Substance and Style in Game Design

Gamasutra has a piece on the elements of substance and style within videogames, and what should be considered when designing with these elements in mind. From the article: "An easy way to understand the difference between style and substance is by example. Many shooter games have traditionally calculated world collision and bullet impacts by modeling bullets as instantaneous line traces and characters as moving collision cylinders. In this case, the line-projecting cylinder is the fundamental nature of the character - the character's substance. The image of a fighter, the sounds he makes and the way he animates is the character's style."

Comments

Strange game physics

[vertinox]

Many shooter games have traditionally calculated world collision and bullet impacts by modeling bullets as instantaneous line traces and characters as moving collision cylinders.

That always irked me in FPS games. In the real world, bullets aren't instantaneous and travel in an arc. Red Orchestra (a UT2k4 mod) does a pretty good job of simulating bullet drop and real life physics.

Re:Strange game physics

[masterzora]

This is where the question of "how realistic is too far?" comes in. I mean, there are plenty of people who would love no more than to have bullets take a bit of time and arc a bit, to be sniping and have to adjust for distance and wind.

On the other hand, this will just serve to confuse a lot of people, or make it difficult. In an effort to find a happy medium between the super-realistic people and the just-make-it-easy-to-pick-up crowds, I think modern FPSs aren't too horrible.

(Though, I would love a more realistic system, personally, but I don't think hyper-realism is for the masses.)

Re:Strange game physics

[HD Webdev]

That always irked me in FPS games. In the real world, bullets aren't instantaneous and travel in an arc. Red Orchestra [clanservers.com] (a UT2k4 mod) does a pretty good job of simulating bullet drop and real life physics.

FPS games are that way on purpose. It reduces the server load by quite a bit and reduces network traffic. Also, it's hard enough to deal with internet latency & client-side prediction of the action without throwing in craploads of flying bullets. (Most FPS games are now built on top of existing network-capable FPS engines).

For example, here is the basic outline of how a modern game server deals with networking for Counter-Strike:Source

Re:Strange game physics

[Lisandro]

I remember Digital Image Design (an old game developer) made a big fuzz about how realistic their new flight simulator TFX was; one of the points they touched was that rounds shot by planes had realistic physics - it was either that game or EF2000, kind of a sequel to TFX. Both were great games and those details added a lot to the experience. If they could do it the DOS days, they can surely do it now.

    But, in a fast paced FPS (a-la-UT2k4, f.ex.), i really don't think it matters much - after all, if we're going to complain about realism, a guy firing a flak gun while doing a flip in mid air comes first :). Tactical FPSs, yes, they could benefit from such touches, and Red Orchestra is a great example. The only other one i can think of is Counterstrike, where bullets don't fly in arcs but you get groupings according to distance instead of bullets that always hit the same pixel.

Re:Strange game physics

[Grab]

Re your midair flip thing, the problem there is that no FPS takes account of recoil. Oh yes, they make your sights move a bit, but that's it. No loss of balance or anything like that, not even firing a SAW from the hip. They don't even reduce aiming accuracy whilst moving. When they get all that right, look for all those ex-leet people jumping around to die in swarms to the one guy who stays stationary with his feet on the ground whilst firing.

Grab.

Re:Strange game physics

[Jackmn]

Try Day of Defeat for recoil.

As far as accuracy reduced while moving, just about every semi-realistic FPS does that.

Re:Strange game physics

[beakerMeep]

Well said. Except for the happy medium thing I couldn't agree more. For me, I'd say keep those ultra realistic features out of my game. I want fast paced action and not a bunch of people learning how to aim and a whole lot of wasted ammo. But that's me. And I fully appreciate that some people want hyper realistic. So it's pretty easy really. Some game makers focus on gameplay, others create super realistic games (Rainbow 6 maybe? i dunno). And to me, I am happy to have the choice. What makes me unhappy is when someone's ego comes strutting along and complains that my UT2004 pyhsics stink. To that I'd say go play ms flight sim :P

Re:Strange game physics

[Grab]

All the ones I've seen move your sighting in a predictable side-to-side way. This simulates running very well. What it *doesn't* simulate well is the behaviour of arms whilst running - I've yet to see a FPS which does anything except fire at whatever's in the centre of the screen. If it comes through the dot in the middle of the screen, bam, it's dead, regardless of how fast you're running. Except that IRL, where your eyes point != where your gun points. No game I've seen makes any allowance for this.

