Slashdot Mirror
The Hard Science of Making Videogames
twoblackeyes writes "PopSci delves into the 10 greatest technical challenges faced by game developers today, and the technology that will hopefully make them a thing of the past. At the top of every dev's wish list is increased realism: realisitic fire, water, enemy AI, material physics, etc. Here directly from the developers where the tech stands today, and where it will likely be tomorrow. '4. Artificial Intelligence - Problem: Once upon a time, the bad guys in videogames wandered around mindlessly, shooting at you while they waited to die. That doesn't cut it anymore. Players demand sophisticated enemies to fight and reliable in-game allies with which to fight them. Thing is, it's freaking complicated, and it eats up processor speed. "We're faking just enough smarts to make it work," says Mathieu Mazerole, lead engineer on Ubisoft's Assassin's Creed. Status: Imbuing characters in a game with lifelike decision-making ability involves employing the kind of high-level logic theories--learning decision trees, mobile navigation, finite-state machine models--used by top robotics engineers.'"
would rank the importance of realistic water simulations above the importance of good artificial intelligence in games.
And to think, I used to subscribe to popsci...
using namespace slashdot;
troll::post();
On the contrary, it is both. However, the topics discussed here are the science portion of it. Things like the textures, character design, music, scripting, etc. are the art of it.
http://blog.heavensdomain.net
The article should be renamed from "10 greatest technical challenges faced by game developers today" to "10 greatest technical challenges faced by first person shooter game developers today" Contrary to popular belief, not all game developers are striving for photorealism.
You'd think with the success of casual games, and less "technically advanced" games for platforms like Wii and the web, game developers would see huge market of gamers who are simply looking for games to be more fun again. Who cares if it has the latest AI or better cloth physics? Leave the better cloth and fire effects to the SIGGRAPH people :).
It's just that even if they solve all these issues with games, and are able to render a true to life simulation with AI that can mimic a real person. It still doesn't even begin to solve the most important aspect of video games, which is FUN.
Fun cannot be realized through more processor power, better looking faces or AI.
All these problems are very hard to get 100% right and all they really need to do is to get it right enough that people pretend that they are in a fantasy world. Which is why old text based games like Zork can still do a good job of pulling a player into the world enough that it's fun. It's all about stimulating a person's imagination, not creating a photo realistic simulation of reality.
Sure it would be neat to have photorealistic fire that can burn the entire environment and interact with water in a realistic manner. But is that what you really want? I mean it's also very fun to be able to play as the human torch, which means that you have to bend the laws of physics in your game world to simulate such a being. So it's not really about being real as much as fun, no?
Art direction, character design, level design, are much more important then these issues, yet we are spending much more time on the motion capture of a video game then on the plot. Crazy, IMO.
Dance Dance Revolution is goofy as can be, and I don't play it, but as a "game" it's a lot more fun and interesting than the anti-utopian fascist horseshit that passes for fun these days.
In fact, the Wii opens up a whole 'nother wonderful can of possibilities, as does Guitar god games an similar things. THAT'S where the creative action is, and that's wher ethe REAL innovation is going on. Not in stupid 12 year old boy shoot 'em up bullshit.
RS
Shoes for Industry. Shoes for the Dead.
would place the importance of graphical detail as shown inside above gameplay. If you don't have a good gameplay, you can have 1000000000000 triangle per square inch of pixels, it won't make your gamer more happy. The Primary challenge of a good game is not graphism. it is a good idea which translate to a good technical gameplay.
C. Sagan : A demon haunted world:
http://www.amazon.com/gp/product/0345409469/
visit randi.org
The article claims "hard science" but instead is a collection of blurbs that read like half-assed filler written by someone without a clue as to the subject.
Coding with assembly is like playing with Legos. Coding an application in assembly is like building a car with Legos.
This list is about making games more real, which doesn't necessarily mean better. There more to it, such as balance, game play, user interface, premise, and plot.
I'd still rather play NetHack than any MMO game, and I enjoy the early Final Fantasy games more than the later ones.
And her processing power is pretty dim compared to a computer.
Your cat is smarter than you realize. Brains do image processing, task/goal tracking, fine and coarse motor control, and a myriad of other complex processes simultaneously. Many of these require advanced (but not abstract) mathematics and ability to react to the result. Example: if I'm moving this speed and the dog is chasing me at that speed, will the dog catch me before I reach the house and bite off a chunk of my tail? If so, run up the nearest suitable tree.
It's been said repeatedly that "the most powerful supercomputer in the world is approaching the complexity of X". Currently, X is somewhere between a snail and a housefly. When it gets to housecat, you won't be able to win the games at all anymore.
They're focused on the wrong things. How about focusing on something fun.. Tetris is fun because it's just fun, not because of some stellar graphics and the AI behind which piece comes next. the same goes for Zelda or Final Fantasy or Mortal Kombat.. These games were all fun even back in 8-bit or 16-bit days......
