Comparing PC Game Physics

John Callaham writes "On Wednesday we posted up comments from Havok about rival AGEIA's use of their physics processor in the PC version of Ghost Recon Advanced Warfighter. Today we have an expanded article with point-to-point comments from AGEIA that address Havok's statements." From the article: "How much interaction do you want in your PC games? It used to be that graphics were the number one factor in picking up a new game but now players are asking more and more about interactions in the environment. One company that has provided such interaction is Havok. They have developed a physics engine that has been used in a ton of games, including most famously in Valve's first person shooter Half-Life 2. Recently, Havok announced plans for a new physics engine, Havok FX, that would use Shader Model 3.0 graphics cards to further enhance game interactions and physics."

Comments

On physics

[remembertomorrow]

How much interaction do you want in your PC games?

Interaction is great and all, but please give humanoid NPCs more rigid joints! It looks silly seeing them flopping around with elastic joints, or doing backflips after being shot in the face.

That, and being able to move enormous metal crates simply by shooting them, breaks any immersion the game has created. :/

Re:On physics

[grammar fascist]

Interaction is great and all, but please give humanoid NPCs more rigid joints! It looks silly seeing them flopping around with elastic joints, or doing backflips after being shot in the face.

Hear hear!

I was watching a coworker play Unreal Tournament, and I had to work to keep myself from laughing every time a player got killed. It looked like someone tossed a dummy.

Also, don't forget that every person in a modern shoot-em-up is nothing but a bag of blood. They must be - it seems like 25 gallons get spilled every time a player gets maimed.

Re:On physics

[Khyber]

I own two Colt AR-15, several shotguns, and no pistols. I can bet if a .45 can make you spin circles by hitting you in the shoulder, a larger round with more power behind it could very well flip you backwards, depending on where you hit. Try firing an 8 gauge. If you can stand up to the recoil, that is.

Re:On physics

[It'sYerMam]

Perhaps the point is that keycard/maze puzzles are old-hat, and that blowing your way through the building is far better. A lack of intuitivity plagues games in this way. Like on 007 Nightfire, I think, where you had the laser watch that could cut through STEEL, but an enemy wouldn't even blink if you laz0red his eyes.

If you give the player rockets, then a simple way to encourage them to use them properly is to ensure that they don't have enough to waste taking out scenery. If you make sure resources are limited enough to force the player to use them only where necessary, then you can still have your godawful keycard puzzles.

Re:On physics

[nugneant]

Allowing the player to remove certain obstacles (in this case walls) necessitates new obstacles the player cannot remove easily in order to prevent the player from blowing a straight line from start to finish. Currently they're using walls and the like to stop you from going somewhere, if you could blow up the wall then they'd have to think of something else, probably a huge sea of fire or something.

And yet again, allow me to restate my central point: the genres aren't stale. The minds behind them are. Rather than brute-forcing everything into compliance, how about something creative? Like along the lines of:

* If I choose to blast a straight line from start to finish -
* And if I do not get caught in the falling debris-
* Then this will no doubt trigger an minor institution-wide event-
* And therefore I might find a platoon of soldiers swamping the start-point of the next level-
* Which, having already made the decision that a "fun experience" for me is the Rambo Tank of Doom approach, will probably be a fun challenge when I whip out my chaingun and mow down as many as I can while frantically fleeing for some cover of some sort.

This can be as simple as I made it, or made even more complex / interesting by incorporating RPG (Role Playing, not Rocket Propelled) elements. Fuck that System Shock stuff - that's for toddlers. How about a system where each gun weighs X units. Where the more units I have for carrying capacity, the less able I am to sneak about (given that a 300 pound muscleman can't prance about like a sleek Ninja of the Night). Where a rocket launcher, a pistol, and a chaingun would pretty much max me out - but you could have your pistol, your dual-pistol, your glock pistol, your AWP, your throwing knives, and your ballerina tutu or whatever helps you prance from choke point to choke point. :-D



As for the telephone pole of doom, well, the Gizmondo CEO found out the hard way that telephone poles are indeed instant doom.

