Hacker News Reader

Comment by antipax | Hacker News Reader

Biggest problem with these games (I've found, anyway) is that your sword controller cannot magically stop in mid air when your virtual sword hits another sword or a wall or whatever. It really ruins the immersion.

View on HN

It will also have to drastically downplay the importance of footwork. In a real swordfight, where to put your feet is nearly as important as where to put the sword.

This is also true for a plethora of sports games as well, specifically racquet based games. Swinging the racquet is less than half the battle -- footwork, in most physical types of activities, is much more important.

Omnidirectional treadmill? http://en.wikipedia.org/wiki/Omnidirectional_treadmill

These things are ridiculous. Researcher: "We need an omnidirectional treadmill, but how?" Engineer: "Oh, that's easy. We'll just make a gigantic TREADMILL OF TREADMILLS."

That engineer would make a good Haskell programmer.

or lisp, take your pick

> Omnidirectional treadmill?

About 3:50 into the video entitled "The weeds" Mr. Stepehnson says "We're going to start with an off-the-shelf controller that anyone can buy today. This eliminates the risk that we'll blow this by failing on some ambitious custom hardware project. From there we'll iterate based partly on your feedback. "

So no treadmills at first.

And intimately related, distance. In Kendo at least you win or lose by knowing where the cutting part (the forward 1/4 - 1/3) of your sword is going to be at the moment you make contact. This means thinking strategically about distance (e.g. I can feint a forearm strike and still have enough space to cut across the abdomen if she raises her sword to block.)

This is why I don't like games talking about how 'realistic' they will be. Firstly, because anything that runs on a standard computer will be far from reality anyway, and secondly because often users don't actually want reality (eg. most FPSs have far from realistic walking speeds - for good reason. Also stylised graphics etc.).

That said, the concept of different styles of swordplay is great - and having more 'realism' in that is something I think is pretty admirable.

Don't get bogged down in the pedantry of realism. When 'realism' is mentioned in gaming, it really means 'realism that assists immersion'. You'll occasionally see articles about balancing realism versus gameplay - if something is real enough to turn you off the game (like boredom from slow movement speed), then it's reducing your immersion.

Johan Andersson, the designer of historical strategy games (Europa Universalis), put it best: "When realism gets in the way of fun, fun wins."

Surely any game where you can respawn even once is already extremely realistic.

Would anyone want to play a war simulator FPS that starts with a cutscene of the Normandy landings and has a 50% chance of immediately being cut down by a machine gun whereupon the game quits and deletes itself off your hard drive?

Relevant PBF comic: http://pbfcomics.com/134/

Wow, there are a couple of new ones!

It's interesting to note that in GTA IV (not exactly an FPS, though) gameplay appears to be slowed down quite a bit compared to previous games in the series. Probably that's facilitated by more realistic graphics, with many little details to entertain our brain while the character gets out of the car.

> In a real swordfight, where to put your feet is nearly as important as where to put the sword.

Arguably more important. I fenced epee in university, and our coach was a little bit old school -- we went weeks just doing footwork before getting a weapon put in hand.

Well, in epee your opponent can score by stabbing you in the foot. That's naturally going to give where you put your place your feet a little importance to begin with.

That's premature optimization. That's the not the important reason for footwork.

This is really true. There are a lot of great stage combat clubs and groups around the country, after a certain point, it seems like it might just be more fun to go spar for real, or go learn some real choreography if you want the added 6 layers of safety (a must if you are using steel, IMO).

That was my thought as well. If I need to buy some fancy hardware and swing it around to play a game... I'll probably just go buy a practice foil and sign up for a class instead!

If it were a real [sword] fight most couldn't play. It's a game. Certain allowances have to made for ease of use.

I mean, do you really think COD is "real combat"?

i almost quit the cod4 single player campaign because it was a bit too ... maybe not realistic, but stressful. i thought to myself: war is hell, so why do i want to play a game that tries to capture a soldiers misery?

"most couldn't play" - but they could learn to play (as long as losing isn't fatal for the player). some games strive to be realistic above all else (some car racing games), because they are tailored for the ultra-enthusiasts.

This.

I took a not-for-credit intramural fencing and kendo classes back in college, and 3/4's of both classes was about learning footwork and movement.

Sword fighting also generally comes down to a few seconds of motion and whoever judged their opponents movements better. Neither of these would be present in a game that wasn't full motion capture.

I had a kendo instructor who said that footwork was 80% of the physical side of it. A fencing instructor said something similar. Boxers and boxing coaches also understand this.