Trouble is that to get it right would basically need two mice (or mice and joystick), one changing how you're facing and the other for changing where you're aiming. The effect would likely be to have crosshairs that move round the screen, like the old Star Wars arcade game or like missile lock on flight sims.

Grab.

Re:Strange game physics

[kniLnamiJ-neB]

Try Delta Force: Black Hawk Down. Reduced accuracy while running (they give you a sort of circular area created by the crosshairs, and the diameter grows when you're moving). Running with a sniper rifle requires a lot of luck to hit anything unless you're standing right next to it. This same effect is used in the "recoil" system... when you use a machine gun, your accuracy goes down with each bullet that leaves the barrel. They also don't allow use of the scope while running... you fire from the hip unless you're standing still. On top of that, the engine arcs bullets and also makes them take time to get there.

I think it's good to see realism in games like this. It's so much more rewarding to fire my sniper rifle and hit a moving target at long range with wind effects... anyone can just point the crosshairs at a dot on the screen and pull the trigger.

Re:Strange game physics

[tartley]

Presumably it would be less confusing if you could give the player some realtime feedback as to what was going on, eg. with 'tracer' lines which showed a bullet's arc and speed. Make them only visible to the shooter, if you feel that negatively disrupts stealth attacks.

Re:Strange game physics

[Minna Kirai]

- I've yet to see a FPS which does anything except fire at whatever's in the centre of the screen.

Meaning, you've never seen a PC FPS set on the modern planet Earth. Battlefield2, Counterstrike, Deus Ex, Rainbow Six, Americas Army... they all slash gunfire precision while in motion. Even the fantastically whimsical Castle Wolfenstein is careful to make an MG42 uselessly inaccurate if fired while standing.

Shmeh

[Eli Gottlieb]

A grouping of game elements into substance and style is somewhat useful, but really couldn't anyone with a decent knowledge of game development have figured it out?

Re:Shmeh

[badasscat]

A grouping of game elements into substance and style is somewhat useful, but really couldn't anyone with a decent knowledge of game development have figured it out?

Well, apparently not, because I don't think Gamasutra quite got it.

A game's substance has nothing to do with how a bullet travels through the air. That kind of thinking is why we're all stuck with cookie-cutter FPS's, sports and racing games these days. People (including most developers) mistake genre conventions for substance. That's still part of a game's style.

The reason we're stuck in this rut of game development today is that developers think they can look at existing games, alter minute stylistic details like how a bullet travels through the air and call it a new game.

The substance of a game is much more fundamental than that, i.e. what is the player avatar (a character, a ball, a group of ants?), what is the goal of the game, what is the gameplay mechanic required to reach that goal. Many developers simply skip these fundamental questions and start out with a pre-made template based on existing genres - these developers are not creating substance at all, they are simply making a stylistic alteration to a set of existing conventions.

Re:Shmeh

[Idealius]

Actually, the way the bullet flies in the air can mean a CRAPLOAD about substance and style.

Multiplayer games and online competitions. I forgive you for thinking entirely in the context of a single player game, though. :)

I'm rather disappointed.

[SharpFang]

The article tries to draw a clear line between substance (mechanics/rules) and style (look/feel/story) of a game - but the story itself has very little of either.
In terms of style it sounds a bit like a lecture for a school of game developers. Readable but boring. The examples are dull or oversimplified. FPS gamers concentrating on Substance, The Sims concentrating on Style? I wouldn't be so sure players of Half-Life 2 are all about substance...
In terms of substance... actually I'm not really sure what the article tries to achieve at all, because it first defines a sharp border between the two and then methodically expresses that the border isn't really important at all, because both are just as important, and by emphacising one or the other you change the style of the game, not whether it's good or if it sells well. I don't think I've learned a thing from the article, because all the info it presents seems pretty much as completely useless analysis. Like analysing structure of a sentence in a poem won't help you writing better poetry, analising whether your game has enough substance or better style won't make better games. And the article doesn't write neither how to improve the style nor how to create better substance...

