By my understanding, at the start of Game of Thrones, summer has been going on for 10 years, this might be wrong though. Anyhow, Petyr Baelish reveals to the small council that Kings Landing has enough grain saved up to last them for a five year winter. This therefore averages at about a 50% surplus.

During the siege of Riverrun the Black Fish reveals to Jaime that they have enough food for two years, so besieging them would be pointless. Looking back it seems obvious that he would have enough food to last them for years. Since winters last for such a long time in Westeros, it only makes sense that each castle would stockpile years worth of food.

So when it comes to a siege, when the main idea seems to be to starve out the defenders, the aggressors would never stand a chance in theory.

So why do sieges work?

5 Answers 5


The simple answer is, they don't work well when winter is coming and people have been recently stockpiling food (like with the hopeless siege of Riverrun). Only commanders as incompetent as Frey & Frey Junior would think a siege was a good idea under such circumstances. Stannis only attempted a winter siege out of desperation, when conditions were bad and getting worse. The Boltons were clear about what his chances were (from SE05 E08):

Our walls have been fully repaired. The gates have been reinforced. We have enough food for six months. We are more prepared for a siege than they could ever be. All we have to do is wait for them to freeze, starve, and mutiny.

However, when seasons are different, and cities and castles haven't recently been stockpiling food, they can work fairly well (but still not very well).

The obvious example is the siege we hear about during Robert's rebellion where Stannis was holed up in Storm's End and were forced to resort to eating rats. It's implied that the only reason they survived was Stannis's strict discipline and stubborn leadership, and Davos smuggling in the onions which earned him his knighthood (from S02 E08):

First we ate the horses. We weren't riding anyway, not with the castle surrounded. We couldn't feed them, so fine, the horses.

Then the cats. Never liked cats. So, fine.

I do like dogs. Good animals. Loyal. But we ate them.

Then the rats.

The night before you slipped through, I thought my wife was dying. She couldn't speak anymore, she was so frail.

And then you made it through the lines.

This was a siege was under ideal conditions: the defenders' supplies were only half full because they hadn't expected the war, and they were defending a completely surrounded enclave deep in enemy territory. Meanwhile, the attackers were led by a well-resourced army (the Tyrells) with easy access to plentiful supplies, in summer when maintaining a camp is relatively easy.

But it still didn't work, thanks largely to Stannis & co's exceptional resilience and Davos's exceptional onion delivery system. This failed siege tied up the Tyrell army for almost the entire war.

Not a great plan.

The only siege I can think of that actually worked anything like the conventional way (by wearing down the defenders) is the Boltons besieging the ruins of Moat Cailin, from both sides, which was defended by fish-out-of-water diseased seafarers whose leader had left them. Even that needed some trickery to actually bring it to an end.

It's also worth noting that in real life, many strategists believed sieges were overrated and a sign of desperation on the part of the attackers. Sun Tzu, for example, in his celebrated Art of War, was particularly scathing:

The superior militarist strikes while schemes are being laid. The next best is to attack alliances. The next best is to attack the army.

The lowest is to attack a city. Siege of a city is only done as a last resort.

...one who is good at martial arts overcomes others' forces without battle, conquers others' cities without siege, destroys others' nations without taking a long time.

(now, why do I think of Littlefinger when I read that last line?)

Something else Sun Tzu discusses which is very relevant to the siege of Riverrun is the importance of controlling the type of psychological position the soldiers are in. In particular:

  • "Dispersive" ground where there is a clear and appealing escape route or alternative that weakens an army's resolve and encourages fleeing, desertion or betrayal (like the situation Jaime puts the Blackfish's troops in, or the situation Robb is in where his bannermen are distracted by amongst other things thoughts of their kin and subjects preparing for winter, with a clear road home behind them)
  • "Desperate" ground where there is no option but to stand firm:

    ...it is the soldier's disposition to offer an obstinate resistance when surrounded, to fight hard when he cannot help himself, and to obey promptly when he has fallen into danger

The army camped around a besieged castle is on something like dispersive ground: bored and complacent, with plenty of opportunities to desert (or, if conditions are harsh, possibly even mutiny). The defenders of the castle are on something like desperate ground, with little choice but to stand firm or die - until a smart strategist presents them with a more attractive alternative (like Jaime does).

