Week 11: Watch

“The Hungarians,” The Major said, as he gesticulated with his imaginary pipe, “used to take lessons without a jacket. Or even a mask.”

“That’s insane,” I said. A flood of negligence claims and surging insurance premiums overlaid themselves onto my visual field. “Unless the coach is pulling every single hit, and the student is some kind of prodigy, how does anyone come out of these lessons without losing an eye?”

The Major flashed me a smug smile. “Well actually,” he said, “every good coach pulls hits and cheats. And a few light strokes” — he feigned a cut to my chest — “never killed anyone. It just makes for a memorable lesson.”

He paused. His eyes glazed over.

“And obviously,” he continued, “you need to teach them how to see first. You lock your eyes here” — he stared straight at me, glassy-eyed, just above my collarbone — “and use your peripheral vision.”

It was classic Major. On this occasion, he was giving us an impromptu seminar on the teachings of some famous Hungarian coach I’d never heard of. That in itself wasn’t all that unusual; I knew hardly anything about sabre back then, still less about its gurus from the byways of Old Europe. But this particular guru was apparently some kind of promethean figure to a generation of sabre fencers and coaches. At least where the Major came from. Which wasn’t exactly the first place one would go looking for top-order sabre fencing, but I wasn’t brave enough to point it out at the time. After all, the Major came highly recommended, was impressively credentialed, and given to the kinds of blithe insults that I, in my younger days, associated with true mastery; the insufferable genius.

So when the Major declared that anyone who knew anything about sabre needed to know the guru’s lessons by heart, I and the rest of the instructors gathered around to get lessoned. Even if those lessons featured a bunch of pale, shirtless, maskless Hungarian boys getting marked up. Not that any of us fit that description; we were the real-life equivalent of a fantasy RPG party lifted straight from 1970’s diversity sensibilities: a Man, an orc, an elf, a dwarf, and a token shield maiden complete with scowl.

The Major, though, wouldn’t have been any more incongruous in sunny Sydney than those boys would have been. He was a pale compact man of professorial demeanour; in his spare time, he liked to write fencing articles for submission to tiny academic journals. I’d hired him on Bond’s recommendation, after Bond decided to return to the home country to support his superior officer wife in her mission to bring some recalcitrant province or another to heel.

Like Bond, The Major also came to us for a bit of a working holiday, during a lull period for his home club which operated out of a charming boarding school hall in a bucolic village located nowhere and mostly catered to its transient denizens. That was his retirement gig; the Major used to be an actual soldier, until he actually saw combat on some forsaken island on the far side of the world and decided that the imperial military life wasn’t all it was cracked up to be.

As far as I could work out, the majority of the Major’s martial career had been spent writing up policies and procedures. He had the kind of zeal for doing things a particular way — in his view, the right way — that the Marines like to feature in their recruitment brochures. He applied that zeal to compiling the first real version of our syllabus, throwing himself into the task with a gung-ho gusto that belied his twilight vintage.

That, of course, meant that he had to go through our notes first. And what he saw so incensed him that he took it upon himself to fix up the things in our understanding of sabre that was not quite up to scratch. Which, in his opinion, was just about everything.

It wasn’t so much that we were wrong, per se. It was that we were so wrong that we didn’t even know we were wrong. It was bad enough that we were fencing Belarussian sabre; the nicest thing he could bring himself to say about it was that at least it wasn’t as bad a mangling as the Korean stuff.

Our very conception of the sport was wrong. The Major was a believer in coaching kids to go for gold; I thought we should teach adults to play for fun. The Major held that governments should be picking up the bill; I wanted the club to be commercially viable. The Major hired his former students as assistant coaches on part-time peanut wages; I bet on salaried professional coaches, of different lineages, who could work together and pool their experiences. The Major scoffed; there was only one true lineage.

For all our disagreements, I liked the Major. And he was right, in hindsight, about as often as he was wrong. At the time, of course, we clashed about as much as two people could under the circumstances: I thought he was a dogmatic diva who spent more time gabbing than teaching; he thought I was an arrogant punk whose pretensions were matched only by my naïveté — a fencing poster boy for the Dunning-Kruger effect.

