Week 1: The Box of Death

“One must dare,” said Vladimir Nazlymov, one of those legendary Soviet swordsmen who gained renown as both a fencer and a coach. He was talking about how a fencer must first guess their opponent’s intention then commit to their own action, even before they see their opponent’s move.

Nowhere is this more true than at the start of each exchange, when the two fencers charge towards each other in the 4 metre long ‘box’. In the mere second or two that it takes for this initial clash to resolve, neither fencer has time to see what their opponent will do, let alone figure out what to do in response. The fencers must enter the box with a plan and see it through, even as they risk being wrong. In “the box of death”, it is better to commit to the wrong action than to not commit at all.

How does one make a plan at the start of the exchange? Fortunately for the sabre fencer, there aren’t that many options. With all due deference to the vast repertoire of actions across the sport of sabre fencing, there are only three types of actions a sabre fencer can do at the start of an exchange: they can attack at short range, defend, or make a long attack.

I had this explained to me by the reigning world champion on a rooftop in Shinsa, well after midnight, surrounded by empty soju bottles. It is like rock-paper-scissors, he explained. Rock loses to paper loses to scissors loses to rock. Short attacks lose to defence lose to long attacks lose to short attacks. It’s the same game, with some extras.

Among those extras is that each type of action — short attack, defence, long attack — can be implemented by any one out of hundreds of possible specific variations. A short attack can be an advance-lunge, or a flunge, or a double-advance-lunge executed at short range; each of these can be launched at any one of a dozen targets along a hundred trajectories.

But the choice of a specific action doesn’t change the underlying game. Every short attack is likely to lose against a defensive action. It doesn’t matter how you launch if your opponent parries; it doesn’t matter where you aim if your opponent makes you fall-short. You might get lucky and hit around a parry with your short attack — and if you had an inkling this was possible, by all means pick the correct short attack — but you shouldn’t have done this in the first place. You should have just made a long attack.

We teach you enough of these variations in this Course to give you variety in your repertoire and to explore the tradeoffs you need to consider when selecting or creating your own actions. But no course can cover all the possible actions in sabre. Nor would it be useful to do so; there are only so many moves that you can maintain at peak effectiveness, and the more you try to maintain the worse your average ability to execute them.

So this class just covers the simplest versions of the actions used in the box. From these we can build the more advanced actions that you may want to use later in your bouts. But you might not: the simplest actions are often also the most effective.


For the short attack, the simplest action is the advance-lunge (or patinando). When I started fencing, this was the most common way to start a bout. The moment the referee said “fence”, I would rush at my opponent as fast as I could, partly to hit them first but mostly to cover so much distance with so much momentum that I could reach them regardless of whether they moved forwards, moved backwards, or stayed still.

But the short attack is most effective when executed against your opponent’s long attack. To win, your short attack must be first. It is not just a case of hitting first — so long as you hit within the cutoff time, the fencer who hits first is not the same who attacked first. Nor, despite what I thought as a novice, is it about how fast or far you go. Instead, it is about who launched their attack first. And, if both fencers launched at the same time, which of them hit at the correct distance.

Who launched first may seem easy to determine; who leaves the ground first? This is easy to see for a solitary action like a step or a lunge or even a flunge. But few, if any, fencers can reach their opponent with such actions in the box. They have to use a compound action like an advance-lunge. Then the question is, did their attack start with the advance-lunge, or only from the start of the final lunge itself?

The reason this matters is because a fencer can use a game-breaking tactic if their (bad) referee judges attacks solely on who launches first. I am not too proud to admit that I exploited this to the hilt in my early days; my students even used it to win adult competitions as teenagers (karmic justice, they later turned the same tactic on me). I would make a short quick advance, see if my opponent was closing in, then lunge as fast as I could to skewer them on the approach. Most of the time the referee would judge it ‘simultaneous attacks’, a draw. Sometimes they would give me the point.

The reason this move was game-breaking was because I didn’t have to lunge. If I saw my opponent hesitate or run away, I could advance instead, take priority and chase them down. So long as I reacted quickly and accurately I couldn’t lose. The worst that could happen was a draw.

We will return to the subject of game-breaking tactics and referees later in the Course. For now, note only that a fencer who advance-pause-lunges loses priority against a continuous advance-lunge. The referee will judge that the fencer reacted to their opponent, and therefore counterattacked.

The referee, of course, has no way of reading the fencers’ minds, to truly know whether one has reacted to the other. They are not Jedi. They are more like Sherlock, inferring the fencers’ intentions from the evidence they can see. And some of the best evidence comes from the rhythm of the fencers’ footwork.

An advance-lunge has three footfalls: front step, back step, lunge. da-DA-DUM. If the beats accelerate in cadence, the referee infers that the advance-lunge was a continuous attack from the start of the advance.

If the rhythm of the beats stalls — da-da…DUM — the referee infers that the fencer paused after the advance. Thus their attack started from the lunge.

With two continuous attacks, the one that launched first takes priority. It doesn’t matter if the fencers advance-lunged, or did something exotic like double-advance-lunge or flunge. Extra (or missing) footfalls don’t matter.

