Second only to guns in the collective gaming armoury, swords form a huge part of our virtual experiences. Yet whereas games have more or less perfected the art of shooting a simulated gun at a simulated face, when it comes to swordfighting, their representation is the equivalent of children fighting with sticks. There’s a very simple reason for this. Presenting an accurate simulation of swordfighting in a virtual space is probably the most difficult thing a game developer can attempt. “It is just fantastically complicated,” says Guy Windsor, founder of the European School of Swordsmanship based in Helsinki. “ What you tend to have in a game is, like, three or four moves, and you can block or strike. And that is so, so simplified that it’s just completely unlike an actual swordfight.”
Guy has been studying the historical longsword for 25 years, deciphering techniques he teaches from a 600-year-old Italian fencing manual. He consulted on Neal Stephenson’s ill-fated sword-sim Clang and has written several books on the subject, most recently Swordfighting—a guide for depicting fictional duels, aimed at writers and game designers. Although not a gamer, Guy is keen to see swordfighting portrayed authentically. “It’s important for it to be represented as accurately as possible, because the better the game tracks reality, the more convincing it is.”
Unfortunately, in games this is very, very difficult. Unlike books or films, where the main obstacle to accurate swordfights is a lack of understanding on the part of the creators, in games there are hard technical problems to be addressed. The first is that real and virtual swordfighting exist as completely different entities. “Swordsmanship is a spectrum phenomenon, and games are a digital phenomenon,” says Guy.
In melee combat games, for example, there are a predefined set of moves, usually comprising fast attacks, strong attacks and a separate blocking function, all of which are enacted by individual button presses. In real swordfighting, while there are a range of techniques with given names, they can be executed at different speeds, from different angles, and will often flow into one another with no separation. For example, with a longsword a block and an attack are often the same move. “ I liken it to a paintbox,” says Guy. “You have a billion possible colours, and a child’s paintbox will have maybe 12 of those representing little shades of paint. So basically, while there are a thousand shades of orange, with the orange in the paintbox there will be just one particular orange.”
Another massive obstacle to good gaming swordfighting is what happens when the blades clash. A huge component of swordplay happens after what is commonly referred to as ‘the bind’, and almost none of it is currently represented in games due to the intricacy of those particular movements. On top of that is the sheer complexity of the physics involved. “[It’s] not just one simple object obeying one simple force, but two objects with multiple forces,” says Guy. “Plus the forces they exert on each other, plus things like the resonance of the blades, the way the blades stick or slide or bounce off each other. It’s so complicated.”
But does any of this matter? Although swords are common in games, actual sword-on-sword combat is comparatively rare. Most of the time you’re fighting someone wielding a different weapon, or a creature that isn’t human at all, like a boss in Dark Souls or a monster in The Witcher . Guy believes it does. “It doesn’t matter what the shape of the person is behind a weapon,” he argues. “If there is a sword-like, stick-like or non-explosive weapon coming toward you, it will behave according to the laws of physics, and solid mechanics and tactics work.”
The problems facing swordfighting in games are considerable, but not insurmountable. The first step is to accept that a certain level of abstraction is necessary. Even our modern understanding of historical combat is an abstraction. Not only do instructors such as Guy have to translate and interpret techniques from centuries old manuals, we also no longer fight with sharp swords to kill, which dramatically affects how people react when duelling.
While games can’t replicate the complexity of a real swordfight, they can accurately portray the ideal of the form.
While games can’t replicate the sheer chaotic complexity of a real swordfight, they can accurately portray the ideal of the form. Guy discovered this while co-developing a duelling card game named Audatia, designed to help teach his students the names and functions of longsword techniques. The cards have symbols indicating whether the blades bind together or not when the techniques are performed. Videogames could use a similar randomising element to trigger blade binds, and then alter the controls contextually to focus on the more subtle post-bind techniques, allowing for greater nuance.
There are also some simple rules developers can apply to avoid embarrassment. For example, swords are next to useless against anyone wearing plate armour, a common mistake that almost every fantasy RPG makes. “It’s like you’re banging your sword against a metal plate,” Guy says. “Why would you do that? It doesn’t make sense. I mean, yes, whack him in the head. You could daze him. If you get a good clean headshot, take it.” Equally, it’s very difficult to overcome an opponent holding a shield using just a longsword, something The Witcher 3 represents to its credit.” There’s a reason shields were popular for 3,000 years,” Guy adds. “They’re really effective.”
Virtual swordfighting still has a long way to go, but perhaps we should view this as an opportunity rather than a problem. Also, we are slowly improving. Guy mentions the upcoming medieval RPG Kingdom Come: Deliverance , as one example he’s been impressed by in terms of their research. Warhorse Studios even backed a crowdfunding campaign for one of Guy’s research books. “Any time somebody makes a sincere effort to make it happen, to really try, they get my undying admiration and support,” Guy says. “ I look forward to the day when you can buy sword-training simulators as good as pilot training simulators.”
