Neal Stephenson wants to revolutionize sword fighting video games.
Hi, Neal Stephenson here. My career as an author of science and historical fiction has turned me into a swordsmanship geek. As such, I'm dissatisfied with how swordfighting is portrayed in existing video games. These could be so much more fun than they are. Time for a revolution.
In the last couple of years, affordable new gear has come on the market that makes it possible to move, and control a swordfighter's actions, in a much more intuitive way than pulling a plastic trigger or pounding a key on a keyboard. So it's time to step back, dump the tired conventions that have grown up around trigger-based sword games, and build something that will enable players to inhabit the mind, body, and world of a real swordfighter.
CLANG will begin with the Queen of Weapons: the two-handed longsword used in Europe during late medieval and early renaissance times. This is a well-documented style that has enjoyed a revival in recent years thanks to the efforts of scholars and martial artists worldwide.
At first, it'll be a PC arena game based on one-on-one multiplayer dueling (which is a relatively simple and attainable goal; we don't want to mess this up by overreaching). Dueling, however, is only the tip of the sword blade. During the past few years, we have been developing a rich world, brimming with all manner of adventure tales waiting to be written--and to be played. In conjunction with 47 North, Amazon.com's new science fiction publishing house, we've already begun publishing some of those stories, and we have plenty more in the hopper. Once we get CLANG off the ground we intend to weave game and story content together in a way that'll enhance both the playing and the reading experience.
"How will this be different than SoulCalibur?" you ask. .....
Low-latency, high-precision motion controller: Critical to a satisfying sword fight is fast, accurate response. This is especially important for CLANG given the depth and complexity of moves that are used in real sword arts. Initially, CLANG will make use of a commercial, third-party, off-the-shelf controller that anyone can buy today
Depth: Roundhouse swings and crude blocks just aren't enough. Real sword fighting involves multiple attacks delivered from different stances, pommel strikes, grappling, feints, and parries.
Expandability: Implementing the longsword style will oblige us to construct a toolkit that can then be used--by us, or by others--to create other examples of what we're calling MASEs (Martial Arts System Embodiments). If your thing is Japanese kenjutsu or Viking sword-and-board, then in principle CLANG should support it.
In Which Neal goes into detail, as is his wont, regarding Clang:
Please read Technical Update #1 for details on CLANG design elements such as MASEs and In/Out of System movements.
Technical Update #2 covers our plans for the combat UI.
In Technical Update #3 Neal discusses timing, lag and tachypsychia.
Technical Update #4 goes more deeply into Sword Sync and Force Feedback.
Video update for reaching 50% of our goal. In this video, Neal takes you on a further tour of our secret facilities and introduces you to two very special characters.
Video update: It is all in the memo.From the video archives of the Top Secret Sword Lab
Mike Yahn, Stunt and Fight Coordinator
Dan Mahon as Olaf
3ric Johanson, Nathan Pegram, and Hackerbot Labs
Jack Mcfarland as The Hipster
And an extra special thanks to Brady Hall, video director and editor.
Mr. Stephenson's Bartitsuit was tailored by Duchess, Clothier
We will build a keyboard/mouse interface, if for no other reason than we need it for testing and development. You can use it if you want, but it kind of misses the point. The “value added” of CLANG is the ability to play a sword game by moving the controller in pretty much the same way that a real swordfighter would move a real sword.
This isn’t a binding commitment, because plans can always change, but: we are big fans of the Sixense technology as embodied in the Razer Hydra controller. The engineering team at Sixense has found a way to make an extremely high-resolution, low-latency controller by making ingenious use of simple and inexpensive components. We intend to make CLANG work on any hardware that supports the same protocol.
We've been thinking about this for years. It's not in the videos because to try to explain it here would get us hopelessly deep into the weeds. We think we have an approach that will work. It’s hard to explain in detail without a very lengthy brain dump. It's not just One Big Awesome Solution. It's a number of separate techniques working together. Some of these are familiar (visual, auditory, and haptic feedback) and others center on some innovative UI schemes. If you allow the controller’s position to get out of sync with what is shown on the screen, you get some feedback to that effect and you get UI cues on how to get back into sync.