I couldn’t understand the Major back then. A sabre master so bad at sabre  — he would brag, inexplicably, about how he once lost to a tween girl in a competition — that he spent more time talking in his lessons than actually demonstrating any moves. A walking repository for every rule and regulation in the sport, yet uninterested in any of the innovations that the athletes were deploying on the circuit.

It was only years later that I came to understand the Major’s Tolkien-esque conception of sabre. It was like he felt that everything that made sabre good and beautiful and noble was done a hundred years ago, the decades since a saga of decay and decline. Wave after wave of barbarians from the East had assailed the bastions that held onto the old ways, each time beaten back, but each time defiling a part that could not be replaced. And the hordes kept coming, their ranks inexhaustible, decimating the watchers on the walls that sheltered the last treasures of the sport. The walls would eventually fall; with every passing year their garrison dwindled in number and diminished in stature. But that only hardened their resolve.

The Major’s dogmas came from what he had seen from the dark side of the sport. He had seen fencing go from forging noblemen to forging athletic ability in college applications. He had seen fencing masters go from being a respected profession to being a part-time job for dilettantes. He had seen clubs torn apart, their students poached off by the former fencers who get stamped out by the hundred every year from the industrial sports programs of upstart nations, fencers who knew almost nothing about the sport that they could play well, but not well enough to make the grade back home.

The Major’s fight was not on the piste. For him, the stakes were nothing less than the soul of the sport. His role was to pass the torch of true sabre to the next generation who could carry on after he was gone. And that was the way he taught me to fence.

***

At the start of the exchange in a sabre bout, neither fencer has priority. To score the point in an exchange, a fencer must hit their opponent first, run them off the end of the piste, or hit them after they’ve gained priority.

These rules compel sabre fencers to rush off their start lines the moment the referee signals for the exchange to start. The speed and ferocity of the resulting clash leaves very little time for the fencers to ruminate on what actions they are going to do. They have to guess their opponent’s intention before the exchange begins. Then they can deceive their opponent into doing it. But above all, they have to execute their own action to counter it, even if they are not confident in its success.

The reality, though, is more complicated than that. Guessing works well for beginners, and guessing with deception is mandatory in the A-grade, where every opponent is some flavour of superhuman monster. But in the long reaches between those two extremes, you will meet countless opponents who neither guess nor deceive, yet can thrash you in the box.

They watch you. They see what you’re about to do as you come off your start line, and react with the correct countermove before you can finish your own. Most of them don’t have the faintest vestige of a plan while they fence; they just wander forwards and rely on their well-trained reactions to handle whatever their opponent throws at them. These “watchers” dominate the middle tiers of sabre fencing, shooting each other down with their hair-trigger reflexes, point after point, like a supercut of cowboy Western showdowns.

I fenced as a watcher. It came naturally to me, as it does to most people. Human instinct, after all, is to watch for and avoid painful threats like swords, not to kamikaze into them or trust your fate to the gods. It takes a lot of training for normal people to fight berserker-style, throwing themselves against armed opponents, risking life and limb (or so your hindbrain thinks), and generally taking a cavalier attitude to being stabbed in the face. Watchers are cautious; they look first.

Anyone can fence as a watcher, but some people do it better than others. Cadets are naturals. They tend to be quick of reflex, fast in action, agile in body — all complementary attributes for watchers, who have to be quick enough to identify what their opponent is doing in time, fast enough to execute their own move before their opponent, and agile enough to recover when they make a mistake.

Cadets are naturals in another sense too: it’s hard to get them to fence any other way. Contrary to common conception, teenagers are incredibly risk-adverse when it comes to combat. They don’t like to guess and dare; they would much rather be sure and use their speed and agility to take potshots from a distance. This is why most coaches train their child students to fence as watchers when they first start. It plays to their instincts, works better as they grow up, and rewards them for diligently honing their technique.