If both attacks launch at the same time, the next thing the referee considers is the attack distance.

The distance is measured from where your sabre is at the end of your lunge. The correct distance is to hit your opponent with your sabre’s tip, rather than further down your blade. The attack with the more correct distance takes priority. A (good) referee will judge attacks to be ‘simultaneous’ only when both fencers launch their attacks at the same time, and hit at equally correct (or incorrect) distance.

So when you make your short attack, make sure your rhythm accelerates and you hit at the correct distance. Aim for where you think they will be, not where they were. Don’t aim for where you can see them; the delay between their movement and you registering their movement — approximately 300 milliseconds — will make you overshoot the target.


The short attack is easy to defend against. The simplest way to make it miss, to ‘fall-short’. You can fall-short with a retreat or a jump-back, but I prefer to run and cross my feet backwards. While sabre fencers have been banned from crossing their feet forwards since the 80’s, crossing backwards is still allowed. Just make sure you end up in on-guard position, ready to advance.

Running backwards, or to ‘crossover retreat’, is the most efficient way to get your body far away from the short attack. You lose some stability compared to the retreat or the jump-back, and you might go too far to hit your opponent back immediately, or ‘riposte’. But it is better to be too far away than to get hit just because your opponent flew further than you expected, or extended their attack mid-lunge, which advanced fencers are able to do. And it is definitely better to crossover than to blow a kneecap trying to do the “right thing” by jumping back with the fall-short — something I have seen way too often, to this day, from tween girls to college wannabe hotshots to Busan pseudo-pro dropouts.

The fall-short, of any form, is more reliable than the parry. Parries are hard. They require you to pick where the incoming hit will go, the path it will take, and the moment it will arrive. You may end up in situations where you have to parry, but you should never use them as your first option.

You can, however, make the fall-short then pull your sabre into a guard position to “parry” the opponent’s continuation of their attack, or ‘remise’. The on-guard position, in 3, is as good as any other. I teach my younger students to ‘turtle’ — yank their arms close against their torsos, like a turtle retracting into its shell. This gets them into a guard position and stops them leaving their arms in front to get hit. It works for adults too.

If your opponent uses the fall-short and parry, you can counter it with the long attack, or chase. The distinction between long attack and chase is academic. The former implies that the fencer chose their attack range ahead of time. The latter presumes that the fencer chose only to move forwards and finished their attack when they got within range. In practice, a fencer who sets their attack range long enough to reach their opponent’s fall-short also takes long enough to convert their attack into a chase.

The most iconic fencer of the 2010s, Gu Bon-gil — come fight me — refined this one move into his entire game strategy. It took years for the rest of the circuit figured out how to stop it, the “Grind”. That is, for those who didn’t just copy it; there are entire national teams that fence like GBG fanboys.

The simplest way to make a long attack is to make a huge advance-lunge. This is the way that Gu does it. But for once, the simplest action is not all that effective or easy to execute. Unless you have preternaturally fast reflexes and range — like Gu — your opponent will see your attack and either attack first or make fall-short beyond the range of your attack.

Instead, mere mortal, you should make a double-advance-lunge. The double-advance will give you the range to hit against the fall-short, and the time to add more advances if you need them. It also gives you time to parry the opponent’s short attack, if you guessed the opponent’s intention wrong. Your parry is not likely to succeed, but at least you have a chance.

That brings us back to the start. You might be thinking that it seems risky to guess what your opponent will do; wouldn’t it be safer to just watch them instead and react appropriately? You would be right; provided you are faster than they are. If you’re not, then guessing is better. You will need to do so anyway, eventually, if you stay in the sport; at its highest levels, everyone is so fast that no one can react in time. Even the fastest fencer has to guess, to some degree, if only so they know what to watch for.

This I learned too late in life. As a young fencer, I was always taught to see and to react. In the beginning it was easy. But as I rose through the ranks, it became harder to see, and even harder to see in time. My coaches and peers would talk of getting better, faster, cleaner. I learned tricks to pare milliseconds off my reaction time, minimise the travel of my blade, and accelerate faster than traditional footwork allowed. All of these things made me a better fencer, but not enough to fight the top guys.

It was not until near the end of my fighting days that some friends who could actually fight at that level pulled me aside and gave away the game over drinks. Guess, they said, not watch. Or guess and watch. 60:40. Guess right and win. Guess wrong and fight. Never lose.

But your guess should not be random. People are creatures of habit. You will find that your opponent is biased towards one move over another, or has a favourite place to hit. “At your level, everybody hits to low line,” said the Paladin, “because their opponents are bad at parry 2.” Or perhaps they execute moves one after another in a pattern: attack first, then chase, then make fall-short. (“Four out of the six people in pools,” said the Paladin).

In the box of death, one must dare. Guess what your opponent will do. Figure them out and your guesses will be correct most of the time. But, correct or not, you must commit. Then you win or fight your way out.

One thought on “Week 1: The Box of Death

Leave a Reply