It's better to give your enemy a mutually acceptable, tempting way out, than it is to push them into a siege mentality backed into a defensible corner: something we see several of the best strategists in the show doing very well.


Having enough food during a normal summer/winter is different to having enough food during a siege.

Mmmm, foood, arhgghghghg!
In the first instance, people are not condensed into a small area such as a castle. Most people have been growing, harvesting and storing for long periods of time. Having enough food to last years is fairly standard.

In the second instance, clearly a war has been happening (and is no different in this case). Most areas surrounding the lands are scoured and pillaged. Much if the castle is overloaded with men at arms. Having enough food to last two years is actually quite miraculous in the Blackfish' scenario.

Patience my young Padawan
A siege by virtue is a patience game. By blocking out all of an enemy's entry/exit paths, you are essentially cutting off their supply (be it water, food, ale or men). The important part is to make sure that whilst you're blocking the enemy, your own levies and stocks are replenished as needed. In this way, eventually the besieged will run out of supplies. The key is to outlast them.

Watch your rear
The risk to the besiegers is that they have to spread their army somewhat. This is to cover as much area around the castle as possible, so that no-one can get in or out. By doing this, you're weakening your own defensive wall. An unseen army can attack you from behind (as Ned did when he lifted the siege at Storm's End). It can even make you impervious to attack from the besieged as a last hurrah (as Théoden and Aragorn did at the Battle of the Hornburg).

  • More than likely, they had 5 years of food stored up, and now that they're under siege that lasts longer due to rationing, but also shorter because of how many more people are in the castle and because no other sources are coming in (deer, fish, etc).
    – corsiKa
    Commented Jul 7, 2016 at 19:28

Have you ever been near a modern military base? There are no walls, no turrets, just a chain link fence that can be breached with a pair of wire cutters.

Successful generals know how to maneuver their troops and know that maneuver is the key to battle. A castle does not allow for maneuver. A cursory study of Alexander will reinforce this fact. As time passed, artillery made strong points (such as castles) obsolete.

Although they died a slow death. As late as 1939 the Belgians invested massive amounts of money building the perfect fort in an effort to keep the Germans from invading France via their country. Yet in about 24 hours 500 men destroyed the strong point. You can read about Fort Eben-Emael here.

It's one of my problems with GoT that was reinforced in S6E10. How can a society with such learning be so backward in regards to warfare, peace and politics? The maesters or at least someone who did partial maester training should be among the most valuable personal in the kingdoms.

A smart lord would send a few people to the citadel to be trained in specific areas, and leave when they gained that knowledge. Sam had use of the library, and I am sure others could be granted that access as well. One great addition to a house would be someone trained in siege warfare and mathematics. They would be handy in the circumstances you describe.

Utilizing technology that has been represented so far, one could take a castle with a three or four barrels of wildfire and a single catapult. The threat of such an attack might be enough to "lower the drawbridge".


So why do sieges work?

A siege can be formally analyzed using various methods, just like many real-world military initiatives. And I would go further to say that these approaches are quite valid for our fictional world of Westeros. As a retired US Army officer, I can assure you that these forms of analysis are done by commanders in the field and by academics who ask themselves why a military operation (not unlike a siege in Westeros) succeeds or fails. Institutions like the US Army War College formally teach the projection of power, and some of these concepts can be used to good effect here.

One classic model is to analyze using the Nine Principles of War as described in US Army doctrine (FM 3-0 Operations): Mass, Objective, Surprise, Security, Maneuver, Offensive, Unity of Command, Simplicity, and Economy of Force (MOSSMOUSE). Most successes against a superior force can be explained by the masterful use of at least one of these factors. Most disasters against an inferior force can be explained by the violation of at least one of these factors. So let’s see where we get in a Westeros siege while referring to Army FM 3-0 Operations, Appendix A.