Combine Kinect with the controller. I'm more concerned with haptic feedback.

Or you can use Kinect or Wii controllers on each limb like this guys do: http://www.youtube.com/watch?v=erMi8kxYFIE

Yep, that and the controller has no weight. To make a realistic game you would then need to make the character have a max swing agility (slow than the player) and also give the character a stamina bar. Anyway that's not to say you can't have fun sword fighting motion games. It just means you can't approximate reality 1:1 in all dimensions.

But you can still approximate reality if you just accept the control dimension will be abstract! Consider that Sim City uses an abstract interface to simulate city planning. Surely thats not the interface real city planners use (meetings, proposals) but it's still approximating a reality of city planning much more closely than other games.

Swords are a lot lighter than people think, even the heaviest armor in use (completely different from the ceremonial stuff) was much less restrictive than usually believed, and the average video gamer is significantly less fit than the average medieval soldier. I suspect that if you used a wiimote or a kinect-tracked stick things would all balance out. It would still require some abstraction, true, but not quite that much.

[1] European longswords weigh 1.5 kg on average: http://en.wikipedia.org/wiki/Longsword

[2] Guy doing gymnastics in armor: http://www.youtube.com/watch?v=xm11yAXeegg

> the average video gamer is significantly less fit than the average medieval soldier.

I am curious if this is just speculation or if you have evidence for this statement? At first it seems reasonable but I am not so confident after you factor in modern medicine and advances in nutrition/diet and discount for stereotyping.

Consider the draw weight of the longbow - "Although the draw weight of a typical English longbow is disputed, it was at least 360 newtons (81 pounds-force) and possibly more than 600 N (130 lbf), with some estimates as high as 900 N (200 lbf)". And they would loose approximately 6 arrows per minute in combat. And this was a weapon for the yeomanry, not the guys whose entire lives were devoted to maintaining military prowess.

Another quote: "Modern longbows have a useful range up to 180 m (200 yd). A 667 N (150 lbf) Mary Rose replica longbow was able to shoot a 53.6 g (1.9 oz) arrow 328 m (360 yd) and a 95.9 g (3.3 oz) a distance of 249.9 m (272 yd). A flight arrow of a professional archer of Edward III's time would reach 400 yds. It is also well known that no practice range was allowed to be less than 220 yds by order of Henry VIII"

http://en.wikipedia.org/wiki/Welsh_longbow

> And this was a weapon for the yeomanry, not the guys whose entire lives were devoted to maintaining military prowess.

Yes in fact they did spend their whole lives practicing. (From http://en.wikipedia.org/wiki/English_longbow#Training):

"Longbows were very difficult to master because the force required to deliver an arrow through the improving armour of medieval Europe was very high by modern standards ... etons of longbow archers are recognisably deformed, with enlarged left arms and often bone spurs on left wrists, left shoulders and right fingers."

Further:

"It was the difficulty in using the longbow which led various monarchs of England to issue instructions encouraging their ownership and practice, including the Assize of Arms of 1252 and King Edward III's declaration of 1363."

The popularization of the crossbow, in part, was the reduceded need for training. In some areas it also paid better.

Perhaps I was a bit ambiguous - they were free farmers, not professional soldiers. Even if they spent a good deal of time practicing, it was not their primary pursuit.

Yes I understand the background, but the difference between the bow and the crossbow was that of both skill and strength. While not their primary pursuit the long-bow was still a skill that the State needed to motivate their citizens to practice between wars. As an aside, were average citizens allowed to hunt? From the little information I've read it seemed like it was considered poatching.

This is quite interesting. I did some archery a few years back (recurve bow) and picked up a second hand, 50lb bow, which, apparently, for a slender early 20s female would be "quite tough" to draw, but I managed it, and got quite good (and strong arms). I never really took it very seriously though, but it was lots of fun. I can only imagine how hard it would be to draw an 81+lb bow (though perhaps longbows distribute the force differently? Otherwise that's going to be some sore fingers!).

Your rate is low. A semi-decent archer (which most men were, practicing every Sunday after church) should be able to loose at least 12 arrows/minute. Most students (~80%) are able to do that, safely, after about 4 3-hour training sessions, 2 weeks apart. Admittedly, they're only using 30-40lb bows, and can only do it for 1 minute, but that's just lack of strength/stamina which comes from practice. A fair proportion can reach 15 arrows/minute, and a very few will eventually be able to get off 18/minute.

This is for shooting at a block of billmen, which does not require much in the way of aim. So long as you aim in roughly the right direction, and get your elevation and draw length mostly right, that's all that's really necessary.

