An overview of the problem with sieges and potential solutions.

weefraze

The criticism of siege battles began prior to the release of the first game, assurances were initially given by CA that these changes would show significant improvement from previous sieges. I believe there is some truth to this claim. It is worth keeping in mind that people have diverse opinions on what exactly entails a good siege battle and some will be satisfied with the state of sieges in the series so far. That said, it is also worth keeping in mind that others are not satisfied with the current state of sieges and legitimate criticism ought to be heard out. What I intend to do in this post is subjective, it is my opinion on the good and bad of siege battles, my hope is that enough people will have similar experiences and thoughts on the issue such that we can push for some changes. I believe this is achievable over the course of the series if enough momentum is gained. CA has shown a willingness to change parts of the game that the community has brought to their attention, this is to their credit. I would like to divide this post into four related but different topics: Maps, AI, Immersion, and Potential Improvements. I cannot cover in detail all of these topics and the nuance they require, what I will give is a summary that hopefully lays out core details.

Maps

In this section, I am discussing the layout of sieges, the design of the maps and what they play like functionally from a gameplay perspective. Firstly, I want to look at walls. There are 2 kinds of walled siege maps in the first game, we have the arrow map which is shown in the first picture below and we have (roughly) the straight wall map which is shown in the second picture below. There are very slight variations on both maps, the more noticeable variations are with the straight wall maps, two such examples are shown in the third and fourth pictures. In my opinion, these slight variations do little to improve or keep fresh the gameplay. It is worth pointing out that if you choose to attack one side in the arrow maps you are functionally speaking attacking a straight wall. This is usually done so as not to spread your forces out and to rush down and control the walls (and gate) quickly. This same tactic is usually used on straight walls as well, where you focus one side. The point I wish to make here is that there is not enough wall variation. We have arrow walls and we have straight walls. Even if you use both sides of the arrow and straight wall maps for a more spread out battle you are looking at very little variation in actual gameplay with regards to walls.

Let's move past the walls now and look at variation in the city design. The cities have a fair bit of variation in obstacles or structures that are inside the walls and at least in terms of design, there is potential for creating chokepoints between these structures as you fall back to defend the capture point. As I have said, there is variation, however, I feel like the city design is overly simplistic. There are usually direct paths from the walls to the capture point, there are exceptions to this, for example in Karak Hirin and Erengrad the capture point is sort of hidden behind a building. Nonetheless, it is not much of a challenge to get to, you march your army down a main road, take a left or right and you're at the capture point for a fairly open battle.


1. 2. 3. 4.

AI

The AI's focus is on defending the walls. This is where the battle begins and is essentially where it ends as well. The wall battles, in my opinion, are actually pretty well done. The AI does a good job of setting up ranged units and artillery to attack the oncoming army, it keeps infantry units and sometimes heroes up defending the walls against enemy invaders. A problem here though is that ranged units on the walls tend to get caught up in the fighting once the enemy lands. Another shortcoming I have noticed is that the AI tends to allocate a weak force to protecting the gate (1-2 units + hero/lord). The problem with this shortcoming is that if that gate is broken down by a large monster or lord unit and followed by other monsters or cavalry, the defending force of the gate is rarely capable of dealing with the attack. In some instances, this gives the force that broke through easy access to artillery units and a free capture point while the attacking units have tied down the defending units at the wall. Anyway, the main problem with the focus being the fight on the wall is once the wall fight is over, the battle is essentially determined. If you are the attacker all that is left to do is clean up the city, it shouldn't feel like cleaning up a city. This ought to be a phase 2 of the battle, the battle for the city.

Immersion

Immersion is incredibly important for absorbing you into the Warhammer world and for creating the kinds of battles that you will remember and enjoy through this fantasy narrative. I would like to discuss some of the positives of the game here that I really enjoy. The aesthetics of the game overall are very good, from the architecture of vampire cities, the models, idle animations, and running animations of skeletons (my favorite). The game does a great job of exploring this world, you get to use monstrous units, legendary lords, and little gits. I believe CA did a good job overall with regards to immersion for this first game. The relevant area where immersion for me breaks completely is siege battles. When I launch an attack on Altdorf it doesn't feel like I am attacking Altdorf, it feels like I am attacking "a straight wall map". This is immersion breaking. Akendorf, Altdorf, Averheim, and so on, should not feel like straight wall maps, they should feel like unique imperial cities rather than the same (roughly) imperial cities. This doesn't just apply to empire, it applies to all sieges - they feel the same and it break's immersion.

Potential Improvements

Thus far I hope to have given a basic summary of sieges. What I aim to do in this section is put forth some suggestions for improvements to this area of the game. Some of these improvements have already been suggested by the community. This is the area that I think requires the most contribution from the community and where those of us who would like to see improvements really need to explain, coherently and constructively, what we would like to see. As a side note to the suggestions that are about to be made, there is no reason why sieges as they currently stand can't remain in the game in some form. This leads to the first suggestion.

1. Introducing AI personalities.

This would, at the start of the battle, randomly select a type of defender or attacker for the AI to act as, their personality. For instance, a defending AI that does not care to fight at the walls and would rather lure you into the city and fight among chokepoints. Another could be a defending AI that will keep ranged units on the walls (securing the towers) and send out units to engage your army. For instance, if you decide to rush one part of the wall, they send cavalry out the opposite gate (if available) to come in behind and take out your artillery or any stragglers. How about an attacking AI that instead of spreading their army out to attack all points, instead rushes one part of the wall like players currently do. Another attacking AI that will rush to engage your units on the wall with minimum force to tie them up while trying to rush to your capture point. You get the idea, set up different types of AI and have it play a random role or perhaps attach AI personalities to legendary lords and have them attack and defend siege battles in unique ways.

2. Variation in wall and city design.

Design more wall types than the straight wall and arrow wall maps. Go back to previous total war game iterations for siege battle designs, have three and four sides for potential walls to attack. Have walls within walls as we had in medieval 2, where you break through the first outer wall and then have to proceed to fight again in an inner siege. Redesigning cities with things like capture points that are themselves hidden away, so that in order to get to them the enemy needs to get through a very defensible chokepoint that the AI knows to defend. Expanding the size of the cities so we get to interact with some of that beautiful background, so we can get the experience of a large city and defend down alleyways and crossroads.

3. Random or planned traps in siege battles.

Imagine during the planning phase before you start the battle that you and the AI could place traps around the city. Oil for the gates, houses that will collapse on enemies as they pass by them, parts of walls that collapse when a unit moves onto it. Take the following as an example of what I mean. Behind the playable area in Altdorf are two inaccessible bridges.

In this hypothetical, the map extends back into that city for further defense. Imagine you are attacking this city and now have the choice of crossing one of these two bridges. The one on the left looks easier because there is no gate and perhaps on the right side bridge there is also archers on the walls and spearmen behind that gate (walls can be added). So, you decide to go to for the easy bridge on the left and will then flank behind them. The bridge then collapses or explodes when you place units on it killing them units, or perhaps oil flows out from the towers and makes it uncrossable and kills units that try to do so. This punishes you for trying to take the easy route and forces you to confront the gate crossing.

These are just examples, I'm sure people will have better suggestions. Maybe the amount of random or planned traps that can occur in a siege battle depends upon the size of the forces in play or upon the perceived advantage of the attacking force. If you as a defender are overwhelmed, maybe you get more traps to play with. Or another way of introducing it might be that for each turn the enemy lays siege to you on the campaign map, the more time you have had to prepare and hence get more traps to use. Additionally, certain hero types like the master engineer or even just race bonuses to Dwarfs or Skaven, for instance, can increase deployable traps.

4. Instancing Sieges.

I believe this has been brought up before, however, it recently was brought up again on Reddit with a discussion here.

The idea is straightforward, we divide sieges (perhaps important ones like capitals) up into multiple stages. This way you could have a fight for the walls in one instance, after that is complete you then move into the next location to fight for the city, after this perhaps a fight for the inner sanctum or capital building, maybe a final battle in the University of Altdorf, or the Colleges of Magic if in Altdorf.

------

I appreciate anyone who takes the time to read this. It turned into a longer post than I initially intended. I hope I have been clear throughout, it is not foolproof, I am sure there will be mistakes and oversights. My desire is to help the community push for siege battle improvements through discussion and by ideally getting the attention of the development team. Any criticisms are welcome.

Hircane

Very welcome ideas. I suspect we wont see any free changes tho. Lets hope they redo the system in the next total war franchise!

