Spencer Ackerman, the well-known national security blogger, recently posted an article criticizing Robb Stark’s military strategy in the War of Five Kings. In the piece, Ackerman argues that:
But the Young Wolf is a case study in the difference between winning battles and winning wars. Robb is an excellent company commander, leading from the front and inspiring his men with both his bravery and his battle prowess. He’s also a terrible general…
Robb’s vainglorious uncle clearly messed up by disobeying orders to hold Riverrun, preferring instead to stop Clegane’s army at Stone Mill from crossing the rivers of the Trident and heading west. Robb rolls his eyes: he wanted Clegane to come west, so the Mountain, who “doesn’t have a strategic thought in his head,” would have been lured unsuspecting toward the eastward-marching Stark forces and killed. “Instead,” the King in the North laments, “I have a mill.”
News flash, Your Grace: Clegane is not worth much more than that mill.
While I think his article does have some important points, I feel that the piece fails to grasp the larger strategic and political environment informing Robb Stark’s military decisions and as such comes to an overly negative conclusion.
A Point of Agreement:
Let me start by saying that I agree entirely with Spencer’s criticism that the show’s handling of Robb Stark’s story line (while perhaps necessary from a budgeting standpoint) “doesn’t make it easy…the gorgeously rendered map displayed during the opening credits never shows you the Westerlands, an omission with inadvertent implications…Without a sense of the terrain, you can’t really understand Robb’s war.”
The War of Five Kings is an incredibly complex war, as each of the different camps’ military strategies constantly forces the others to readjust their own in response to unforeseen events, and it’s especially difficult to understand what’s going on when the show elides events due to the narrative economy imposed by budget restraints. Case in point: between the end of Season 1 and the beginning of Season 2, Robb makes a trip to Riverrun that wasn’t shown in the show (in order to obviate the need for new locations and casting actors who wouldn’t be prominent until season 3) that dramatically changes the political and military imperatives for the Northern war effort.
Which brings us to where I think Ackerman’s missing the bigger picture.
Where Did Robb Stark Stand at the End of Season 1?
In the books and in Season 1, the War of Five Kings begins with the Lannisters in a commanding advantage. Following the capture of Tyrion Lannister, Tywin Lannister begins raising troops at Casterly Rock and sends out Gregor Clegane to raid across the Riverlands, initially hoping to draw out Ned Stark and capture him so as to revenge himself for the insult to the Lannister name and gain a captive for an exchange, and to distract his enemies while he raises a force of some 35-40,000 men at Casterly Rock. However, Jaime’s wounding of Ned Stark forces the Hand to send out Beric Dondarrion, Thoros of Myr, and 120-odd men after Gregor Clegane, preventing that plan from coming to fruition.
Tywin adapts to this change in plans in two ways: first, together with his lieutenant Gregor Clegane, he ambushes Dondarrion’s Men at the Mummer’s Ford near Pinkmaiden (2 on the map), decimating this force and forcing them to turn guerrilla fighters and become the Brotherhood Without Banners; second, he splits his forces in two and puts Jaime in command of half of his men as the Lannisters invade the Riverlands in force. Tywin sweeps through the southern Riverlands (2-4 on the map), taking the critical castle of Harrenhal which commands most of the southern Riverlands and the key north-south route to King’s Landing; Jaime overruns an outnumbered Riverlands force at the Golden Tooth (1 on the map, a castle that commands the pass that marks the border between the Westerlands and the Riverlands) and then succeeds in a surprise attack that smashes Edmure Tully’s forces mustering at Riverrun (3 on the map), although enough men manage to make it inside the castle to hold the castle, which Jaime places under siege.
This is the strategic situation facing Robb in Episode 8 of Season 1: his enemies have split their forces, such that Robb’s 18,000 men (plus the 4,000 Freys he will soon acquire) more or less equal either army, but if they combine, they outnumber him 2:1. If he attacks Tywin’s force, Riverrun might fall, ending all hope of assistance from the Riverlanders; if he attacks Jaime’s force, Tywin can march up the Kingsroad and cut Robb off from the North, while fresh reinforcements from the Westerlands threaten him from the other side. Robb’s decision is both characteristically bold and strategically brilliant: he divides his own army and attacks both forces, march-blocking Tywin’s army at the Battle of the Green Fork of the Trident (3 on the map below) so that it can’t help Jaime while capturing Jaime at the Battle of the Whispering Wood (4) and destroying his army at the Battle of the Camps (5). (Notably the show departs from the books by changing the nature of the Battle of the Green Fork from one in which Robb gambled his 16,000 foot against Tywin’s 20,000 to a sacrifice play where Robb sends 2,000 men)
As a result of these battles, Robb’s army combined with the resurgent Riverlanders now outnumbers Tywin 2:1. Tywin being no fool retreats to Harrenhal, where the strong defenses of the castle neutralize Robb’s numerical advantage, sends Gregor Clegane with 500 men to burn the Riverlands, and calls for fresh troops to be raised in the Westerlands so that he can regain the numerical upper hand. So far, Ackerman and I are in agreement.