For target shooting, most competent archers will be able to hit a 1m target at 20m >75% of the time, loosing in time with a 12/minute count.

At 12/minute your not going to be able fire again before someone 20m from you get's in range. Second 80lb is light on the light but useful end of a medieval bow. Also, their arrows where much heaver and less stable in flight than what most people use now days.

PS: There is something of an arms race with bows, if you use a 100lb bow and your opponents use 80lb bows you can slaughter them before they get into range. (Ignoring wind, and terrain issues.)

No, but that's why you aren't a lone archer on the field. You're in a block of dozens, and you're behind a block of billmen who stop anyone with a hand weapon (bill, sword) from getting anywhere near you.

If that fails, and someone is anywhere near 20m, you drop your bow and a) draw your own hand weapon, or b) more likely - run. Because if you're up against someone whose primary weapon is a sword, they're likely to be better than you and have more/better armour, so your odds are low. On the other hand, if you're up against a billman, your odds are basically zero.

I only put the accuracy for a 20m target because a) that's what we shoot at for practice in my reenactment group, and b) to demonstrate that most people can aim moderately at that distance while shooting quickly - they're not just pinging arrows off in any direction in order to get the rate of fire up.

As a longbow archer, your primary job is to put as many arrows as possible into a large block of men 200m-300m(-400m?) away before they get to the billmen you're protecting (and who are protecting you.)

I can recommend the book "The Time-traveller's Guide to Medieval England" - it mentions the amount of military training that both knights and commoners were expected to perform. Basically a knight was trained from a very early age to fight - no way would an average 21st century male be able to compete in skill or strength.

Even for commeners they were still often legally required to spend a significant amount of time working on their marshal skills - mainly archery.

http://www.amazon.com/Time-Travelers-Guide-Medieval-England/...

e.g. Here in Scotland there was a time that games like golf and football were officially banned as they were seen to distract men away from military training for the wars with England.

Thanks for the book suggestion, I just purchased it. I had been looking to read a book about Medieval times but hadn't taken the time to find out what one to read. The reviews look promising.

Bit of a difference between overall health and sword-swinging fitness. I may be almost completely disease- and parasite-free, but the guy that's been training five hours a day since he was five will always lift more than me and run further than me.

Does “overall health” mean being disease- and parasite-free, though? There's at least one disease I heard of from which the case may actually benefit with increased lifespan[0]. That said, I tend to agree that to equate fitness with health probably isn't correct either.

[0] http://en.wikipedia.org/wiki/Celiac_disease#Screening

Look into the physical fitness of Amish and the less developed world in general. You will also find the answer in any researched history of medieval life.

The short answer is that a typical modern person (gamer or no) is much less fit than anybody from the age of melee weapons, let alone knights.

> The short answer is that a typical modern person (gamer or no) is much less fit than anybody from the age of melee weapons, let alone knights.

But I don't have smallpox. Or cholera. Or an iodine deficiency, because I live so far from the ocean.

That is another discussion since we are talking about fitness, specifically upper body strength and endurance, not rates of illness.

So if you had most of those things you'd be dead and therefor not factor into this discussion. The ones who didn't die were much stronger than the average person today. They had to be.

14th century knights had the physiques of professional rugby players:

http://www.bbc.co.uk/news/10138060

http://www.stirlingcastle.gov.uk/home/newsevents/bbc2history...

http://www.dailymail.co.uk/news/article-1279306/Face-mystery...

On the other hand, that long sword has more rotational inertia than the slightly heavier roman gladius, let alone a much smaller and lighter controller.

Having fenced a wee bit in high school I was shocked to get to waive a claymore around - a ~1.5m long two-handed broadsword has a lot of inertia.

I remember Gaston Phoebus' war armour displayed in Foix castle. It's made of relatively thin plates but reinforced with 5 mm thick iron bars and weigh at least 60 kilos. Of course the guy was a hell of a beast but... It doesn't look anything like the sort of tournament armour such as the one in the video or those most currently shown in museums (hint: if it's beautifully engraved, it's probably not war equipment).

Using a stick won't work because a typical living room doesn't have enough space to swing it around withoud damaging the furniture.

Perhaps use a gyroscope to simulate the inertia of moving a larger weight?

Fortunately, human apathy does not yet rule all of reality.

I wonder if it is possible in some way with gyroscopes. Not going all the time, but spun up quickly, to provide counter resistance for a specific direction of rotation. You could spin up two, in opposite directions, to counter the torque from acceleration. Difficult to spin them up fast enough, but perhaps just the sensation of some resistance would be enough feedback.