Hahaha... +1, Funny. This is why I love discussing videogames, because I find that I go from neck-and-neck "yo mama so fat" low blows (see the last line of my above paragraph) to fucking laughing out loud.

Anyway. Point well taken.



At some point the FIFA games allowed unnecessary brutality, it was removed for later games. Probably because people were abusing it.

Or possibly because FIFA wanted to make a family-friendly image, similar to the NHL's misguided efforts in the mid-90s. Solution, of course, being - who needs licenses? I remember the days of the NES, where Bases Loaded was king and champion, Baseball Simulator 1.000 was the fun alternative, Baseball Stars was the Otaku-favorite, and the MLBPA licensed RBI baseball was usually "top of the second tier" at best. And the only (that springs to mind) baseball game licensed by the MLB? Everybody agreed that it blew goats (MLB, by LJN, possibly the worst baseball game of all time - even Jeff Rovin hated it. It's so bad I can't even find a Google result for it that isn't COMPARE PRICES BUY SELL TRADE. Some things are truly best forgotten, it seems).

I fault game companies not willing to take risks, burnt out programmers unwilling to fight for what they know is right, and stockholders who are paranoid about profits. In that order.

Another note: STOP the POV summaries.

[Avillia]

A 'comparison of PC Game Physics' should not have a summary obsessed with one technology and one company (Havok).

growing older

[Deep Fried Geekboy]

I find myself buying fewer and fewer games as time goes by, and I believe it's thinking like that that really shows why.

mmm... have you controlled for 'growing older'?

quite a significant variable

btw, those games you think were so great? they aren't.

I still have fond, fond memories of the original UNREAL TOURNAMENT and have been sorely disappointed by subsequent releases... and yet when I go back to play UT1 I can't stand it... it pales in comparison to the more recent versions, even though the underlying gameplay is better.

Re:growing older

[Sparr0]

What? If you think Fallout and Starcraft were not in the top 5 games of their years, and among the top 25ish games of all time, then you really don't belong in a discussion about the quality of games. This is not a game preference thing; it can be said objectively that these games embody everything that can be good about games in general, and specific to their genres. They were revolutionary, evolutionary, spawned good sequels (WC3 is a functional sequel to SC, not WC2, regardless of the story), sold insanely well, and pretty much cleaned up by any other metric you care to apply.

Re:growing older

[Tim C]

You do realise that Starcraft is still insanely popular, don't you? They have televised tournaments in Korea - people not only play the game for money, but other people (lots of other people!) watch them.

Not too shabby for a game that's what, 7, 9 years old?

Re:Missing the point?

[AndyAndyAndyAndy]

Graphics was never the number one factor. Great gameplay was. This type of "graphics makes the game" type thinking came along with stuff like Far Cry, FEAR etc. In fact it almost seems like the game content was an afterthought.

Yes, and these games SOLD because there are still MANY people out there who DO think graphics make the game.
Not that I do, but you certianly can't say that graphics isn't the most important factor to some (if not most) people. Don't be daft.

Wake me up when...

[ecorona]

Wake me up when a game world isn't a static 3D environment. Wake me up when I can walk up to any tree, pick off a branch, chop the tree down, squish some ants living on the tree, and can rip a moist leaf on the tree like a sheet of paper. Wake me up when I can knock down a building, wall, and can permanently remove bricks from a house. I want to be able to drive a car through a wall, have grass that actually grows, and can cause wildfires (just like in real life). I want to be able to take some sand from the beach with a bucket and pour it all over the nearest NPC and see all the little grains of sand stick to his shirt. Wake me up when it's time because I can't wait to play. Imagine MMORPGs where you can actually DIG A SECRET TUNNEL underground to invade your enemie's territory. Imagine being able to dig holes to hide in and cover them up with leaves. Well, you get the idea. Possibilities are endless. Seriously, how long do you guys think it'll take for some crude implementation of what I listed above comes to fruition?

