Week 2: The March and Siege

Sabre fencing favours the attack. But that doesn’t mean a sabre fencer should just berserk rush off the line. It is often better to trick the opponent into falling short, gain priority, and execute your own attack down the piste.

I used to think that being the attacker with priority is easy. And, indeed, it is the easiest phase of a sabre bout. But unless the attacker is much taller and faster, the defender can do a lot to thwart the attack. They can beat the blade, parry, make fall short, and counterattack — especially against shorties like me. An attacker cannot simply close distance and hit.

I hit this particular wall when I left the novice leagues. Until then, my natural speed and agility had sufficed to catapult me through my opponent’s defences. They might be taller but I was faster, so fast, like an F-16 eating up a Tupolev. I could accelerate so forcefully from a standing start that I could literally bowl my opponents over, them tripping and falling to flee my rush, then butt-planting on the piste.

It was only when I started facing people who were not only taller, but taller and faster, that I realised that this whole attacking business was not as easy as it seemed. These guys could pull away. These guys could counterattack. These guys could hold me at bay, like a schoolyard bully holding some kid’s head at arm’s length. These guys could bowl me over too.

So I asked my grandcoach what to do. He was one of those Soviet types for whom words like bear, brilliant, grizzly, hewn, mad, metal naturally trundle to mind. Battle tank, not fighter jet. Scorn, he had, for tricks and speed. Better, he said, to confront and control.

He envisaged the attack as a medieval siege. The attacker was an army on the march, trampling down the fields in pursuit of the fleeing defender. The army must not rush the march. A fleeing defender was still dangerous, and might be setting a trap besides — an echo, perhaps, of my grandcoach’s latent genetic memories of Mongol and Parthian battle tactics, inherited from his Belarussian forebears.

There was no rush anyway, because the defender has to stop eventually. There they must make their stand. Then the attacker can lay siege. The attacker must be patient. They must dismantle the walls and gates and moats that the defender hastily erected in their path. They must starve out the defender, drive them mad with impatience, bat away the petulant counterattacks and sallies that serve only to deplete the defender’s dwindling reserves. They must wait for the breach.

Then the charge. Shields up. The attacker must crush the defender, through whatever obstacles the defender can muster up as a last resort. That is, provided the defender didn’t flee again. No matter. March again and lay siege. And again. Until the defender refuses to flee any further, or can do so no more. There is only so far they can run.


In sabre, a defender who is balanced and ready is dangerous. The danger they pose extends into the space in front of them for as far as they can reach with their blade — the ‘danger zone’. The attacker is safe when they are outside the danger zone, but must go across it to hit.

First the attacker has to approach the edge of the danger zone. The attacker can simply advance, or use more exotic footwork like skips and steps and skitters and hops and ballestras.

The mechanics of getting to the edge is not the problem; the problem is working out where the edge is in the first place. Smart defenders rarely reveal how far they can actually reach. Their danger zone lengthens when they are on balance, and it keeps moving around as they move. The attacker must take care not to inadvertently wander into the defender’s zone.

Once at the edge, the attacker must clear the zone. This is obvious when the opponent’s blade is in the way, but is just as important when it is not — it takes only a few milliseconds for the defender’s blade to snap into position. The attacker can make ‘preparatory beats’, cuts to the distal two thirds of the defender’s blade, to clear the zone. Whether the beats actually make blade contact is not important: either they hit the defender’s blade away, or they don’t hit because the defender already moved their blade away.

When I started fencing, the idea of deliberately seeking the blade defender’s blade on the finish was considered with some distaste. It was a crutch to be used only when necessary, or by inferior fencers. Good fencers with good technique and good distance didn’t need to make blade contact on the attack. They could just avoid the blade and disengage through. I was never that good, but it took me a few years to shake the residual shame. Until, of course, I noticed everyone on the A-grade spamming beats shamelessly.

An attacker can make as many preparatory beats as they like. They just have to make sure that they are moving forwards the whole time, to maintain priority, and that the motion of their beats looks different from the motion of a hit at the defender. Otherwise, the referee may judge that the attacker has missed their attack.

Once the path is clear the attacker may rush through the zone to hit. The operative term is rush. The attacker has as little as 180 milliseconds — the cutoff time — to rush through and hit the defender. Any longer, and the attacker risks being counterattacked, single-light, at the last moment.

What makes this difficult is that the attacker cannot simply launch across the zone — barring significant physical supremacy — without the defender easily parrying the hit or moving out of range. The attacker has to enter the zone first. But while they are in there, they are exposed to the defender’s actions. A good defender can hold the attacker in the zone for a long time, retreating in sync with the attacker’s advances, close enough to land potshots at the attacker while far enough away to negate any attacks.

The attacker, therefore, has to first clear the zone of the dangers they can see. Then, as they enter the zone, they must shield themselves from the dangers they can’t. Guards and binds work well. Nikolay taught me to do so with the testudo-like parry 5, and the more dynamic circle-4 “reaper” — there’s a clip somewhere of his son knocking an opponent completely off the piste with this move. But, most of all, the attacker needs to hustle: go in, go through, go fast.


While an attacker may loiter as long as they like at the edge of the zone, doing so also allows the defender keep their distance and spam their own beats to gain priority. The odds are in the defender’s favour: they can make an unlimited number of attempts, and only have to succeed once. The baby Busan kids sweep their blades around like miniature propellors; one clip, and they rush you back.

A defender can also sneak forwards, to place the attacker in the danger zone at a time of the defender’s choosing, and without the attacker realising. This is not difficult; humans are bad at depth perception and use cues like the defender’s blade to measure distance — cues that the defender can spoof. One neat trick — originally from the Italian Stallion — was to go “fishing”. Reel the blade back, give some slack, reel it in again, keeping his arm bent so it looked like he had less range than he had. Then boom! Counterattack in the face.

Better, then, for the attacker to step into the danger zone at a time of their choosing, not the defender’s. Now both fencers are close enough to hit. Doing so forces the defender to choose from two bad options: counterattack or beat at a moment when the attacker is prepared, or run away before the attacker has time to launch? Most choose the latter.

If the defender runs, the attacker should abort their attack. It is dangerous for the attacker to run down a fleeing defender. Better to abort the attack and resume the process — close, clear, charge — further down the piste.

A defender can run away several times down the length of the piste. Each time, the distance between the attacker and defender expands and contracts, their motions not quite in sync, like hands squeezing an accordion. Each time the attacker steps into the zone, the defender dances away. The dance always ends when the defender reaches their back line.

Few defenders let it get to that stage. They usually try to force a resolution earlier. But not too early. The longer the defender can keep this up, the more likely that the attacker will make a mistake and get caught by one of the defender’s torrent of beats.

Good attackers time their steps such that they accelerate forwards just as the defender stops fleeing, crushing the distance between them. Eventually the defender can no longer pull away, or no longer wants to. The defender stands their ground, and prepares to counterattack or parry.

This is the moment for the attacker to commit! The attacker must not hesitate. To do so is to give the defender a chance to win.

I am always most nervous at this moment on the attack. Because despite the best preparations, there is always a chance that the defender will somehow thwart the attack. But their chances are small. This close, the attack can smash through parries, angulate around guards, and reach all but the quickest of fall-shorts — especially if the attacker brings their feet together just before launch. Some add a sprinter’s crouch to power up just before launch. Others just hit hard — I once had the Brony crush my own parry against my chest.

But if the stars align the defender may prevail. My advice to the attacker is to just accept this. Commit to the attack again, next time — perhaps to another target. Sabre favours the attacker. Lose one battle, win the war.

One thought on “Week 2: The March and Siege

Leave a Reply