Re:I'm rather disappointed.

[dynooomite]

I agree, I don't really see any point in this article other taking up a bit of server space. It seems to me that whoever wrote this is obviously a pompous windbag who wanted to try and boast about how much they know about games. This is underserving of being /.ed.

Re:I'm rather disappointed.

[Idealius]

Methodical analysis is all well and good but some things beyond human analysis are better handled organically.

His conclusion is organic. Despite /.'s dreamworld, one doesn't have to react analytically to every stimuli.

In true /. form I wrote the above before reading TFA. It's sad because the FIRST paragraph is this:

"Nobody can completely understand the entire field of game design. There are too many interacting elements, too much information, for the human mind to perceive and consider simultaneously. Thus nobody can hope to think about all of game design at once. The only solution available to the designer is to conceptually split the field up into manageable chunks, each of which can then be considered separately." ..which basically drives my point home. Don't blame the author for a confusing dyanmic that exists in video games. He's not making it up, he's describing it. So what if the subject is boring, say that, don't say the article is style-less and devoid of mechanics. You're disappointed because this wasn't the article you expected, critique the article based on it's merits please!!

Re:I'm rather disappointed.

[SharpFang]

"There are too many interacting elements, too much information, for the human mind to perceive and consider simultaneously. Thus nobody can hope to think about all of game design at once"

This is wrong.
Nobody can analytically examine all of the game design at once.
All the rest is possible. Not as business, science, engineering. But as art, pretty pure at that. And like any art, game design has some prerequisites that are hard to put in words, impossible to purchase, difficult to understand. Things like talent and inspiration, letting the ideas "click into place", a rough sketch taking shape as it grows, gaps getting filled by good ideas that come at the right time. Ability to turn your thoughts into words, images, code, action. System way too complex to be understood or analysed by any scientific method just works fine connected by genius of a talented artist. You can't engineer a good poem, a good movie, a good song. The idea of "good" is too complex and too abstract to be understood by science and all attempts to approximate it by ballancing slightly less abstract ideas like "style and content" is bound to fail. With some luck you'll be able to engineer a product that sells well, but it will NOT be a GOOD product, because it will be just it, a product. Maybe with certain "wow" factor. But it won't be a piece of art, something the author put a part of his soul into. Something to make you laugh or cry.
There's no chance for good games without talent of the creator. And no amount of methodical analysis of the market can replace that.

Re:I'm rather disappointed.

[Idealius]

"Nobody can analytically examine all of the game design at once."

That's the primary difference between organic and methodical analysis.

Organic analysis is used for subjects too complex for human's to completely understand methodically.

Re:I'm rather disappointed.

[SharpFang]

Yes. Then why teach failed methods of methodical analysis? Maybe one day we will be advanced enough to analytically examine the process of game development, but for now it's of no use, purely academic problem completely useless for game developers. I wonder who's the target audience of the article - I suppose there are maybe 2-3 people in the world genuinely interested in pushing this kind of analysis ahead, who might profit from it...

little things

[gl4ss]

suck.
some that could be fixed rather easily, too.
like models that keep guns at hip, and then bullets fly out of their eyes. pretty annoying when you see someone peek out and shoot you, while their gun is still behind an obstacle.