Thus, while watchers only make up a plurality of sabre fencers in the middle tiers of the sport, they make up virtually all of the sabre fencers in the youth leagues. If you fight an opponent too young to have graduated from college, it is almost guaranteed that they will be a watcher. The older they are, the better they get. Until they hit the plateau in their 20s, when their physical development stalls and their technical skills reach their apex. That’s the age at which they start losing to younger watchers; when two watchers fight, it pretty just comes down to who is quicker on the draw.

They lose to the older pros as well. Watching doesn’t work so well as an approach when the opponent can spoof what you’re watching for. Most watchers quit in their 20s, taking their aches and losses as a sign to retire.

But a few adapt. They are the ones who stay pro. Because all the pros used to be watchers too, and they still watch — to some extent — even after they have celebrated their 25th birthday and crossed over into the realm of old man game. It’s just that they learn to do other things as well. And so, if you have any ambitions of becoming or fighting them, you need to learn how watchers approach the game.

***

Excellent watchers are fast, agile, technically proficient, and have great reaction speeds. The first three attributes are straightforward to hone, but not necessarily easy: train hard, train often, train right. (Also eat chicken, advised my grandcoach. Supermarket chicken, all the hormones, one a day.)

The fourth attribute, reaction speed, is the most important one for watchers. But unlike the first three, the process to improve your reaction speed is not straightforward, even though in one sense it is easy.

Strictly speaking, you can’t improve your base reaction speed. The delay between you seeing something and you reacting to it clocks in at around 300ms for the young and twitchy, and gets worse as you get older and more mellow. This limit is hardwired: it is the time it takes for light to hit your retina, convert into an electrical signal, conduct via your optical nerves to your brain, translate into instructions, travel down your central and peripheral nerves, then induce your muscles to contract to your desires.

Even getting to 300ms is incredibly difficult; under real life circumstances, people’s reaction delays are closer to 500-700ms. And even that assumes the person involved is ready to receive a signal, knows what to look for, and how they are going to respond. If they don’t, reaction delays can be very long indeed; deer-in-headlights long.

So what I really mean when I write about training reaction speeds is to train the ability to recognise the right signal and make the right reaction in the shortest possible amount of time, not to somehow improve the speed of the underlying physical mechanism. 300ms is the limit. The aim is to get as close to that limit as possible.

There are three main sources for the extra delays in your reaction pathway: the time it takes for your brain to figure out what it had just received, the time it takes to for your brain to work out what to do about it, and the time your muscles need to effect whatever it is that you direct them to do.

The first source of delay, identifying what you’ve seen, is the easiest part to train. You just need to cut down the number of things that you are looking for. This is Barry Schwartz’s “Paradox of Choice”: give someone three choices for ice cream, and they’ll pick one (and be happy) in a few seconds; give the same person thirty choices and they will stand there for ages.

The fewer choices you have, the faster you can make your decision. Deciding between three choices takes longer than deciding between two; deciding between four choices takes longer than deciding between three, and so on. Each time you add a choice, you have to try to fit what you see into it before moving onto the next. Your brain asks: is what I am seeing A? Or is it B? Or is it C? Which of them, A, B, or C, is the closest fit?

In sabre, your opponent can execute an almost countless variety of actions at the start of the exchange. What a watcher does is to group those actions into as few categories as possible. At its most simple, there are only two categories: the opponent can move forwards; or they can not. If they move forwards, they are attacking. If they don’t, they are defending.

In each of these categories, your opponent could be doing any one of hundreds of specific actions. If they move forwards, they could be making a short attack, or a long attack, or a chase, or a setup before the defence. If they move backwards, they could be defending with any one of a dozen parries, or variations of fall-short, or those things plus accessories like stop cuts and jumps. Simply identifying whether the opponent is moving forwards or not is insufficient for the watcher to figure out what to do. But neither can they spare the time to add additional choices to their decision-making, or wait long enough to positively identify the opponent’s action.

What watchers do instead is to break up what they see and how they respond into a decision tree, each fork of the tree having as few branches as possible. Then they start reacting to each fork as they make the decision, continue their reaction while evaluating the next fork, and finish their reaction when they run out of time to watch or they feel confident that they know what the opponent is doing. 