Here’s what happens next that Ackerman doesn’t include in his analysis: in relieving the siege at Riverrun (and freeing his uncle Edmure from temporary captivity), Robb Stark is hailed as King of the North…and the Riverlands. This changes the political and strategic imperatives of the new Stark-Tully alliance completely: not only does Robb have to “attack the Lannisters’ home turf until they sue for peace — and acknowledge northern independence,” he also has to protect the Riverlands because the lords of the Riverlands now make up half of his armed forces. Robb doesn’t really have the option of retreating to the North – he’s leading a coalition army that’s half Riverlanders who will not follow a king who abandons them to the enemy (which would have an impact far worse than breaking his promise to the Freys).
The second thing that happens is that Edmure makes his next blunder of the War after the disasters at the Golden Tooth and Riverrun: he lets most of the Riverlands Lords go free to retake their lands from Gregor Clegane, which scatters much of the 20,000 Riverlands forces as they take and retake Raventree, Stone Hedge, and Darry in inconclusive hit-and-run warfare.
That’s the true strategic situation Robb was dealing with at the end of Season 1: he slightly outnumbers Tywin, but can’t hit him at Harrenhal (Ackerman doesn’t really take into account the importance of castles as defensive force multipliers), he’s slowly retaking the Riverlands but at the cost of dispersing half his forces, and there are new armies being raised in the west. Thus, when he invades the Westerlands in Season 2 (or A Clash of Kings), he’s not just trying to “attack the Lannisters’ home turf until they sue for peace;” he also is trying to prevent himself from being flanked and outnumbered, take the fighting out of the ravaged Riverlands, and capture supplies to feed his armies (since most of the war has been fought on the Riverlands, it’s been stripped clean), and looking for a broader strategic victory.
Analyzing Robb’s Western Offensive:
In addition to a more complex strategic situation, I think Robb’s offensive has more strategic merit than it’s been given credit for. Ackerman argues that “even in A Storm of Swords, Robb is hitching his hope for the war on a masterstroke, which is terrible wartime leadership. He’ll win — if Tywin pursues him west; if the ensuing battle breaks his way; if Stannis wins at the Blackwater,” and that “lay[ing] siege to Casterly Rock and Lannisport with the additional strength of Riverrun, bleed[ing] the west… [is] a last-gasp plan — the North and the Riverlands may not have the manpower; and if Robb loses, he dies — but it has the benefit of taking the west away from Tywin, Joffrey and Cersei. The smarter plan is to retreat to the North.”
This doesn’t really get at what Robb was doing, and fails to recognize the nature of feudal politics.
In attacking the West in Season 2, Robb sought to further his own strategic objectives (preventing himself from being flanked, resupplying his forces, taking the fight out of his own territory) while forcing a political and military Hob’s choice on Tywin Lannister: either Tywin comes West and risks the loss of King’s Landing, or he rescues King’s Landing and risks the loss of the Westerlands and Casterly Rock itself.
As we have seen in Seasons 2 and 3, the loss of one’s capitol city is politically devastating, especially in a feudal context. The armies of House Stark and Lannister are not standing professional armies; they are made up of bannermen who serve their overlords because the overlords offer them protection from outside invasion and can threaten them with retaliation if they betray them. When Robb Stark lost Winterfell, it showed that he couldn’t protect his own home or the homes of his bannermen and undermined his position with his vassals, above and beyond the issue of the loss of resources and reinforcements. Losing Casterly Rock would have done the same thing to Tywin Lannister; the mastermind of the “Rains of Castamere” inspires fear and respect from his bannermen, but none of the love that still inspires Robb’s men to fight for the memory of Ned Stark. Without his aura of invincibility, without the ability to call up fresh troops, without the gold in Casterly Rock to pay for the war effort, Tywin’s army would have melted away like snow.
The same holds true for King’s Landing: if King’s Landing falls to Renly or Stannis (and Renly and Stannis have to make for King’s Landing, since it’s the center of all political power) while Tywin is away, he loses his daughter, Tyrion (again), and the grandsons and granddaughters who represent his House’s claim to the Iron Throne. He also loses half of his territory and instead of holding the entire middle of Westeros and keeping his enemies from combining against him, now faces being backed into the Westerlands and destroyed. Politically, he ceases to be the Hand of the King, putting down rebels and traitors to King Joffrey, and now becomes a rebel and a traitor to the rightful Baratheon King.