Re:Wake me up when...

[Musc]

So you want a completely detailed model of the world, down to bricks and individual grains of sand?
You want it all be simulated with physics so that you can interact with everything in a plausible way?

Well, I can tell you that any one of these things currently is a struggle to get to work at all,
even assuming you are willing to wait hours per frame. You want a pile of thousands of bricks
falling into a pile, with correct collision detection? This is an area of active research.

http://www.cs.cmu.edu/~djames/

You want the water on the beach to swirl and splash?
How about a piece of paper that you can burn?
Again, a challenging set of problems that we are just beginning to solve in a way that looks good.
http://graphics.stanford.edu/~fedkiw/

How about the snot you pull out of your nose?
You want to pick your nose and have the snot squish in a gooey fashion?
We can do it, but just barely, if you want to wait all week for a few seconds of animation.

http://www.cs.berkeley.edu/b-cam/Papers/Goktekin-2 004-AMF/index.html

Now, what you want is to combine all these simulations, plus many more.
Also you want it to run in real time on a desktop PC.

I predict we will have this in 50 years, and that is being extremely optimistic.
If Moore's law is really ending, then maybe much longer.

Hardware physics cards may be just the thing we need to make it possible one day.

Re:Wake me up when...

[Anonymous Coward]

Don't hold your breath.
Things like that have been doable in muds for a long time. They're not implemented often, though, for a good reason. "Combinatorial explosion" is what it's usually called. The possible actions the player can take begin to overwhelm both the game's system and its programmers, dragging everything down and breaking things. Tying a rope to a tree is fine, but what if the player wants to tie the rope to an ox, hold the other end, point the ox at a cliff, and kick him while wearing a set of wings he carved out of wood from a tree? At this point your game's programmer is breaking down in tears, realizing what ELSE that would mean is possible, and must be taken into account.
  In a single player game you can have this sort of freedom (well, a programmatically reasonable amount thereof), but multiplayer is a serious problem as this makes game balance nearly impossible. (See: Nethack. A sufficiently skilled player can quickly become a seeingly unstoppable engine of death due to the vast array of tricks at his fingertips)

Re:Wake me up when...

[zokrath]

It depends on the amount of abstraction that you are willing to accept. Game physics are currently focused on accuracy rather than results, which is why ragdolls go haywire and objects get stuck into corners and bounce out of the world at relativistic speeds.

A bucket of sand is a bucket of sand, so if you get a bucket from the beach, it does not matter exactly which grains of sand wind up in the bucket, or on the NPC. Thus, you 'scoop', and wind up with x cubic centimeters of sand in your bucket. You dump the bucket on an NPC, and the NPC gets covered in random sand particles.

A wall can easily be abstracted to a single entity, rather than the individual components. Drive a car through it, and the associate component's properties are used to determine how the wall breaks up into smaller pieces, but again, it really does not matter exactly how the wall breaks, or exactly how the bricks scatter. What matters is that there is now a hole in the wall, and the wall is now divided into smaller discrete objects and a mess of random bricks. Because the discrete entities do not form a solid body from the top of the previous wall to the bottom, they no longer offer support to whatever is above, potentially causing a chain reaction.

Digging a tunnel in an MMO is problemtic not because of technology, but because of other players. An engine could certainly be developed that allowed for the construction of tunnels, with location-based criteria for starting a tunnel, and valid tunneling areas defined underground. It could even have advanced engineering aspects such as shoring, cave-ins, and flooding. But the more robust you make a multiplayer system, the easier it is for one individual to ruin it for numerous others.

The idea of sneaking into the enemy's tunnel system and causing a cave-in is certainly enticing, and would be filled with peril and what not, but what about someone on your own team going down and causing that same cave-in, due to malice or incompetence? And where does all of the excavated dirt go? Player made mountains are perhaps of even greater concern...