Is the opponent moving forwards or not? Forwards. Start moving back. If they are moving forwards, are they chasing or attacking? Attacking. Get ready to pull away. If they are attacking, are they attacking to the left or the right? Left. Start moving your blade into guard position. High or low? Low. Ok, we make a guard parry to 4.

***

The next part is to make the identification process itself as fast as possible. Most of the information entering your eyes is irrelevant to the practice of sabre combat. You don’t care about the colour of their shoes, or the brand of their mask, or exactly how they are holding their sabre, or how nice their special Olympic-edition Omega is. None of those details matter. They just take up cognitive effort.

The only things you really care about are the closing speed of your opponent, how far away they are from you, the location of any open targets on their body, and where their sabre is. Hence the Major’s advice to use peripheral vision instead of regular focused vision. Peripheral vision uses rod cells: low resolution, high sensitivity light receptors that respond faster but with less information than the cone cells that cluster near the centre of your retina for focused vision. Rod cells strip out the details from what you see, leaving only the things you care about in a sabre bout: black-and-white imagery, motion, gross shapes, and depth perception.

All you have to do to use peripheral vision to not focus on your opponent, or anything else. Exactly where your eyeballs point don’t really matter. The Major preferred to blankly stare forwards. Master Shifu recommended looking at the opponent’s sabre guard. But not the blade; the opponent’s blade moves around too much and too quickly to track, and can hypnotise you into a dumb trance. The same goes for anything else that might snap your attention back to focused vision, like the opponent’s face or hand.

When I use peripheral vision in a bout, my opponent looks like a blurry silhouette; a fuzzy off-grey blob with a harder edge where their sabre is. Then all I do is track the blob. Blob closes distance, opponent must be attacking. Blob stays still, opponent must be defending. Blob comes forward with a flash of an edge, opponent must be about to hit. Time to parry.

I can simplify the view even further. After a few exchanges with an opponent, I can get a sense of how fast they can move when they attack and when they defend, and how much distance they cover. Then I can set up mental tripwires along the piste, to trigger a response if the opponent crosses them at specific times. If the blob crosses this tripwire at T+ 200ms, trigger A. If the blob doesn’t, trigger B.

I’d tweak the tripwire locations as I learned more about my opponent over the course of a bout. Back in those days, I often did badly in the 5-point pool bouts because I didn’t have enough time to calibrate the tripwires if my initial settings were too far off. I did better in direct elimination bouts; even if I lagged behind in the first half to 8 points, as I often did, I would usually make up for it in the second half — to the annoyance of my opponents.

***

Human vision doesn’t work as a continuous stream of signals from eye to brain. It works like video systems: a series of snapshots, 10 frames per second, a set from each eye, the brain stitching them together into a model of the world around us. The frames aren’t even all that complete; your eyes raster from side to side several times per second, painting their images in broad strokes across your visual cortex like a cathode ray gun.

Why this matters for sabre is that you make your decisions in much the same manner as you see with your eyes. That is, it is easier for you to make a snap decision at a specific time based on what you see in that precise moment, than it is for you to make a decision at an unspecified moment based on what you see over a period of time. I’m not enough of a neuroscientist to understand how this works, or enough of an evolutionary biologist to understand why, but the difference in response times matter.

The upshot is that you need to time your decision just as your opponent is about to commit to their action, even if that “action” is to see what you are doing first. This where preparations and rhythm becomes important. At the start of an exchange, both you and your opponent will move — usually forwards — during your preparations. At some stage, both of you have to complete your preparations and execute your action. But you can’t pick just any moment to do so; the way your feet move restricts you into launching actions only at particular times — for example, you cannot lunge until your back foot lands back on the ground. So if you know your opponent’s preparation, and roughly what kinds of actions they like to do, you can time your own preparation such that you decide just as they commit.