In general,if you drill down deep enough on the actual sword techniques, the tree of possible outcomes gets pruned way down. It turns out that you rarely have to solve the fully general problem of one sword stopping another sword traveling at top speed at an arbitrary location in space. Which is a hard problem!
If you are "swinging for the fences" with a sword attack---which is to say, if you are assuming a long follow-through---then you're probably doing it wrong. You don't have to cut the other guy in half. You just have to hit him. In most of these arts, you're trained to pull the attack and stop with the sword between you and the adversary. If the attack succeeds, you're done. If it fails, you have stopped with your blade in a tactically sound defensive position instead of swinging all the way through and taking your sword completely out of the action.
Not at first.
We don’t have unlimited resources and so we have to focus on a single clear achievable goal. After evaluating all of the options out there, we’ve decided that a PC game using the Sixense technology is the best thing to start with.
Latency and spatial resolution.
Tachypsychia (slowing down time). To some extent this has to happen anyway because the movements in real swordfighting are just too rapid to be transferred into a video game. We don’t want to slow time down a lot, just enough to make the game fun for normal humans with normal internet connections.
Indeed, the first release will be multiplayer only, with no single player option.
Since ⅔ of Subutai is left-handed, the question is not falling on deaf ears. Again, however, we have to make a distinction between “things we can plausibly achieve with $500K” and “things to add when it looks like we’ll survive.”
One useful fact about the longsword (and two-handed weapons in general) is that handedness simply doesn’t matter as much. Both of your hands are always in action. In some respect, a lefty fighting right-handed actually has some advantages in that there are some techniques that involve taking the left hand off the pommel and using it to do tricky stuff.
The video mentions "developing tools for others to use". Does this mean Open Source, or do you intend to license this technology?
Our tools will be free to use for user created content (UGC) for CLANG, much like any other modding/UGC tools. If another company wants to use them for their own games, we will develop a license. We have not yet throught through all of the implications of making all of our stuff fully open source and so we can’t commit to that one way or the other yet.
That is the plan!
Its not a patch (update) for the game, but an actual physical patch with the logo of the OMVI.
Hey I sent you a question/submitted my talents for consideration/have a valid gripe! You haven't responded to me yet!
We are getting a lot of questions and comments. Frankly, more than we are able to handle. We are but a humble band of rogues who, among our other endeavors, are doing our best to address what we consider the important issues associated with this project. We are also getting a lot of press inquiry. We will try to update the FAQ as frequently as we can, and this will address most of the stuff we are seeing in the message queue, and if we see more detailed questions getting asked frequently we will add that to the the FAQ as well. We are simply unable to personally answer all the the messages we are getting, but we see them, and we really do appreciate the response this is getting and if we had the bodies and time we would be answering each and every message individually. If you are submitting yourself for consideration of the team to make this game, now is not really the time for that. There will be a period of time after we get funded (optimism!) where we need to get everything sorted out, get our rewards done, get our heads together and only after that will we start putting out the net for talent. Please be patient!
The Razer Hydra has thumb sticks (among other controls) which will allow the user to move their character around in the world while directing the sword with the position and rotation of whole device.
Anyone who has played video games has had the experience of accidentally hitting a thumbstick at the wrong moment and thereby causing the character’s POV to swing around to a completely useless direction with the result of falling off a ledge or getting butchered by an attacker who is now behind you. This kind of thing is annoying because it’s so stupid. A real person would never make such a mistake. A real person looks where the danger is.
So, it seems that too much control can be as bad as too little. Too much control gives you the power to do unbelievably stupid shit that no real human would ever do.
Any swordsman will go on and on about how important footwork is. We know it to be true. This doesn’t mean that a good swordfighting UI should give the player unlimited control over how the character’s feet move. Actually, it kind of means the opposite. Footwork is determined by what you’re doing with your hands and with the sword. So, the right way to do this is to cue the footwork based on what the hands (and the sword) are doing.
The Razer Hydra doesn’t track body movement like a Kinect, but it is very good at knowing what the sword is doing. So the controller won’t be telling your character where to put its feet. Instead--during a duel, at least--footwork will be driven by what the user is doing with the sword. For example, moving the sword forward and to the left in a certain manner will cause the character to step forward, as that’s what one would naturally do it they moved their arm and weapon in such a manner.