Maneuver. Place the enemy in a disadvantageous position through the flexible application of combat power. [Emphasis mine]

This gives a KEY advantage to the attacker. For the US Army, loss of maneuver is very, VERY bad! A scaled down version of this sentiment will flow to the defender in Westeros. He has lost the ability to affect very much outside his walls. Meanwhile, the attacker has the ability to continue operations with forces that are not directly needed for the siege.

Objective. Direct every military operation toward a clearly defined, decisive, and attainable objective. [Emphasis mine]

This will be an advantage to the attacker because he gets access to everything outside the siege walls, which can be a lot. Additionally, this constitutes success IF he has an overall campaign plan that uses this as a step along the way. This is a clear disadvantage to the defender because his objective has now been lowered to “do not die” (as opposed to something vital like protect his realm).

Offensive. Seize, retain, and exploit the initiative.

This will favor the attacker, as long as the defender is forced to stay under siege.

Unity of Command. For every objective, ensure unity of effort under one responsible commander.

This will favor the attacker. All forces loyal to the defender have their senior leadership cut off, with the small exception of messenger birds that routinely get shot by arrow as they leave the castle under siege.

Mass. Concentrate the effects of combat power at the decisive place and time.

This will favor the defender in Westeros. Without the means to overpower the enemy in a siege, the attacker would be violating this principle – if it were an end to itself (more on that later). Without siege engines (catapults, trebuchets, battering rams, etc) and sappers (military engineers), this may be a wasted effort for the attacker in the long run.

Simplicity. Prepare clear, uncomplicated plans and clear, concise orders to ensure thorough understanding.

This will be an advantage to the defender. In Army-speak, the defender has interior lines meaning all his actions to survive are in one place (short lines). The attacker will have exterior lines (his logistics will be longer, and relatively more complex).

Economy of Force. Allocate minimum essential combat power to secondary efforts.

This will be an advantage to the defender. His walls will do most of “his talking”.

Surprise. Strike the enemy at a time or place or in a manner for which he is unprepared.

This does not favor either side, unless the defender can perform operations outside the walls.

Security. Never permit the enemy to acquire an unexpected advantage.

This does not favor either side, unless the defender can perform operations outside the walls.

Recap. Even if the attacker does not win in the short run, he has a net gain.

Advantages to the attacker:

  • Maneuver – this is HUGE
  • Objective – this is Big
  • Offensive
  • Unity of Command

Advantages to the defender:

  • Mass
  • Simplicity
  • Economy of Force

Partial Answer #1: The Principles of War favor the attacker if he is able to execute it. Please note that taking over the castle does not need to be the goal of the siege. Tying down the defender can be all that is needed in order to choose to do a siege. Please see below.

Example: Stannis was advancing on Winterfell in order to put it under siege. The Boltons knew this and decided to attack the attack (a sound defensive strategy). Twenty riders went in at night and set fire to crucial materiel that crippled Stannis’ ability to proceed. The Boltons showed masterful execution of Surprise and Simplicity, while Stannis showed woeful execution of Security.

Key Idea. What are you trying to gain with a siege? What is the end state? There’s more to fighting than a single battle. The overall geo-political context is quite important. At the end of the day, it’s diplomacy (or negotiations on a smaller scale) that will end the fighting with a good result. Combat is an extension of that (diplomacy by other means as they say). “I want to be King” is usually the answer in Westeros (though not the case with Jon Snow).

Below that, we have a campaign plan. This is what should decide whether a siege is called for – not the fact that someone can simply do it. What is the campaign plan? Answer this, and the necessity for a siege will be self-evident. What are the steps that must occur to be able to win at the final negotiations? The need for a siege should be justifiable only when it becomes “one of the moves on the chess board”. By the way, the Principle of War are applicable at the strategic level as well.