TyrionLamperouge

The only thing that could redeem sieges is a total rework of the mechanics (no teleports and no instant ladders) and map more like attila/shogun2/med2 in size and complexity .
With all the money they have made and after claiming that they would focus in land battles (fater saying no to naval battles) i really hope to see a total rework.

Vanilla_Gorilla

TyrionLamperouge said:

The only thing that could redeem sieges is a total rework of the mechanics (no teleports and no instant ladders) and map more like attila/shogun2/med2 in size and complexity .
With all the money they have made and after claiming that they would focus in land battles (fater saying no to naval battles) i really hope to see a total rework.

How can it possibly be like Shogun 2 and Medieval 2? Those two are opposites. Shogun 2 had pre-TW the best siege gameplay of any TW game, and was very simple, units could climb walls (they didn't even need pocket ladders, they could just straight up climb them), and the siege maps themselves were relatively small and simple. M2 had the most complex sieges of any TW game, and in my humble (but always, always correct) opinion had the clunkiest, worst sieges in the series to date.

TRex

For me, i hate the scale, it's all wrong. The settlements in Rome, Attila etc. seemed real. The walls of Couronne, for example, probably suit a city of more people than the old world actually has. Thousands of towers everywhere. But it isn't a city of 100 million people and the armies are tiny skirmish forces. They've got their scale all mixed up and it makes it seem like a silly video game. The sieges in Attila etc (forgetting of course their limitations) seemed real. They are so much better than now, so much better. To me, the Bretonnian cities in particular are awful looking. The art is good, but it looks like it was placed by a 5 year old, who doesn't understand the concept that less is clearly more. A few towers = good, millions of towers in a city of several thousand people is so...naff.

In over 1000 hours, i've yet to manually finish a single siege battle. They are so [email protected], i just won't even think about bothering to do one. And they used to be my favourite bit.

Layzan

This is the kinda stuff I wish CA would work on instead of total war 2. I'd pay the 60 if they made non wall settlements, maps, sieges, the A.I much better and more interesting.

Mods are showing how good things like non walled settlements battles can be. CA should be on top of this.

Itharus

Honestly this game doesn't even have siege battles. Just field battles with a cool skybox and some destructible terrain. I know that sounds harsh... but the siege defenses don't even constitute a speed bump.

I hate to say it but the sieges in TWWH are the worst in the whole series other than the original Shogun.

Luckily it's so gd fun watching these fantasy races go at it that you don't notice it unless you're dwelling on it.

Andkat

IF you copy pasted siege maps at complete random from preceding TW games instead of the maps we have now, it'd be more enjoyable while being about equally as immersion breaking as these recycled monowall cityfronts where the most interesting part of the map is the inaccessible background texture.

LestaT

I would certainly love more variation of the siege map. While I certainly love this kind of battle (second only to Attila), the downside is that, once you see one, it feels like you have seen them all. I cant even remember if there's any difference between Aldofr or Marienburg (is there any?).

Now, I dont expect that each settlement to be unique (though I would very much like it) but as of now, I can't even differentiate between level 3 or 4 or 5 or even level 3 minor settlement with walls.

My dissatisfaction was not about the battle itaelf, but more about the map design.

I would love for example if I get to fight on 2 walls if I bring 2 stack (more if more stack) and also gets to choose which wall to attach, each have different layout. Maybe attacking from the north wall will have different layout than attacking the east wall, in terms of tower placements, gates position, gate numbers and even road that leads to the center.

This way, even if we have to attack a same settlement twice, it will feel like a new settlement.

Overlord87

My opinions on your points:
1) Siege maps: well, it's hard to disagree here. Maps variety is a priority and in my opinion it's absurd this has not addressed so far by CA (ok, mods fixed it, but still, no excuse). In particular, all capitals and at least a few major settlements should have special maps. I can understand the concept of reducing the number of walls to help the AI, but still, there must be some degree of flexibility CA can use to make maps more interesting. When I attack Karaz-a-Karak, I want to know it will be a bloodbath, with engineered landslides, lots of fortified towers with artillery, and so on. And I wan to see landmarks. So, in short, totally agree here. More maps, CA, more maps.
2) AI: not many complaints here honestly. I think AI makes a decent job at defending and attacking, not always, but most of the time. I think in previous games it was much more easy to exploit the AI's dumbness honestly.
I'll add that in most sieges, in fact, once you conquered the walls, the battle was effectively over. That is quite clear, obviously. If you resorted to lock yourself in a castle, it meant the enemy had significantly more forces than you, and if they break the walls, well, that's it. Skirmishing in the streets won't change the outcome. I wouldn't like the AI to leave the walls to place troops in the streets.
3) Immersion: no problems here for me.