Of course we want to eventually have a hardware setup that both reads the position and movement of the “sword” in your hands, and also reads the motion of your limbs. That, however, is beyond the scope of anything we can accomplish right now, so we are focusing on something that we have a chance of achieving without delving into a potentially bottomless pit of hardware development.
When you’re not actually fighting someone, you should be able to look and move around in the manner we’re all familiar with from other games.
Short answer: you’ll be able to move around.
To be realistic you have to be able to feel the impact of sword on sword! How are you going to do this?
Since devices that provide enough physical feedback to simulate the collision of swords are beyond the scope of the initial project, and because this IS a game and not a simulation, we have to accept that physical response is going to be limited at best to a bit of haptic feedback (a technical term for vibrating). Visual and auditory cues will have to supply most of the information. Consider a driving game in which you are running into a wall. You can’t feel the wall, but you can see what you are doing and know you need to pull away from that wall before you can move freely again. Of course a sword duel moves faster than a car race, but trust that we are very aware of the feedback issue and it is one of the main priorities of this project.
There is a middle ground to be struck here between “flailing my arm around like a jackass and winning the fight” and “realistic, steel-on-steel contact simulation that reverberates up your bones”. We’re trying to get as close to the latter as possible with the technology available.
Yes, even with the limited number of characters we will develop for our first release there will be female options. But don't expect them to be chain mail bikini clad bimbos! These are some serious ladies that are just as interested in protecting their delicate flesh as their male counterparts!
A good longsword isn’t that heavy--on the order of 1 to 1.5 kilograms. That’s pretty light compared to the mass of bone, muscle, and flesh that has to move in order to swing it. Factor in the weight of gauntlets and other armor, and the disparity becomes even greater. Two hands are being used to move the blade, so that’s twice as much muscle power, operating through much longer lever arm, compared to a one-handed sword. All in all, weapon weight just isn’t that much of a factor in longsword fighting. Which is part of what makes this such a fast-paced and exciting martial art.
Of course this doesn’t address the issue of maces and really large two-handed swords like the spadone or montante, which, with any luck, we’ll have to address in the future. Precisely because those weapons are so heavy, though, the art of wielding them frequently relies on keeping the weapon moving. Starting and stopping takes too long in a real fight.
Yes, but not in the first release. The first release is going to be the longsword. $500,000 is a frayed shoestring of a budget for a game, and we are going to focus on the core of ours to make a strong foundation. Because each new weapon involves a lot of animation for the character and lots of programming for the combat tree, each new weapon costs a fair bit to develop. But we will, yes we will eventually. We want to see you fighting katana vs glaive and falchion vs kusarigama.
Certainly, the game will come with a “how to play” manual. The illustrated manual offered in the rewards will be a deluxe edition based on the traditional “fechtbuchs” of the late middle ages and Renaissance, such as those by Fiore, Talhoffer and Lichtenauer. Only ours will be of a more modern illustrative style and teach you how to play CLANG with style and finesse.
The controller that we are working with has two separate identical-looking units, one for each hand. So, moving the hands independently is not ruled out on technical grounds. This feature could be exploited in a number of ways in addition to having a separate weapon in each hand. Many longsword techniques involve removing the left (or non-dominant) hand from the hilt. Eventually we would like to be able to capture those techniques by exploiting separate hand gestures. Two independent hands could also facilitate sword-and-shield fighting, empty hand fighting, or even archery. Again, though, it is important to emphasize that all of these will not be covered by our initial $500K fundraising target.
We would like to support the highest level of accuracy with regards to damage that we can, while balancing it with performance and development expense. Being able to target vulnerabilities in the joints of plate armor and so forth is a pretty important element of effective longsword fighting. We would love to see severed bits of our opponents flying off of them as appropriate.
Yes, Unity supports Mac very easily and we will ship for the mac at the same time as Windows.
We hope so, but we will only commit to it if we get at least $800,000. Even though Unity 4 will support Linux we are pretty sure it will be a bit more difficult to package up and build than Windows and Mac, so we want to be sure we have plenty of extra time to devote to making it work well.