV and once you have researched key units you just have to spam them at the AIs. If you have science parity it is not hard to steamroll even the warmongers as AIs are just bad at war.

 

[mbbcam]

Thank you. Those are sensible recommendations. I've used worker baiting in the past. Perhaps the trick is recognising the window of opportunity, which I think I am not good at. I've tried bribery, too, though I've not often played on a map that made sniping easy. As to science, I've rarely managed to get ahead until really late in the game.

Perhaps the last point is the key lesson -- overwhelm the enemy with numbers. That I am not good at doing. I find it pretty hard to produce enough or earn enough gold to get a really large force. Perhaps I need to stop making buildings and just make units.

Cheers!

 

[vadalaz]

If you're fighting a strong AI without a tech lead or bribes, you're going to lose units. A few ways to deal with that are:

- Capture easy cities first, difficult cities last. This way you'll have more units, including well-promoted veterans, and Great Generals for the difficult city. And you can afford to lose a lot of units when the city you're attacking is the last one you want to take anyway

- Capture/found a city next to the difficult one and plant a citadel or two. Fortify melee units on citadel tiles and make sure you have other melee units that can replace it if it takes too much damage. Make sure the damaged unit has an escape path and doesn't get zone-of-controlled

- Try to only lose units that aren't very valuable to you. This includes scouts, your starting warrior, very low XP pikemen etc. Try to put those pikemen in positions where they'll take a lot of damage but survive, then use insta-heals and pillages so they can tank another round of shots

- Bring/capture workers, have them build roads on the AI's border for better mobility and repair pillaged tiles

- Ally CSs around the AI before you DoW, so that the AI has to defend on two or more fronts

 

[mbbcam]

Capture easy cities first, difficult cities last.

I think that may be one of the crucial ideas. Over the last 24 hours I've been trying an experiment and concluding that it may well be necessary to capture cities you don't want in order to get to the one you do. In other words, there may be a sort of "strategic path" across the terrain that you have to follow. In this last experiment I was attacked by my neighbouring warmonger just as I reached Machinery (with enough cash to upgrade my CBs), but the neighbouring CS was their ally, with 145 points advantage. There was no way I could flip it, though I could see the attack coming. The upshot was that I decided I had to capture the CS, which I did not really want to do because of the diplomatic penalties, before I could attack the next city, which belonged to the warmonger AI. I didn't really want that city, either, but taking it made it much easier to attack the AI capital. It is probably this sort of path across the map that I am not good at evaluating (among many other things!).

 

[Tiberiu]

Hello! Great topic. You have also inspired me to post a story from one of my militaristic games, if you care to look at the story it contains a lot of pictures that you might find useful in how the heavy use of citadels, forts, terrain, can be used to overcome even the most difficult of enemies (almost runaway medieval shaka neighbour, in this case)
https://forums.civfanatics.com/threads/the-tale-of-a-great-war.668095/


Now, I will try to add a few general things to this discussion, other than what the previous posters said.

- It's good to know the damages, to accurately aproximate before hand how much damage units receive in combat. I used to be very afraid to be in range of cities during war, fearing I might lose units, but this is inefficient and loses out on the XP that could have been obtained from attacking and also from being attacked. I had for myself some charts where I noted down how much damage archers and composite bowmen receive from various city strenghts so I can consult during play. Crossbowmen are much tankier, they can take a lot of beating from cities and also from the melee units. Watch out for civs who have Tradition (Oligarchy +50% city attack) and for those with the God of Protection pantheon (+30% city attack), they need extra care when approaching in order not to lose units. Also, care when in Desert/March is needed for their combat penalty.

- An alternative way of taking AI cities is by just killing their units *and* creating a high enough difference of strenght that the AIs will capitulate and give them in exchange for peace. For this, monitor your army strenght and target's army strenght. Check with Military advisor from time to time but take everything he says with a grain of salt.
Obtaining cities this way in turn conserves the population and can make it harder to raze the city but on the other hand it conserves the buildings and can be worth puppeting. Sometimes cities that are not important or their place is not important can be sold to a weaker third party.
War can be resumed in 10 turns, repeating the process until there is only the capital city left.

- In some situations it can be worth trying to pillage the Horse and Iron resources from the enemy, this cripples their army temporarily, even a few turns can be enough to kill more units.
A horseman can be good for this purpose for his mobility or also a chariot in open ground.

- In the average game, success can be obtain without heavy use of citadel/forts but in those cases where the enemy has become too strong and he spams units faster than they can be killed by our army, or when the terrain is very advantageous for the enemy in a few choke points, they can make the difference between making progress or being forced to retreat.
For this purpose, the Honor Policy that increases their spawn on top of a really free general, can be a good idea in militaristic games as it gives you more of them to use.