Partial Answer #2. The ultimate desired end state will drive the campaign plan, that then answers the question of whether a siege serves that purpose – and whether the siege must result in the destruction of the forces and aristocrats inside. Will the siege work if we do not break down the walls? The end state and campaign plan will answer that.

Example: The Boltons needed all of Winterfell. Getting inside those walls was a required step for their end state.

I’ll end this with one way to determine how you decide what to do in the campaign (that drives whether a siege is needed and how far it needs to go).

Centers of Gravity (COG) FM3-0, Paragraph 7-30: Those sources of power that provide moral or physical strength, freedom of action or will to act [fight].

Find and defeat your opponent’s COGs and you will own him. Likewise, you must protect yours. These two actions will help define your plan. The seminal book, Understanding Centers of Gravity & Critical Vulnerabilities by Dr. Joe Strange (no relation to the other Dr. Strange) lays down exactly how to do this. (By the way, his doctrine bears great applicability to modern business operations.) He builds on Clausewitz’s universally and globally studied book On War.

What makes your opponent effective as an opponent? That’s the COG (or one of them).

What are the Critical Capabilities (CC) that makes them COGs (the abilities that make him dangerous)?

For each of those, list the Critical Requirements (CR) that are essential to enabling the CCs (For CC A, we have A1, A2, A3: those things that are needed to make the danger happen).

Find the Critical Vulnerabilities (CV) that will neutralize the CRs.

Partial Answer #3: Use intellect to determine how best to use your forces. Execute a siege if it furthers that end.

Example: One of Stannis’ enemy COGs would be the Lannisters and the Queen in particular. While they exist as an organized clan, they will prevent Stannis’ success. A COG = the Lannisters.

One of the Critical Capabilities of this COG is gold: “A Lannister always pay his debts” is a family motto whose violation would make them the laughing stock of Westeros. A CC = gold and their ability to pay debts.

Two Critical Requirements are the Lannister gold mines and overseas credit. these are what's needed in order to make the Critical Capabilities possible.

The Critical Vulnerabilities are the security of the Lannister mines and the “credit rating” of the Crown. Stannis may have been attempting to siege the wrong place and the wrong idea. If it were possible (it’s not explained if it is), Stannis should attack/siege the Lannister gold mines. Additionally, he should figuratively siege the Crown’s credit rating by possibly negotiating something with the creditors. (This example is a stretch whose only purpose is to illustrate how COG analysis would work.)

[If you’ve read this far, you may need a nap, and you should not drive a car for a while.]


I'd disagree with some of the other answers. They absolutely DO work if the ultimate goal is to take the castle/city and there is no urgent time factor involved.

The Lannisters were not involved in any other wars/campaigns, so the Blackfish being able to wait him out for a couple years is not an indication that Jamie would have been thwarted, just that he would be a pain in the butt to the Lannisters in the process.

While those inside would slowly starve, those outside would be fed and relatively unmolested.

Also, it's not as if, once inside and delving into the food stores, that those outside have no choice except to wait out those inside. Ramsey Bolton notes that he could hold out indefinitely against Jon Snow because Snow's army was not equipped for a siege. However, an army that doesn't have fear of being hit by raids or battle can bring siege engines (catapults, trebuchets, battering rams etc) and can put their engineers to work digging tunnels to undermine the walls.

Think about Lord of the Rings, when the army of Mordor was besieging Gondor, and they were flinging chunks of old ruins into the city, and bringing the gate-busting "Grond" to bear to move the process along.

In the Song of Ice and Fire books, when Mereen is under siege by the slave masters, there is some kind of plague/disease that runs through the camp of the army besieging the city. They promptly take those bodies and catapult them into the city, infecting their enemies who have nowhere to go and no resources available to avoid the spread of the illness.

There are a TON of methods available to make life miserable and inflict suffering to those within. If you want a quick victory, a siege isn't going to work. If you have time and resources and have vanquished all foes outside the city, it's a pretty safe and certain way to go.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .