Please register for Total War Access to use the forums. If you're an existing user, your forum details will be merged with Total War Access if you register with the same email or username. For more information please read our FAQ’s here.


Let's brainstorm how to make Sieges Epic in Historical Titles

jamreal18jamreal18 Senior MemberRegistered Users Posts: 12,504
edited September 1 in Total War General Chat
Sieges historically took days and couldn't be won with 1 day only.

What are your ideas on how implement it in Total War?

Let's help CA. What are your suggestions?


  • CommisarCommisar Senior Member Registered Users Posts: 1,949
    Would expect assaults to still end in a single day as we end turn and months fly past so it's messed up either way.

    Main issue I tend to find with sieges is the defences are quite often worthless. Attackers tend to be able to get on to the walls quicker than the defender can get on or off them, possibly pathfinding issue with the limited staircases.

    I think Shogun 2 had the best sieges in recent titles although the AI still failed to defend them effectively. Part of that would be the layered defences as you upgraded the castle, so if the enemy captured the outer layer you could still defend the second or third. Was quite interesting moving my units back and choosing which to sacrifice while my others got in position.

    Oh and of course not having siege equipment be standard issue battlefield items would make sieges and assaults take more time and resources, although that could also be seen as annoying.
  • jamreal18jamreal18 Senior Member Registered Users Posts: 12,504
    edited September 2
    Siege Engines:
    No Siege Equipment, the siege cannot progress.
    This is why simple ladders that can be picked up by anyone shall return. Replace pocket ladders with ladders from Medieval II so it's not that easy to overwhelm the defender with superior numbers.

    Last Stand:
    Make the Siege of Capitals have its army constantly defend the city so there are always 40 units defending and attacking. Make it epic.

    Make the Buildings part of Battlemap:
    Change the inner walls so that skirmishers can be placed on them. There is really no reason why that shouldn't be possible. Capital Cities must have garrison-able buildings for defenders. It's their capital so they should be able to utilize some of its buildings for ambushing attackers.

    As Attacker: Before you assault Capital Cities, you must be prepared and must have multiple armies assembled to attack them. One army shall not be enough or it should be hard. You must have more armies ready to deploy. You are attacking CAPITAL CITY though.
    As Defender: You must be ready to hold until reinforcements arrive if ever there will be.

    Ammunition Replenishment:
    They must have a place somewhere in the city for re-supply of ammunition. Defenders must protect it or the attackers will steal or burn it.

    Day and Night Cycle:
    As siege progress, the time shall also progress.
    *Ancestors Legacy have this feature.

    Units Merging
    Depleted Units of the same type shall be able merge in battle. Giving more room for reinforcements and to make them more effective.
    *Praetorians have this feature


    It should be applied to sieges especially Capitals though. Capitals are the heart of empires/factions except Nomads. So it should be a challenge to take them.
    Post edited by jamreal18 on
  • TejaSchwarzhaarTejaSchwarzhaar Registered Users Posts: 309
    General approach

    My general approach is the following: I will take a potential Medieval 3 as the basis for what I write, because this would be the perfect scenario for all kinds of siege battles.

    First of all, I think the separation of castles and cities should return to reflect the medieval reality. When I speak about both of them together, I will use the term "fortifications".

    Cities should be economic and cultural centres, with a large population, many buildings and a lot of income, the possibility to build universities etc.
    However, they should be relatively hard to defend. They should normally be on relatively flat terrain and only have one or two layers.

    Castles however shouldn't generate wealth, they are there to control the land. They should be something you have to take and nothing you want to take.
    You should need a giant number of men to capture them and a lot of preparation.

    Campaign surrounding the siege

    The most important thing to me is that sieges should be risky for the attacker. The player and the AI shouldn't be able to completely plan how the siege goes. There is always stuff that can happen which makes you abort the siege, here are some possible examples:

    Your general dies, leaving the army leaderless.
    You run out of supplies or money making the army revolt.
    Smallpox break out, killing 20% of the army and the rest will be unable to fight, so there is a risk that you get slaughtered in a sally-out.

    All of these will affect how your siege goes and in the worst case force you to retreat. The likelihood of something like that happening increases with the amount of men you have and the amount of time you have spent sieging. YOu don't want to have 5.000 men sitting in a dirty and unhealthy sieging camp for years.

    A high amount of defenders should also have negatives, like supplies running out faster and the likelihood of deseases rising.