Also, in my opinion, I think it's good to post for discussion various situations that you encounter in your games that you find difficult to deal with. It's much easier to discuss something concrete from a game looking at a screenshot, so that particular strategies that work in that case can be discussed. Sometimes general things are too general.

I hope any of this helped!

 

[mbbcam]

@Tiberiu Thanks, that is very useful. I'll have a look at the thread later. I'll have a think about some screenshots.

What you say about damage calculations is interesting. Everything in the game is really about numbers and calculations, I suppose, and I am not good at that. Are your tables resources that could be shared? I know I ought to take more trouble over evaluating damage, but I've never been a numbers person. I guess my whole approach to Civ is a bit too "fly by the seat of the pants" and not analytical enough.

 

[Tiberiu]

I don't have something good enough to share at the moment, but I found a better documented resource! I didn't know about this previously, I think it's pretty useful.

The most important principle about units is that you need a 3:1 ratio to make an instant kill. For example A pikeman should 1-shot an archer since 16 : 3 is more than 5, the str of an archer.
This is more or less how I used to calculate things, but I was using a 2.5 : 1 ratio just to account for those cases where the variance is not in your favour. So a Rifleman will not kill a Crossbowman in 1 attack, according to the resource, 39 strenght would be needed, 13 str of crossbow x 3 = 39. They only have 34 so we are safe (all else being equal, but you have to account for great general bonus, for promotions, etc). Also, by "safe" I mean, we will remain with around 5-10 HP...

What this translates to, in a game, is that in a theoretical situation where your crossbowman encounters a rifleman, one has the option to attack the rifleman instead of running away. (not always best idea, but the fact that you know you have this option it is good enough to improve the warmongering). This could yield bonus experience for the units (faster logistics+range), and bonus great general points.

https://steamcommunity.com/sharedfiles/filedetails/?id=170194443

 

[mbbcam]

Very useful -- thank you.

 

[Tiberiu]

@mbbcam I want to add something about damage calculation since I have tested multiple times now. It's about cities.

There seems to be a different formula that calculates a city's damage. For example a city of strenght 25 (no oligarchy or damage pantheon) will only deal ~15 damage to a swordsman with 3 shock and cover 1, while a composite bowman (11 is much less than 25 even with 15% from general) will also deal ~15 damage to the same swordsman. (see picture)

What this means, in practice, is that cities are much less dangerous than we perceive them because of numbers - especially those that lack Oligarchy (tradition +50% from garrison unit). A musketman of 24 strenght deals much more damage to a unit, then a city of strenght 24 does.

so the 3:1 ratio for killing a unit does not apply to cities. I expect that a city with 60 str will not kill a cannot that defends with 20 str.... This makes siege weapons more appealing than they seem at first glance. The real issue with siege weapons is not that they are bad, but that vs AI they aren't in fact needed, as ranged units do the same job almost as effective, while being stronger vs units too. For me, realizing how things work, this makes the unique siege weapons slightly more attractive, but only vs non-Tradition civilizations.

 

[mbbcam]

Glad someone is doing the calculations!

BTW -- I notice you usually write "strenght" -- it should be "strength".

I spent about ten years teaching English, and God knows how long marking English exams. I'm programmed to notice these things!

 

[Tiberiu]

Glad someone is doing the calculations!

BTW -- I notice you usually write "strenght" -- it should be "strength".

I spent about ten years teaching English, and God knows how long marking English exams. I'm programmed to notice these things!

Thank you for helping me correct and please let me know if you find other mistakes.

 

[mbbcam]

Your English is very good. One reason why that particular spelling caught my eye was because the rest of your writing is so correct. This was an outlier, and more noticeable for that.

 

[plus]

use the terrain,its your best friend.
melee on front,ranged behind.
start barracks,armory etc on few designated cities with good production.start wars early game,your army will improve via promotions,once u get 3 archers to ballistics and range u can take down any city till end of medieval.than apply same strategy on trebuche,must start early.use honour tree,if not all of it than at lrast the 50% to gg.

 

[martindo]

I think that may be one of the crucial ideas. Over the last 24 hours I've been trying an experiment and concluding that it may well be necessary to capture cities you don't want in order to get to the one you do. In other words, there may be a sort of "strategic path" across the terrain that you have to follow. In this last experiment I was attacked by my neighbouring warmonger just as I reached Machinery (with enough cash to upgrade my CBs), but the neighbouring CS was their ally, with 145 points advantage. There was no way I could flip it, though I could see the attack coming. The upshot was that I decided I had to capture the CS, which I did not really want to do because of the diplomatic penalties, before I could attack the next city, which belonged to the warmonger AI. I didn't really want that city, either, but taking it made it much easier to attack the AI capital. It is probably this sort of path across the map that I am not good at evaluating (among many other things!).

Very important if you have a strong airforce -- move the bombers to the newly captured city.

