Engineering bulb concern

[Mantic0re]

When your post-engineering war is your first, how many Catapults do you mix in your stack with the Trebuchets?


Engineering gives you a choice of which siege unit to bring to a fight. With trebs only, your open field and city defense siege options are weaker. Using Trebs in the open field when you can bring catapults means more dead (and more expensive) trebs and a waste of hammers. On the flip side of the coin;capturing cities with all Catapults, when you have the ability to bring trebuchets is a waste of hammers.

Since we have to defend our new cities and can't always control where the combat occurs it seems that bring a mix is ideal. In the past I've tended to the simpler "more is better" response and walked with a bigger stack ignoring combined arms. This simple answer can become it's own problem when trying to improve to a higher difficulty. See the spoiler for my reasoning / rough math on the mix. I'm hoping the forums can point to some good rule of thumb that reasonably solves my dilemma. Maybe the simple answer is the best, if so, confirmation is appreciated.


I also need some clarification on collateral damage. Does collateral damage done depend on anything other than the siege unit type and the number of units attacked. Do damaged siege do less collateral damage?

Spoiler :

Trebuchets cost 60% more to build (and loose) than a Catapult.
Trebuchets have 60% more strength on attack than a Catapult. (vs cities only)
Catapults have 20% more strength than Trebs in the open field.
Bombard differences are ignored here (weakened siege bombard fine)

The problem isn't perfectly symmetrical. Hammer costs are straightforward but strength comparisons are not. Survival chances are not directly determined by attacker strength but must also consider the defender side of the problem. To muddy the waters further, the defender situation can change dramatically when considering siege, mixed siege options, Python defender selection from a stack, etc,.


In trying to keep a balance between the two extreme cases it makes sense to bring a mix of catapults and trebs. Roughly 40% catapults, or 2 Cats for every 3 Treb should give you equally efficient options when you can't determine where the opposing SoD will be confronted (and don't want to dig deeper into the maths).

 

[cseanny]

4:1 or 5:1 should be fine, or conversely, keep a good eye on SoD and 3 Cata's with collateral II upgrade should be ok. After all, you're gonna have Mace vs AIs with weaker melee units and a few pike/spear for mounted defense.

I have no clue about numbers so I can't help there. Just going from in game experience.

 

[Zx Zero Zx]

I only bring Catapults when I have attacked another civ with them first. Otherwise I just build city takers before I get Eng.

 

[capnvonbaron]

I'll stick a lot of catapults in my SoD if I suspect I'll run into the enemy's SoD or the enemy is fielding lots of catapults. The catapults do indeed "absorb" the collateral damage somewhat if your stack is attacked by siege, because they are still included in the checks the program does to determine what units are damaged. So there is no harm in bringing them along.

I would say I bring 2:1 cats:trebs if I expect light resistance, more like 4:1 if resistance will be heavy or tedious. In that case, I'll use bombard trebs to quickly strip city defense and maybe sacrifice a couple of cats to soften the garrison a touch before launching the flanking mounted or nut-crackers on them.

 

[Mantic0re]

Hhmmm, You don't throw enough siege at the city that the basic city-taker has winning odds?

I almost never take mounted with my siege stack (I will take elephants). Partly because my option at this point is HA's if I bothered to get archery + HBR. Knights are on a different tech path and have no base withdrawal chance (seems like an oversight imo, other horse units have a base chance).

I got into a habit of bringing more siege and looking at the collateral damage. When less than half a stack(or 3, whichever comes first) is taking collateral damage I switch to city-taker assaults.

Looks like I need to re-evaluate how I take cities.

 

[capnvonbaron]

Horse archers are good too, if you don't like knights. You're almost best shooting for Cuirs anyways, even if it does delay your offensive, because they aren't that much further up the research path. That way you can use nationalism to help draft a bunch of basic units while slow-building/whipping siege. Once Cuirs are in, you can really bring the pain.

 

[babar]

With a treb based army I wouldn't make a special effort to get HBR, but a few HAs/knights in a stack is very helpful for clean up, pillaging and flanking. They also get to the front much quicker, so I like to build stables in far away cities.

 

[Chuwe3y]

I'm still learning but I usually don't build siege unless I need to. For example 4 archers with 20% culture defense I won't really need it when axe men, has, or swordsman can take out. But if it has 100% culture defense with same unit I need to reconsider. I try to check enemies promotion also because city defenders has a possibility to take some of your units. I also check if city is on hill and what direction I'm attacking from. So far I'm learning to trick ais to routing units to defend another.

I've play this game but it was on monarch (was busy playing endless space and came back to civ 4 a couple weeks ago so I'm rusty) ai was defending with his Sod with 80% defense but instead of bombarding it I just use siege to attack and weaken them then took out the city with elephants. Does this strat also work on higher difficulty to reduce the turns of bombarding?

Or maybe the rng god loves me

 

[BornInCantaloup]

When your post-engineering war is your first, how many Catapults do you mix in your stack with the Trebuchets?