I once fenced a competition in Thailand where I was getting thrashed in the pools. I just could not work out what those guys were doing, but they seemed to pick my moves with preternatural precision. It wasn’t until I scraped through my first DE that I finally worked it out: my step-bounce preparation was finishing so early compared to theirs that they could see what I was going to do almost before I had even committed to it myself.

But simply slowing down my preparation didn’t work; now the referee was accusing me of waiting too long, and giving all two-light attacks in the middle to my opponent. I ended up having to switch to a double-advance preparation — my worst — for the rest of the competition. I hated every moment of it, but it was better than losing.

***

The second source of delays in the reaction pathway is working out what to do once you’ve identified the opponent’s action. If you ever take one-on-one lessons with a coach, this is half of your training: drilling the correct responses in the correct timing at the correct distance. Orthodox coaches overload the student’s responses, forcing them to make response after response without a break, while being pushed to their physical limits, in circumstances they would never encounter in a bout. The idea, of course, is to make it so that the student will be able to execute the correct response by instinct no matter how tired or how scared or how injured they might be. The student also gets a decent workout, and the lessons look really cool — dozens of parry ripostes and counter-parry ripostes in rapid succession, a real-life Dragon Ball Z flurry.

These well-drilled responses, though, can also be a watcher’s Achilles heel; they always respond the same way, so an opponent who can feint actions convincingly can virtually “make” the watcher do what they want. This is what coaches mean when they yell at their students to “make the opponent do X” from the sidelines.

The antidote is for the student to learn two or three alternative reactions to what they see, and to get better at identifying whether an opponent is really going to execute an action or just spoofing. Watchers tend to have massive repertoires for this reason, and require frequent tune-ups with their coach and club-mates to maintain their internal calibrations of when to do what.

They also need to be able to execute those responses very quickly and accurately; the third source of delays in the reaction pathway. The quicker and more accurate they are, the more time they can give themselves to watch the opponent before they commit to their response. It also allows them to make a second response in quick succession if their first reaction gets spoofed — witness Gu Bon-gil’s counter-parries or Aron Szilagyi’s compound parries.

Human muscle strength isn’t the limiting factor here; technique is. Your opponent might be twice as strong as you, but they won’t be ten times stronger. The distance your blade travels if it wanders around and overshoots and winds up — that can be ten times longer than your opponent’s blade path. Being able to identify your opponent’s action is important, but not if you can’t do anything about it. That’s why the other half of one-on-one lessons is practicing how make your actions as efficient as possible under any circumstances.

So far, the things that I have written about watcher training are applicable to fencers of any archetype. The cynic in me might think that coaches train kids as watchers because it’s a great way to make money, but the truth is that the training process makes good fencers, period.

Where things start to diverge between watchers and other sabre fencers is in their repertoire of actions and their underlying philosophy. Watcher actions are balanced, short, and rapid. They start with a slow preparation phase and end with an incredibly fast finish, even for simple actions like parries. Watchers don’t commit until the last possible moment, and they always hedge: the start of an attack can become a parry, the start of one parry can transform into another, a lunge can convert into a chase.

Go watch a competitive cadet squad and marvel at them doing hundreds of squats and holding stances so low that you can see their muscles bulging with tension. Fight them, and note how they would rather get hit than cross their feet backwards to make fall-short. See them practicing their bladework in the mirror — turn guard slowly this way, guard parry; turn guard that way, start parry make different parry. Listen to their coach berate them after every remise, even if it was successful. Feel the crunch of cartilage when they lunge: short, stable, sudden, and stopped in stance the moment they land, no matter how much it hurts.

It doesn’t matter whether these actions are successful or not, though they often are, or whether they hurt to execute or not, though they always do. They matter because these are the actions that embody what it means to be a watcher. You cannot pick and choose the ones you want, any more than you can leap from one railway track to another. If you choose to follow this way, then you must follow all that it entails — the way you think, the way you see, the way you move, the way you face the opponent.

One does not simply fence as a watcher. You can ape the moves, and learn the tricks. But until you accept the mindset that binds them all together you cannot be a real watcher; a sabre fencer of the one true lineage.


One thought on “Week 11: Watch

Leave a Reply