 

[martindo]

@mbbcam I want to add something about damage calculation since I have tested multiple times now. It's about cities.

There seems to be a different formula that calculates a city's damage. For example a city of strenght 25 (no oligarchy or damage pantheon) will only deal ~15 damage to a swordsman with 3 shock and cover 1, while a composite bowman (11 is much less than 25 even with 15% from general) will also deal ~15 damage to the same swordsman. (see picture)

What this means, in practice, is that cities are much less dangerous than we perceive them because of numbers - especially those that lack Oligarchy (tradition +50% from garrison unit). A musketman of 24 strenght deals much more damage to a unit, then a city of strenght 24 does.

so the 3:1 ratio for killing a unit does not apply to cities. I expect that a city with 60 str will not kill a cannot that defends with 20 str.... This makes siege weapons more appealing than they seem at first glance. The real issue with siege weapons is not that they are bad, but that vs AI they aren't in fact needed, as ranged units do the same job almost as effective, while being stronger vs units too. For me, realizing how things work, this makes the unique siege weapons slightly more attractive, but only vs non-Tradition civilizations.
View attachment 590547

When cities grow much bigger than that (say 20 population and more than 50 strength), they can wipe a whole unit from 100 to 0 in one shot.

 

[martindo]

I don't have something good enough to share at the moment, but I found a better documented resource! I didn't know about this previously, I think it's pretty useful.

The most important principle about units is that you need a 3:1 ratio to make an instant kill. For example A pikeman should 1-shot an archer since 16 : 3 is more than 5, the str of an archer.
This is more or less how I used to calculate things, but I was using a 2.5 : 1 ratio just to account for those cases where the variance is not in your favour. So a Rifleman will not kill a Crossbowman in 1 attack, according to the resource, 39 strenght would be needed, 13 str of crossbow x 3 = 39. They only have 34 so we are safe (all else being equal, but you have to account for great general bonus, for promotions, etc). Also, by "safe" I mean, we will remain with around 5-10 HP...

What this translates to, in a game, is that in a theoretical situation where your crossbowman encounters a rifleman, one has the option to attack the rifleman instead of running away. (not always best idea, but the fact that you know you have this option it is good enough to improve the warmongering). This could yield bonus experience for the units (faster logistics+range), and bonus great general points.

https://steamcommunity.com/sharedfiles/filedetails/?id=170194443

View attachment 588942

There's another factor going on here, though, when your unit gets hit by more than one unit/city on the same term. Although it obviously loses health, the second hit from similar type of unit as first hit does LESS damage. I've seen damage decrease two or even three times on same turn hits for a strong unit like XCOM.

 

[martindo]

Bombardment works well on cities. Astonishing to see how much increase there is from Artillery (already a powerful unit in mid game) to Rocket Artillery. I always go for the "shield" icon (resistance to bombardment) as second promotion for range weapons.

Bombers with the "hammer" icon (repair even if used during the turn) are a serious threat later in the game.

 

[martindo]

In regard to the OP, cutting off supply lines when you're invading is very useful. Drop a paratrooper or XCOM onto a road that connects your target city with the rest of its territory. You can immediately pillage both the improvement and road/railroad even though you can't attack on the "drop" turn.

As overview of warfare tactics leading to a NON domination victory, I just had a come-from-behind victory as Maya at Emperor level. Korea had 3 or 4 spaceship parts completed and I had only 1. So I declared war and dropped a bunch of XCOM around their two main cities Seoul and Busan. They put up a defense based on land and sea bombardment with no air. I was only able to take Busan before winning the space race, which they would've won if I hadn't destroyed their production capacity.

 

[Megalou]

I would like to ask forum members for some thoughts about the question of tactics in warfare. It seems to me that the small details of how to fight and how to capture cities have generated surprisingly little discussion in the past. Perhaps this is because others find it easy. I certainly do not! A common experience for me is facing a constant stream of AI generated melee units that pop out of the cities, requiring me to keep ranged units at a "polite" distance to avoid being minced. I read reports of others capturing cities in five turns, and wonder how they do it. They are obviously doing something that I am not. So I would like to request others to share what techniques they use. It would be good if experienced fighters could give a little detail. In the past I have read advice on the level of "wait till the right moment, then move in", which, though well meant, was not too illuminating!

Cheers!

Hi! Your description actually lays the groundwork for how to do it. That stream of melee units needs to be vanquished in a defensive manner before it is time to take the city. Of course, you use 80-90% ranged units to do this, and your losses should be close to zero on any level. The number of defenders to defeat may be reduced like so_what said, but it also depends on the difficulty level.

Patience? Yes. Rushing to capture a city may cause one or two recaptures by the AI, resulting in a huge population loss that is very bad in the long run.

Lastly, the AI seems to be well prepared when they themselves declare war, but rottenly prepared when declared war upon...