Taking leaves from trees is reaching into assinine territory, but tree limbs are perfectly reasonable, and chopping down trees has been done many times. Perhaps not with molecular simulations of axe versus wood, but why bother? You hit the tree at a given location, it gets a notch. You hit the notch, and it gets bigger. Once ht notch is big enough, the tree falls over. If you spread your swings around, you get a bunch of little notches. Accuracy, strength, and technique could all be factored into the one end result, the goal of chopping down a tree.

Re:The Physics of Brick Out

[nugneant]

This, actually, is a perfect counterpoint to the "realistic physics are ALWAYS better" line of thinking.

If it weren't for these deliberate anomalies, Breakout, et al, would be thrown into "loops". I remember a port of Breakout for the TI-83 graphics calculator that suffered from this - you would eventually have the ball at such an angle that no matter how you hit it, it'd always travel along the same pattern.

Face it - even today, this applies. Would it really be fun if your character could only jump 6-12 inches off the ground? If you ran at a rate of around 20MPH? My stipulation is that it would not be. Game designers must fudge the physics to keep a game playable. And frankly, I find the physics of Mighty Final Fight for the NES to be light-years ahead of the supposedly "revolutionary" physics of, say, Trespasser. More complex != more funriffic.

In defense of graphics

[cgenman]

The point is not the gameplay. The point is the experience. If your experience is reduced if you can't get over the fact that the graphics look bad. Or don't evoke the images they are supposed to evoke. Sure, Starcraft has a limited visual experience now, but A: it was amazing for the time and B: there have been a lot of amazing games released since which players just couldn't get into the experience because it was a cheap, unbelieveable 2D sprite engine. Certain games it works for, but to get into the experience of others, you have to kick it up a notch visually.

The same can be said for movies. Anyone can do Clerks. But nobody can do Titanic without a large budget going to visuals. Anyone can create the next Tetris. But nobody can create the next Final Fantasy without reasonably engrossing visuals and expansive, expensive vistas.

Which is not to say that gameplay isn't important. It's just that people know (or think they know) how to do amazing visuals, but nobody knows how to make amazing original games. Even Blizzard, a consistent hitmaker in the industry, basically takes existing genres with major flaws, fixes all of the flaws, and throws in a ton of aesthetic polish.

Now as a side note, you can get a hell of a lot of bang for your buck out of good sound, especially considering how few people do. Sound is subliminal, so it frequently gets forgotten when budgets are getting allocated. But you can spend months prototyping and sketching and modeling and mapping your main enemy to make them seem as massive and powerful as possible, or you can get a sound engineer who will mix a bowling ball dropping onto a piece of steak with someone punching through aluminum foil, and getting the most amazingly visceral reaction from the audience after one afternoon of experimentation.

Look at Oblivion.

[Vo0k]

Most of people asked say Morrowind was better than Oblivion.

What could make Oblivion better, or at least equal to Morrowind?

These:
Better grass distance?
More details in the LOD (distant textures) area?
More objects covered by the physics engine? (furniture, rocks, plants)
Items possible to shatter, smash, break, dent?
Containers displaying their content in 3D and not in 2D menu?
Better voice acting?
Books that burn?

Or maybe these:
Less linear quests not forcing the next step on you?
Shorter load times of locations?
Not removing levitation, slowfall and a dozen other classic spells?
More factions to join, interesting quests?
Dialogues and text that always makes sense, never seeing hearing the same thing less than 5 seconds apart?
New, interesting books you haven't read in Morrowind already?

Re:Crowded House lol yea, wait for mp processors

[Anonymous Coward]

we are going dual core, soon quad core. let the cpu take the load of physics. it might be slower but it will improve and its more than enough for most people. i think most of us gamesr would rather spend more on our cpu's than some 300 dollar physics card that is useless for anything else but gaming. the need just isn't there. we are already tapped out paying for huge lcds and expensive video cards.

and yea the cell is a joke. at first they claimed it would do the graphics. then they came back to earth and tacked on a nvidia gpu for the ps3:P