Do damaged siege do less collateral damage?

2nd question: no, it is a fixed value.

1st question: it stands out that post-Engineering, Trebuchets and Catapults won't serve the same purpose.
Proportion of Trebuchets or Catapults in your army should depend upon the use you could make of them (opportunity) AND upon how much you value that use.

e.g.: stack field combat vs city combat. If both were equally frequent and desirable, it'd be advisable to field as much catapults as trebuchets.
However, city combat is much more appealing and, unless the AI outnumbers the player greatly, it is also more frequent. If you don't expect the AI to counter you in the field, then there is close to 0 reason to build Catapults.

Trebuchets downgrade the Catapults. They acquire the niche use of a counter unit; as such, scouting info can justify them.

 

[shakabrade]

Since we have to defend our new cities and can't always control where the combat occurs

Wrong. You can and should control where the combat occurs. If you have narrow front line, you capture a city and can be 100% sure that city will be targeted by SoD. Then you leave it for enemies SoD to capture it, attack with CR2-3 trebs (no cats needed) and clean up. After that, there are no stacks left and there is no need for cats. If you have wide front line, you can expect an AI to attack the city closest to its borders. And if there are multiple of them meeting the condition, probably the one closest to AI capital, or one with least defense. I prefer founding borderline sacrifice city to be sure where the battle will be. But I do that only for very strong and technologically more advanced AIs. When fighting normal AI with wide front line, I split the stack first and take 2-3 cities with low defenses, then merge and fight the SoD wherever it is. It is predictable. Especially if you scout. Moral of the story is: bring plenty of siege or have fast reinforcements.

 

[Manco Capac]

^
True that. I vaguely recall in AI function for targeting cities that cities that once belonged to the AI are pushed foward in target score.

 

[Mantic0re]

I also need some clarification on collateral damage. Does collateral damage done depend on anything other than the siege unit type and the number of units attacked. Do damaged siege do less collateral damage?

2nd question: no, it is a fixed value.

Thanks for the confirmation. I hoped this was the case but I vaguely remember a thread that from a code-diver or someone running a lot of simulations to determine the best promotions for siege. I think it was determined that bonus damage from Barrage does not apply unless the siege unit survives and after a lot of testing CR1 Bar1 was picked best for the first two promotions. I thought tests were run that determined all siege damage was reduced by some % if the unit did not survive. Supposedly this was to prevent massive catapult sacrificial pushes.

It's been so long, that discussion may even predate BTS.


@Shakabrade:

It can be difficult to maintain the siege initiative deep in AI territory where their road network is against you. I was recently caught in the open field by a SoD I lost sight of with nothing but trebs/crossbow and trash from previous wars. It slowed down my war/game considerably.

If I'm feeling cautious I'll pick a city semi-close to the core AI cities (making sure I bring plenty of troops) and capture it, then abandon it in sequence a few times. The AI tends to fill it up and save me the hassle of scouting/moving units around.


These long front-lines are one of the problems I struggle with. I know the AI tend to target their former cities, but when you've taken three quickly it can be hard to know where to defend. High UnitProb AI suck in this regard. If I have too much trouble keeping these cities I'll just start razing them until I can consolidate my forces and cap the target. (If I can't play here then no one can.)


BIC makes a good point; scouting is a huge weakness for me.

 

[BornInCantaloup]

BIC makes a good point; scouting is a huge weakness for me.

Note I mention scouting in the larger sense Scouting encompasses information you have on your opponent, as a whole.

- From difficulty level + unit spam value of a leader, you can evaluate your military needs. With a little experience, maybe.
- Era, number of cities current attitude (AI is RExing, researching or building up) are more situational factors that you can use the same way. They're multipliers over the base factors, if you will.

Of course, having some units outside your territory during peace time will help preparing for war.
Knowing the lay of the land is essential to drawing invasion routes and determining defensive points. Civ4 becomes a whole different game, once you start considering the map's strategic value: land plots aren't just for developping cities.

- Ultimate scouting comes from espionage: that gives vision over the target's cities. It is VERY easy then to spot weak points. It is very likely that an AI that is on good terms with you and you don't spy on will not spy on you either. So it won't accumulate many EPs on you... so it will be very easy to have vision over the Ai once you target your EPs on it.

The whole point of this is to know what units you need and where. With a unit mass or another, units shouldn't be attributed the same roles (of course!).
Critical mass is another key word, here. With good scouting, you can know the critical mass of units you need to invade. Then you add in a few counters. Once you start capturing, lesser military cities can start training garrison units...

If you're facing heavy forces and you are declaring war, it might be best to set up one or multiple defensive points in your territory (there you might have catapults).
Either you can use your main stack to crush the ennemy's stack in your territory (wait, counter, advance).
Either your stack(s) can advance while you defend some place(s) else. How you should/can stack (larger sense) very much depends on the map layout and available opportunities.