Interesting idea. However, would that create an unnatural, 'valley of the uncanny', type feel? Approaching realism but falling short and ultimately distracting?

I believe this is what Blade of Honor (Tsurugi) did, though I can't find any deeper reference to the controller.

http://www.arcade-museum.com/game_detail.php?game_id=10210

More ambitious idea: Make the controller a grip on the end of a robot arm.

I bet you could come up with something passable between 2 off-axis gyroscopes, 2 large induction motors (magnet suspended in a coil that vibrates back and forth – can achieve below 1hz - ~300hz vibration) and some more standard vibrators.

This only helps with angular momentum, not with linear momentum.

Is that really a problem? If something hits your sword it's going to make you rotate your wrist, not stop your hand, because it will be blocking the blade and not the handle.

Not if you're using a sword intended for cutting strokes and strike correctly. Distal to the "sweet spot", your hand will continue while the blade is rotated; proximal, and the distal end of the sword will tend to continue while your hand is levered backwards. Just as when using a tennis racket or a baseball bat, though, the moment is balanced around the point of contact when you strike correctly (the proximal part of the blade has more mass but less velocity; the distal has more velocity but less mass) and the blade, as a unit, stops (or is at least slowed by a uniform scale when the target can be cut through). Transitioning from a "chop" to a "slice" requires additional input from the swordsman.

They addressed this in an update:

>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!

There's a longer explanation here: http://www.kickstarter.com/projects/260688528/clang/posts/24...

Next biggest problem is that, in reality, Medieval duels rarely lasted more than a minute before one side was fatally injured. Not nearly long enough for a video game.

It doesn't seem like many video game battles last more than a minute. If you're still fighting somebody after a minute, that's either a very slow game or one or both of you are playing very conservatively.

Disagree, no harm in being short as you can have many duels.

I'm curious; do you have a source for this?

I was one seeing an olympic fencing event on TV. The duels were incredibly short.

Ogami Itto can slice up 15 people in 2.3 seconds. Does that count?

A minute is a long time for a single fight in a video game.

Are you not familiar with rounds?

A lot of fighting game rounds are over in under 10 seconds.

I'm surprised no one has mentioned haptics in this thread. Novint for example produces commercial haptic controllers. Haptics are what can get your controller to magically stop mid-air.

This is why haptics is so important. About 15 years ago I did a virtual reality race car simulator at SAIC. The steering wheel was controlled to simulate the feel of hitting a curb, the smooth feeling during a spin out, etc. Good for suspension of disbelief.

Two years later I did a ritual reality raft simulator for Disney. The paddles had no feedback mechanism and it hurt the experience (paddling with no feel for paddle going through the water). At the same time someone I worked with did a sword fighting simulation that was very cool, but also did not had force feedback effects.

Most shooters don't bother making you deal with recoil either though, or the weight of the gun.

But in shooters, opposing weapons don't regularly make physical contact with each other. In sword fighting, that's a necessity.

Non-physics geek here. Would it be possible to embed some gyroscopes on servos or something so that the game engine can deflect the controller? It wouldn't be ideal but it could provide some directional feedback.

I agree. Maybe the best way to approach a sword-fighting game would be to make a cheap sort of sword-fighting robot/dummy thing. Maybe like a mannequin on wheels that could quickly extend and retract one of several plastic rods and spin around and whack you with them or something...

I googled "sword fighting robot" and found this: http://www.bit-tech.net/news/bits/2011/07/20/kinect-used-in-...

I agree. A robotic arm (even a weighted stand with a hard plastic arm and thin plastic sword) seems pretty much required to get realistic sword fighting in. A good camera on the base that keeps an eye of where the "sword" is could do a reasonable job of keeping it from hitting most things, yeah?

And I think people wouldn't have a problem with it. They made room for the entire Rock Band ensemble. Just make it fold flat-ish to stow under a couch or in a closet.

Indeed, and Kinect requires enough clear space for a simulated duel to take place (especially for two players).

(Time to ramp up my iaido skills)

my approach to realistic fighting is: go to kendo keiko as often as possible. even the bruises seem real.

Kendo's pretty far from realistic fighting. The fact that it's a contact sport is cool, though.

I think it is more realistic than a game will ever be. You can block and have your strikes blocked. That seems pretty impossible to do with a controller.

Given that the project is called "Clang!", they'll probably want to incorporate that!

wonder if you can use a gyroscope to add some of this effect....

Reply on news.ycombinator.com