    A siege needs preparation.
    You need to build siege equipment (please, remove those damn pocket-ladders!), starve the enemy out and wait for good weather, your men won't be able to push a battering ram through thick mud!

    Castles should force you to take them, maybe beacuse they block strategic paths or because guerilla troops coming from them ruining your supplies or just because there is a small army in there that could sack your land if you don't destroy it.

    Some good mechanics to return here would be army supplies from ToB and 3K, the necessity to build roads and severe penalties to marching through land without roads.

    The decisions that the player makes on the campaign map should be reflected on the battlefiled: If there is a reinforced gatehouse, a watchtower, a fortified palace or a ditch around the fortifications, these should not just give bonuses but actually be there on the battle map.

    Lastly, it should be possible to capture parts of the fortifications, like a gate house area of a city and then continue the siege on the campaign map, creating a possibility to rest your troops, get reinforcements and have negotiations about a capitulation. This also reflects the historical reality, i.e. in the siege of Harfleur.

    Real-time battles

    The strenth of a garrison should not so much be in their numbers but in the situations they get to fight in. It was possible to defend castles against enemies that were 20 times stronger in numbers. This dimension would probably be a bit exaggerated in a Total War game, but defending a castle with 500 men against 2500 should be possible if the attacker isn't well prepared. And for that it shouldn't be necessary that the AI glitches out and sends 2000 men into a unit of pikes, which is basically how I won my Sparta campaign in Rome II on very hard when I was an absolute noob.

    Cities should have their outer walls and gatehouses as their main defense. Inside, it should be pretty easy to take with flat streets and maybe a last stand position. Still, the walls should be really hard to take requiring siege towers, rams, catapults, so at least two or three turns of preparation even if it's a weak city. This obviously shouldn't be required for unfortified settlements. (By the way, these should definitely exist to create some diversion from just siege after siege.)

    The attacker should probably need about twice the strength of the defender to take a city, even if well prepared.

    In castles, things get more difficult. Many areas of the walls should be unattackable, whether it's a water-castle or a castle on a hill. On the attackable parts there should be some gatehouses creating the weakest points because a ram can be used there, but the defender should be able to take precautions like boiling oil, stones and stuff like that.

    Generally, a castle should be very hard to take. There need to be two or three layers, maybe even more in big castles and the whole design should be made to be able to defend the castle.

    However, there can't be as many supplies in a castle as in a city, so the defender should be rather limited in terms of numbers. Even a big castle won't be able to supply 1000 men in a Total War game for a long time, so if you don't want to get starved out: Don't pack your castle full with troops.

    If the castle is well defended, the attacker should need about five times as many men to take the castle in an assault.

    But most importantly:
    The AI needs to start working: No clumping up in choke-points, no just defending the walls, no attacking without equipment, no suicidal cavalry charges, no completely unprotected archers etc.!

    Now, I already hear you saying: The AI won't be able to make it!
    And you are problably right, but I expect CA to be able to make the AI better, this has been the main complain from the community for ten years or so, so please, CA, use more of your resources on that!

    Until then, I will appreciate every upgrade to the AI, but I think that it's fair to request that the AI stops failing at choke points the same way it did ages ago.

    One last very important point in this section: I don't think that this would eliminate land battles, because it would be too risky to keep putting your armies in fortifications. They could get surrounded really easily and have attrition really quickly because there are not enough supplies. The AI obviously has to realise that too.

    In general, it should be balanced that way, that a fortification with a slightly reinforced garrison should be the perfect mix of being able to put up resistance in a fight and being able to be supplied for a long time until reinforcements arrive.

    If that balance can be created, it makes sense to only use your bigger armies in field battles, offensive sieges und as reinforcements for besieged fortifications.


    Regarding capitals, the first thing that needs to be noted is that they should normally be big cities reflecting the reality in medieval times. However, this makes them harder to defend creating the necessity to give the defender some options to make the defence easier.

    This could be done with unique buildings, i.e. one that gives a palace guard to the garrison or a King's Keep building which creates an extra layer of defense that is perfect for a last stand by the defender.