Myth's and Facts

[Daysaway]

From the perspective of a Game Developer, there are quite a few myths that buzz about the game industry, and the majority of the complaints about the quality of games these days fall on deaf ears. My favorite is "This game would be so much better if X was done instead of Y". X = a pretty damn good idea which chances are was already thought of, and more times than not, in the original design doc to begin with but had to be cut. Y = what the developers settled for due to time constraints / engine shortcomings / limited funds from publishers etc... The developers of the games you play know that X would be super sweet, but they also have a chain of command which unless followed to the letter results in best case, termination of employment, and worse case, the plug pulled on the title completely. The designers know that enough X's would put the game quality over the top, and change a sub-par 'b' title to a 'AAA'. Unfortunately for games, features are cut early, and cut often to save production time and costs. And in the end, you have just another clone of a game you played last year, and not much to offer otherwise. But strangely enough, this is what pays the bills, and as long as the boss can pay his employees, they can all stay in business. Not that X ideas are bad, in fact they are what keeps the game industry bringing in more revenue than Hollywood year after year. If there were no good ideas realised, there would be no reason to buy new games... but just because a good idea didn't make it into a game, doesn't mean the game was any less difficult to make it into your console. Another myth is that to make games more real, X should be implemented. X = some realistic physics, or models, or anything along that nature. The answer to this is pretty simple, who wants to play "Real Life 2.0?" The reason games are fun, is because they do not follow the real world too closely. This is the same reason action heroes never get hit by the flurry of bullets through chain link fence. Sure, bullets do not fire in a straight line, as per suggested by just about every FPS out there. Have you ever calculated the physics of a bullet? Did you calculate in everything including wind resistance and gravity? And because I know this will reach at least one person who has done this, how long did it take you to do the math? Could you imagine doing it 60 times a second? Even if you simplified it greatly, you can't argue that it will still take up a considerable amount of processing power to do it that often, not to mention the bandwidth to transfer it. This is why a 'ray to cylinder' collision is used instead.. its fast and easy.. and aside from the select few who want physics to feel even more realistic, its good enough for the mass audience.

Re:Myth's and Facts

[Daysaway]

[pwned by the html selection, new and improved comment, now with more carriage returns]

From the perspective of a Game Developer, there are quite a few myths that buzz about the game industry, and the majority of the complaints about the quality of games these days fall on deaf ears.

My favorite is "This game would be so much better if X was done instead of Y".
X = a pretty damn good idea which chances are was already thought of, and more times than not, in the original design doc to begin with but had to be cut.
Y = what the developers settled for due to time constraints / engine shortcomings / limited funds from publishers etc...

The developers of the games you play know that X would be super sweet, but they also have a chain of command which unless followed to the letter results in best case, termination of employment, and worse case, the plug pulled on the title completely. The designers know that enough X's would put the game quality over the top, and change a sub-par 'b' title to a 'AAA'. Unfortunately for games, features are cut early, and cut often to save production time and costs. And in the end, you have just another clone of a game you played last year, and not much to offer otherwise. But strangely enough, this is what pays the bills, and as long as the boss can pay his employees, they can all stay in business.

Not that X ideas are bad, in fact they are what keeps the game industry bringing in more revenue than Hollywood year after year. If there were no good ideas realised, there would be no reason to buy new games... but just because a good idea didn't make it into a game, doesn't mean the game was any less difficult to make it into your console.

Another myth is that to make games more real, X should be implemented.
X = some realistic physics, or models, or anything along that nature.

The answer to this is pretty simple, who wants to play "Real Life 2.0?" The reason games are fun, is because they do not follow the real world too closely. This is the same reason action heroes never get hit by the flurry of bullets through chain link fence.

Sure, bullets do not fire in a straight line, as per suggested by just about every FPS out there. Have you ever calculated the physics of a bullet? Did you calculate in everything including wind resistance and gravity? And because I know this will reach at least one person who has done this, how long did it take you to do the math? Could you imagine doing it 60 times a second? Even if you simplified it greatly, you can't argue that it will still take up a considerable amount of processing power to do it that often, not to mention the bandwidth to transfer it. This is why a 'ray to cylinder' collision is used instead.. its fast and easy.. and aside from the select few who want physics to feel even more realistic, its good enough for the mass audience.

If you want substance...

[Anonymous Coward]

Battle for Wesnoth has lots of subtance. They released 1.0 a few days ago, but the /. editors don't think that the first stable (as in unchanging) release of the highest rated Linux game deserves a story...