So, regarding your suggested improvements, I totally agree with point 2 and partially with point 3, not so much with point 1 and I absolutely wouldn't like anything like you suggested in point 4.
My 2 cents.

Lin_Huichi

Excellent post. Brilliantly written and it outlines the flaws very well. Honestly, I can see little merit to Warhammer sieges, not even how they supposedly help the AI. Hence the wider streets in WH2 is little consolation to the bigger problem.

The best part about previous sieges is their three hundred and sixty degree structure. I guess I took that for granted, I never thought in my wildest dreams that CA would change it. I don't think there will be drastic changes to siege structure in this trilogy, but I hope these sieges disappear completely after WH3 is done

Andkat

AI: Err....what? The AI literally will stand in a breach as you shoot at them point blank (they will respond if you shoot at them through a broken gate, after a while). If you bombard enough to induce them to abandon the walls, you can mount ranged troops on the walls and half the time they won't try to remount the walls to fight them off . They'll let their army get pounded to dust by mortars and rockets rather than withdrawing vulnerable troops into the town. They'll often piecemeal deploy troops to the gates as they are being overwhelmed, rather than positioning their entire army to respond coherently to the breach. The siege AI is terrible.

doclumbago

I still think that the AI was overburdened with flyers, magic, monsters etc.
Remember those last units that assambled in the center and were letting themselves
shot to death witout moving in Rome 2 ?
AI in former sieges was never 100% perfect. That´s why they broke it down to basics.


Here´s some stuff I´d like to see if the old siege maps don´t come back


1:MOST important and easy to do:
make the architecture non-material for the mouse cursor.
Somehow those big towers stand in my way when I want to click on units.
I have to scroll and turn like mad and basically guide single units through the city.

2:Less range for towers.
If you bring a piece of artillery chances are you loose it during the battle. The towers make sure of that.
Also you can´t have a reserve. You have to charge everything you have at the walls

3: Have some buldings in village battles like Rome 2.
That makes small town battles actually interesting enough to not aut-resolve them.
That wooden palisade in villages in Rome 2 was also a nice touch.
Maybe something for the Norsca Settlements ?

4:Unique mechanics for factions and cities.
Middenland (example) is famous for it´s single giant-loaded cannon.
Even if it´s only represented as a bombardment spell, I want it in.

Have lore-friendly special siege units for each race, not the same tower and ram painted in different colours.

->Dwarf miners could burrow underneath walls.
->Forest Spider units could climb walls
->Treemen could destroy walls
-> Chaos should have access to Flayerkin/Siege Giants
-> Wood Elve should not use rams and towers but maybe grow vines and trees
directly on the walls

If you build special buildings, you get special bonuses
->Temple of Sigmar Warrior Priest/Witch Hunter/Morale Buff for garrison
->Magic College: Adds a wizard/WoM buff to garrison
...

irurobin

It takes a total rework.

weefraze

Andkat said:

AI: Err....what? The AI literally will stand in a breach as you shoot at them point blank (they will respond if you shoot at them through a broken gate, after a while). If you bombard enough to induce them to abandon the walls, you can mount ranged troops on the walls and half the time they won't try to remount the walls to fight them off. They'll let their army get pounded to dust by mortars and rockets rather than withdrawing vulnerable troops into the town. They'll often piecemeal deploy troops to the gates as they are being overwhelmed, rather than positioning their entire army to respond coherently to the breach. The siege AI is terrible.

I rarely use artillery in sieges (mostly auto resolve now). But what you have said is true, they respond very poorly to bombardment on the walls. They will use their siege equipment in the city to try and take out yours at times. But they are willing to just let their army be fodder to defend walls that are honestly not worth defending. By standing there, they suicide their army. This could be sorted out with AI improvements, which is a fairly common request by the community.

doclumbago said:

1
make the architecture non-material for the mouse cursor.
2:Less range for towers.
3: Have some buldings in village battles like Rome 2.
4:Unique mechanics for factions and cities.
Middenland (example) is famous for it´s single giant-loaded cannon.
Even if it´s only represented as a bombardment spell, I want it in.

Have lore-friendly special siege units for each race, not the same tower and ram painted in different colours.

->Dwarf miners could burrow underneath walls.
->Forest Spider units could climb walls
->Treemen could destroy walls
-> Chaos should have access to Flayerkin/Siege Giants
-> Wood Elve should not use rams and towers but maybe grow vines and trees
directly on the walls

If you build special buildings, you get special bonuses
->Temple of Sigmar Warrior Priest/Witch Hunter/Morale Buff for garrison
->Magic College: Adds a wizard/WoM buff to garrison

I like these suggestions, they are good quality of life improvements. 4. in particular, as someone who has played a ton of wood elves, it's something I would like to see.

Buguba

I think a lot of your suggestions are well-thought out and interesting. I especially enjoy the idea of the AI personalities.

I have struggled to enjoy sieges in almost any iteration of Total War because I feel they diminish the parts that make Total War combat fun (flanking, morale shocks, maneuvering, etc) while emphasizing what are often the buggy and/or unpolished components of combat (units randomly suffering from flanking penalties due to narrow wall spaces, the AI not understanding how to address map fixtures, or grind out melee fights).

I think walls should be a settlement fixture exclusive to only very special capital cities (such as Altdorf or Black Crag), and that the vast majority of settlements would be designed akin to settlements from Rome 2 and the like. These settlement battles were fun because the attackers could approach from any angle, but streets and defensive emplacements allowed the defender to funnel the attacker through expensive choke points. Settlements normally had elevated mezzanines or levels that could allow the defender to place skirmishers in advantageous positions, but attackers were not always forced into boring melee grind outs just to reach the victory point.

Pr4vda

I liked the small settlements battle from Attila and wish them to be back.

It was something between the "normal" battle and the siege. With the possibilites to put barricades and also put the town on fire, combined with the small towers and the chock points you could create, it was really interesting. This system could have been improved for TWW, depending on the faction and the level of the town (or the "yellow" garrison building).

For instance :
- An empire settlement level 1 should have nothing special. Level 2 or 3, you could have barricades and some towers. With the garrison building and a level 3 city, you would also have wooden walls with wooden gates, more barricades and defensive abilities (remember what you had in Rome2) available with the garrison units.

Concerning the sieges, I like the big battles on the walls. But as you said, the AI can't defend the gates, but in the same time, you have zero option when a monster unit rush your doors. You can not shoot them from above or from the flanks, there is no oil, etc. and that's a shame. Moreover, artillery units are useless for the defenders and the attackers.
As an attacker, the unlimited ranges from the city towers will crush you. As defender, you can not put your artillery on the walls, and units such as the cannons have zero utility.

Finally, the battles inside the cities are meh, there is only a single capture point, and you don't have the feeling of an "epic" siege. Imagine some buildings could hidde some of your troops for an ambush inside the city, some towers could be inside the city to have support fire.

That's why your idea of dividing siege in multiple battles is interesting.

Sieges are really the worst part of TWW1 (and I suspect, of TWW2), and that's sad to say it because the art of the maps are GORGEOUS and the modders can do really great things, so CA could do even better.

Arthas_Menethil

Pr4vda said:

I liked the small settlements battle from Attila and wish them to be back.

It was something between the "normal" battle and the siege. With the possibilites to put barricades and also put the town on fire, combined with the small towers and the chock points you could create, it was really interesting. This system could have been improved for TWW, depending on the faction and the level of the town (or the "yellow" garrison building).

For instance :
- An empire settlement level 1 should have nothing special. Level 2 or 3, you could have barricades and some towers. With the garrison building and a level 3 city, you would also have wooden walls with wooden gates, more barricades and defensive abilities (remember what you had in Rome2) available with the garrison units.

Concerning the sieges, I like the big battles on the walls. But as you said, the AI can't defend the gates, but in the same time, you have zero option when a monster unit rush your doors. You can not shoot them from above or from the flanks, there is no oil, etc. and that's a shame. Moreover, artillery units are useless for the defenders and the attackers.
As an attacker, the unlimited ranges from the city towers will crush you. As defender, you can not put your artillery on the walls, and units such as the cannons have zero utility.

Finally, the battles inside the cities are meh, there is only a single capture point, and you don't have the feeling of an "epic" siege. Imagine some buildings could hidde some of your troops for an ambush inside the city, some towers could be inside the city to have support fire.

That's why your idea of dividing siege in multiple battles is interesting.

Sieges are really the worst part of TWW1 (and I suspect, of TWW2), and that's sad to say it because the art of the maps are GORGEOUS and the modders can do really great things, so CA could do even better.

They aren't going to do multiple capture points after Rome 2 the game in which they turned capture points into morale buffs instead of city capture after all the complaints.

weefraze

Pr4vda said:

Concerning the sieges, I like the big battles on the walls. But as you said, the AI can't defend the gates, but in the same time, you have zero option when a monster unit rush your doors. You can not shoot them from above or from the flanks, there is no oil, etc. and that's a shame. Moreover, artillery units are useless for the defenders and the attackers.
As an attacker, the unlimited ranges from the city towers will crush you. As defender, you can not put your artillery on the walls, and units such as the cannons have zero utility.

Finally, the battles inside the cities are meh, there is only a single capture point, and you don't have the feeling of an "epic" siege. Imagine some buildings could hidde some of your troops for an ambush inside the city, some towers could be inside the city to have support fire.

That's why your idea of dividing siege in multiple battles is interesting.

Sieges are really the worst part of TWW1 (and I suspect, of TWW2), and that's sad to say it because the art of the maps are GORGEOUS and the modders can do really great things, so CA could do even better.

I would just like to point out that dividing the sieges into multiple battles was not my idea. Nonetheless, it is an interesting idea, if you take a look at the Reddit thread, I believe, some concerns have been raised. An example of one potential problem would be that if the defending AI does retreat to Phase 2 after getting destroyed in Phase 1, are they really going to put up much of a fight? Aren't their units going to be severely weakened anyway? It's a valid criticism, I do think there are ways around it though, like the introducing of traps, reinforcements, last stand buffs or a number of others. The balancing and implementation of these ideas would be the main problem.

Artillery actually can be quite useful as an attacker, you need to bring enough to quickly take down the towers on the walls and then you can pretty much abuse the AI by keeping your army back and having artillery attack the defending units on the walls. It is worth mentioning this is best done by attacking one-half of the wall, if you spread out, then, so does the AI and as a result, they will capture and use more towers. The AI does not retreat from artillery fire and, therefore, the units are sitting ducks up there. You can wipe out a good portion of the wall defense before engaging your troops. I'm not a big artillery user due to the races I play, so, I couldn't say what the ideal amount of artillery to take is, but I suspect there is a very viable build for attacking with siege artillery (against AI).

I share your concern for TWW2, this was part of the motivation for making the OP and also looking to TWW3.

Itharus

Towers don't need less range. If you put artillery on top of a wall -- it's gonna have better range than artillery on the ground, cuz physics.

Sieges are supposed to be deadly. Normally a besieged fortress can fight off many times its number in enemies. Sieges basically never favor the attacker. That's why fortifications were built in the first place.

boyfights

Imo it's good the way it is, keep it up CA

boyfights

One wall: good
Long range towers: good
No chokepoint tedium: good
AI can find a path inside: good

Seems good to me

Itharus

Chokepoint tedium? Seriously? So... you just want field battles on an open field basically.

Vanilla_Gorilla

Sieges are crying out for better AI. Maybe not attacking (they might be less fun that way) but defending they need to focus more on the walls, and avoid players shooting them like they're in a fishtank.

Itharus said:

Chokepoint tedium? Seriously? So... you just want field battles on an open field basically.

A straw man is a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent's argument, while refuting an argument that was not presented by that opponent.

I don't want choke points, they take away all the fun and strategy by making the fight easily winnable by using said points.

boyfights

No I dig the walls part, but just don't find anything engaging about units grinding each other down or enjoy feeling obligated to play siege defenses with five defending units because I know I can cheese out the AI

HorseWithNoName

Arthasmenethril said:

Pr4vda said:

I liked the small settlements battle from Attila and wish them to be back.