Far from blindly gambling, Robb has weighed the odds with a keen eye to the realities of Westerosi politics, and as a result is proved right: Tywin does take the bait and come west, and had not Edmure blocked his march, King’s Landing would have fallen. Moreover, Tywin placed himself in real danger by marching west and trying to cross the Red Fork with 20,000 men – had he succeeded in crossing, he would have been outnumbered and surrounded, with Robb Stark’s 6,000 men in front of him, Edmure Tully’s 11,000 men to his right flank, and 10-12,000 Northmen under Roose Bolton’s command behind him.
Even with Edmure’s screwup, Robb Stark’s offensive showed substantial results: at Oxcross (shown in Episode 4 of Season 2), Robb Stark wipes out an army of 10,000 men who were mustering on his western flank; this victory is followed up by battles at Ashemark and the Crag which protect Robb’s rear and right flank while ending Lannister resistance in the northern half of the Westerlands (which in turn means that there isn’t really a force in the west that could threaten his army at the moment, contrary to what Ackerman says). In quick succession, Robb’s bannermen seize the gold mines of Castamere, Nunn’s Deep, and the Pendric Hills (simultaneously improving their finances and cutting Tywin off from his source of funds), raid up and down the northern coast, and bring back thousands of head of cattle to feed a hungry Riverlands. In strategic terms, Robb’s “scouring” of the Westerlands eliminates Lannister numerical superiority (which in turn forces Tywin’s hand), conquers half of the Westerlands, and puts him within a week or two of Lannisport and Casterly Rock.
Analyzing the TV Version:
I agree that great violence is done to this strategic vision by having Robb be at the recapture of Harrenhal (in the books, Roose Bolton is simply ordered to retake the castle once Tywin marches west, to cut him off from behind), and by changing the strategic significance of the Battle of the Red Fork to a fight between Edmure and the Mountain – although I think the bigger mistake was not doing any foreshadowing of the Battle of the Red Fork last season, which makes it seem like it came out of the blue and was unimportant. (It could have been very easily done as a fakeout in Episode 8 when Tywin marches out of Harrenhal with a quick scene where a message shows up at Robb’s camp in the West with news of a battle at Riverrun, filling Robb with false hopes that his stratagem is succeeding).
However, I think Ackerman is wrong when he says Robb taking Harrenhal is “inexplicable” and that “Clegane is not worth much more than that mill”:
- Taking Harrenhal isn’t stupid. In itself, Harrenhal controls the southern Riverlands and any approach from King’s Landing to either the North or the Riverlands, and is a huge defensive force multiplier. By taking the castle, Robb has a much better chance of preserving his territorial gains against a numerically superior opponent. We often forget, given its dilapidated appearance, that Harrenhal is a valuable military asset in its own right, but we shouldn’t.
- Eliminating the Mountain isn’t stupid. Gregor Clegane isn’t just “one admittedly mountainous henchman;” with Jaime gone and Kevan Lannister in King’s Landing (although we haven’t seen him this season thanks to narrative economy), Gregor is Tywin’s chief general in the Riverlands. We’ve already seen how Tywin relied on Gregor in the opening act of the War of the Five Kings; Gregor was Tywin’s hit man when it came to wiping out the royal family during the siege of King’s Landing; and Tywin will go to great lengths to keep the Mountain as a military asset even when there are compelling political reasons to give him up.
- It wasn’t just about Clegane. Leaving aside his person, I think the showrunners made a crucial mistake when they left Clegane with just a garrison command of Harrenhal with no more explanation – if that was truly the case, there’s no way in hell he would be marching 300 miles to Riverrun with Robb’s army in the neighborhood, or could have lost multiples of the 208 men that Edmure lost in the TV version of the Battle of the Red Fork. I think a scene was missed somewhere where the Mountain was put in charge of a larger force, which must have been at least a few thousand men (given that Edmure’s defense suggested at least a 2:1 casualty rate, which means the Mountain lost at least 400 men, and unless we’re assuming a more profound defeat than the scene suggests, he had to have 4-5 times that many men under his command). Given that Robb’s strategy at the mountain requires him to “take the fight to the enemy” and inflict the kind of lopsided battlefield losses that could even the odds against the combined Lannister/Tyrell forces, the chance to bring down their numbers by around 2,000 men isn’t a bad idea.
If Ackerman wants to criticize Robb Stark’s skills as a commander-in-chief, I think there are ample grounds for doing so. Edmure acted contrary to his orders, but that could have been avoided had Robb explained his battle plans to all of his subordinates. Had Robb Stark not married Talisa Maegyr, he’d have 4,000 more Frey soldiers at a time when he needs every man. Had Robb not let Edmure discharge the Riverlords after the Battle of the Camps, he’d have had an army of 40,000 men concentrated and capable of taking Casterly Rock in one go. And for the sake of the old gods and the new, you don’t ever, ever trust a Bolton.
But Robb’s attempt to lure the Lannisters into the west where he could defeat them on the field isn’t one of them.