    I just barely touched many areas because I wanted to keep this as short as possible, but I hope you got an impression of what I mean and I'm happy to discuss it!
  • WarlockeWarlocke Senior Member Registered Users Posts: 3,722
    Add a Mythic DLC.
  • IchonIchon Senior Member Registered Users Posts: 5,546
    edited September 22
    1. Siege escalation so that it is possible to storm the walls immediately but with few ladders and no siege artillery, however at any point after their first unit is routed, the attacckers can choose to stop the siege and hold 50% of what they have gained. IE- they hold 2 towers and 3 sections of wall, they would lose 1 tower and 2 wall sections but on turn 2 there would be an automatic siege escalation that might give a slight advantage from controlling part of the walls already + if they reduced the number of defenders by 2-3 units and only lost 1 routing unit they get a more beneficial attacker vs defender ratio reducing the number of turns until the defenders surrender without having to risk their full army.

    2. Automatic AND event driven Siege Escalation decided by 4 factors:

    A. length of time
    small siege escalation each turn with events and the ratio of attackers to defenders affecting the size of the escalation modified by defensive structures and commander traits

    B. ratio of defenders to attackers with the best results for the attackers at 2:1 or better while defenders get better results in siege escalation and random events if the ratio is less than 1:1 IE- siege at 1:1 takes 5 turns until surrender, with 2:1 ratio for attackers the siege will take only 3 turns until surrender while if the attackers are fewer than the defenders the siege will take 7 turns until surrender.

    C. traits/skill of each commander, commanders with terrifying or rich reputations might get bribery events while commanders with merciful or pragmatic traits get political cost events (-authority, -PO, -replenishment, etc)

    D. random events for both sides on the same turn with costs attached to each of 1 of 5 outcomes ( 1 large benefit to the attackers, 2 slight benefit, 3 neutral, 4 slight benefit to the defenders,5 large benefit to the defenders ).

    If for example the Attackers elect to pay the highest price, of either gold, units, negative trait for the commanding general or Faction Leader and the Defenders pay nothing then the Attackers get the full result such as siege escalation moving the surrender timer from 4 turns to 1 turn, if the Defenders also pay the highest price the result is neutral.

    If the Attackers pay for a slight benefit and the Defenders pay the highest cost for a large benefit they get a small benefit, such as disease striking the besiegers camp and 5% attrition for the next 2 turns (which could lower the ratio from 2:1 to 1.8:1 and in a siege vs a capital with full defensive structures that was going to take 4 turns to reach surrender will not take 6 turns.

    3. defensive structures based on the population/type of settlement being attacked (3K did this) with small settlements with limited defensive structures being relatively easy and not super different from open field battles to the faction capitals with full defensive structures having raised platforms and multiple layers to the attackers have to breach.

    4. besieging automatically counts as Raiding but with the difference that Raid income decreases per turn. It never made sense that besieging the center of population in a region generates 0 income for the attackers despite putting them at greatest risk by being near a garrison army and subject to sally attack + reinforcements.

    The main reason players don't wait for siege escalation and find every exploit to break the AI siege defense is because of the opportunity cost of having an army tied up for several turns laying siege when they could gain the regions income + automatic replenishment often will give better results to take 40% casualties in a siege assault and get 32%replenished in the next 4 turns than wait 4 turns and end up losing 4 turns of the regions income and having a 100% vs 92% strength army on the 5th turn.

    5. regions which surrender give 100% of their income on the 2nd turn after being captured. regions which are captured by siege assault give 33% of their income on the 2nd turn increasing by +11% per turn so it will take 8 turns to reach full income so their is a further economic advantage to surrender vs capture. Replenishment is for surrender is 50% of normal for 2 turns while replenishment for capture is 50% for 6 turns to reflect units are required to keep an eye on the sullen inhabitants or actually do some of the grunt labor themselves while new colonists arrive to fill the roles of those who fled the siege or died in the assaults.
    Post edited by Ichon on
    YouTube, it takes over your mind and guides you to strange places like tutorials on how to talk to a giraffe.
  • Lotor12Lotor12 Registered Users Posts: 609
    jamreal18 said:

    Sieges historically took days and couldn't be won with 1 day only.

    What are your ideas on how implement it in Total War?

    Let's help CA. What are your suggestions?

    I just finished the book about siege of Ujvár (Nové Zámky) in 1663

    I can say, that Sieges are implemented right in TW,

    Siege itself took days/weeks/months and this is represent by turn based system on campaign map - siege for 3 turns or 10 turns;

    At the same point, attackers/besiegers try for a assault, and that is represent as Siege Battle in Total War

    I would focus on supply mechanics introduces in Thronnes of Brittania (and later used in 3K) , the supplies played critical role in besieging,
Sign In or Register to comment.