It was something between the "normal" battle and the siege. With the possibilites to put barricades and also put the town on fire, combined with the small towers and the chock points you could create, it was really interesting. This system could have been improved for TWW, depending on the faction and the level of the town (or the "yellow" garrison building).

For instance :
- An empire settlement level 1 should have nothing special. Level 2 or 3, you could have barricades and some towers. With the garrison building and a level 3 city, you would also have wooden walls with wooden gates, more barricades and defensive abilities (remember what you had in Rome2) available with the garrison units.

Concerning the sieges, I like the big battles on the walls. But as you said, the AI can't defend the gates, but in the same time, you have zero option when a monster unit rush your doors. You can not shoot them from above or from the flanks, there is no oil, etc. and that's a shame. Moreover, artillery units are useless for the defenders and the attackers.
As an attacker, the unlimited ranges from the city towers will crush you. As defender, you can not put your artillery on the walls, and units such as the cannons have zero utility.

Finally, the battles inside the cities are meh, there is only a single capture point, and you don't have the feeling of an "epic" siege. Imagine some buildings could hidde some of your troops for an ambush inside the city, some towers could be inside the city to have support fire.

That's why your idea of dividing siege in multiple battles is interesting.

Sieges are really the worst part of TWW1 (and I suspect, of TWW2), and that's sad to say it because the art of the maps are GORGEOUS and the modders can do really great things, so CA could do even better.

They aren't going to do multiple capture points after Rome 2 the game in which they turned capture points into morale buffs instead of city capture after all the complaints.

As long as those multiple capture points are not in the same map, I don't see why not. I would actually like to have multiple stage sieges for very big cities where for example in the first stage you fight for the outer walls and in the second stage you fight for the inner walls and central place (with a seperate garrison and units that managed to flee from the first fight). This would for once add diversity and also allow for battles to be more challenging without having more units on the map at once (danger of easily devolving into blobs and all-out brawls if not on walls).

Vanilla_Gorilla

Fight a second battle after winning the siege? That's like defeating an enemy army then chasing them down to officially kill them. I get it on the field, but for sieges? Sounds too tedious.

Itharus

Vanilla_Gorilla said:

Sieges are crying out for better AI. Maybe not attacking (they might be less fun that way) but defending they need to focus more on the walls, and avoid players shooting them like they're in a fishtank.

Itharus said:

Chokepoint tedium? Seriously? So... you just want field battles on an open field basically.

A straw man is a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent's argument, while refuting an argument that was not presented by that opponent.

I don't want choke points, they take away all the fun and strategy by making the fight easily winnable by using said points.

That's literally half the logic behind fortification... to trap the enemy in chokepoints so you can slaughter them. Sure it makes defending easier -- and makes attacking much harder. Which is kind of how it should be for a siege. Wanting fortifications without chokepoints and kill zones is just... pointless. To make defending harder just script the AI to send 2+ stacks to every siege. Overwhelm the chokepoints, make it costly. Settlements shouldn't just change hands willynilly without cost.

Vanilla_Gorilla

See, when you get into the "just send more stacks" suggestion you start to mess with field battles. Sure, you could have it that it takes 4 stacks to take a city with a stack in it, but then you need to script the AI to build far more stacks, which is fine if you intend the game to be all about siege battles and few field battles, but CA are going the opposite way. Siege battles right now have a force multiplier of around 1.5-2 times to the defender (assuming it's ai vs human)

This begs the question; how do you make siege battles better without impacting sieges?

I suppose you could increase the force multiplier of sieges without increasing the number of troops it takes to take them, but I can't see an option for that which is better for the player.

Itharus

Games prior to Empire required either elite armies or more than one army to successfully siege a defended castle (unless you cheesed the ai). Those games also had quite a lot of field battles. Honestly I feel like since Rome 2 there have been less and less field battles in the franchise, since they started making having more armies penalize you more and more. In M2 for instance I might have 30 armies in the field and have half that many battles in a late game turn. The only downside to that was that it made turns take much longer -- not that I minded, the battles were fun.

Somehow Shogun 2 managed to get a good balance going on. Like, what slipped between Shogun 2 and Rome 2?

Layzan

Someone once wrote here that sieges are are more like a road bump to the enemy, I'm inclined to agree.