I find it amusing that an entire piece about the "hard science" challenges in game development doesn't even deal with the development process itself. I mean, when you code in a higher level language, dealing with strings is now easy, whereas it was tedious in C. There's no library sharing like in every other language. There's no #include , whereas there's really nothing new to moving a bitmap anymore.
Even with something like OpenGL, you're still basically given a pile of bolts, beams, and sheet metal and asked to make a car. If I had a nickel for every time game developers reinvented the wheel, I'd be Bill Gates. Heck, I'm still coding font routines and sprite handlers for companies. I heard that even the Wii doesn't have a system level call for the main menu stuff...it leaves that up to you.
Someday, the tools will come along enough that people will be able to work with something higher level like Python or Ruby and not have to worry about twiddling their own framebuffers. We're still in the dark ages in game development this way. Having a CPAN for games is DECADES off. Instead, game developers are stuck trying to make a rock fall or a torch look right, and when they're done tweaking that crap, THEN they remember they have to make a game, not a shadow simulator. Thus: Doom LXXXVIII.
All animals, man included, spend their whole existence building up "logic trees" which give us our reactions in certain situations. Someone throws something at you, you block it, dodge it, catch it, or let it bounce off your head. We categorize, extrapolate, and induce...It's all logic trees.
Cats are cursory hunters; they lack stamina, and they hunt by stealth and lightning attacks. This being the case, knowledge of their range is critical to their success as hunters, therefore they spend most of their non-sleeping time engaged in exploration. That's why cats are always into stuff. Startle a cat, and he's gone, under something, up something, behind something. Do you ever see them stop and think about where they're going? They already know.
And the processing power isn't dim. Jesus, just because it's not sitting there doing philosophy its not a highly specialized and successful hunter? You're talking about an animal with enough instinct and reflex control (and that is brain power, just as much as problem solving) to do acrobatic things that people strive in vain to accomplish, calculate a thousand variables while flying through the air to snatch something, or land on its feet after a drop of meters, and we can't even get robots to accurately match the range of mere human dexterity. It's like the DARPA robot challenge; it took years to get a robot through that course, and it's not because the vehicles couldn't make it, its because the processing power wasn't up to making the decisions needed to get across the terrain.
You can't look at intelligence in pure terms of math. When you take them out in the world, and tell them to apply that computational power to walking, talking, and chewing gum at the same time, you see how far we still have to go.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
1) So now the enemy only throws grenades when he's out of bullets? You're missing the point. You gave a very simplistic "If he's within range, throw grenade" explanation. If you use something that simplistic it's predictable and exploitable. More complex scripting is better AI...this is the easy shortcuts you outlined originally aren't groundbreaking.
2) Again, simplistic AI. Move one player into range that can handle the grenades and the AI will switch to lobbing grenades at him while his allies shred him. To be realistic the AI would need to assess the situation and prioritize targets as well as which weapons to use on those targets.
3) So flanking, pincer attacks, getting the high ground, strategic smoke concealment, etc.. are governed under your "Charge or Cover" mechanic? Again, now you're talking orders of complexity higher to achieve anything other than a "wait for player" or "move along quickest route to player" routine.
4) So a scenario should never even be considered because it means the player is already outclassed? That's ridiculous! What if it's 4 on 20? You're outnumbered and the bad guy pin you down with covering fire while they evacuate a VIP. As they pull out of the room they toss some grenades behind them to keep you in cover and delay your advancement. This is exactly the sort of real world tactics missing from games today. With your version they would either run at you and shoot, throw grenades in a confined space or take cover and wait for you. 3 options.
5) The designer should be able to do whatever the hell he wants. What if you have a compound filled with armed guards, who have grenades. You sneak into and get close to your objective, a large reactor, without being spotted. You get spotted though and guards are quickly called from surrounding areas (rather than magically appearing). Should they just start lobbing grenades next to a reactor? Probably not. Again, it's not weakness on the developers part, it's realism. What are they going to do, drop the grenades before they come in? Or maybe in your game non-grenade carrying enemies will just magically appear...
6) Okay, so if there's ever a tank, the designer's an idiot... I guess that rules out a LOT of WWII scenarios then! Oh, and nice job of completely glossing over the need to prioritize weapons!!!
7) Fine, you have a sum-machine gun and he had an M60. You don't need to "adjust" much on an M60 in order to fire it effectively, and it's clearly superior to an SMG in terms of shots down range, but not in accurracy. Some situations it would be good, others it wouldn't.
8-10) For you last couple of points, let me illustrate it like this. You have three people standing in a line, when a whistle blows they either jump up or crouch. That's multiple different combinations, but of the same behavior. They don't automatically start flanking and using squad tactics because they can randomly jump up or crouch, they just randomly pick between some pre-scripted events. The overall effect though is that some guys are jumping and some are crouching...big whoop.
"Life's short and hard, like a body building elf." -- The Bloodhound Gang
"It was a cheat that only looks good close up if nothing much is happening. But why don't you get the wakes from the patrolling boats out there rolling in or making it hard to stay in your own boat? If you blow up a helicopter and the pieces fall in the water, where are the waves from that? That's right, they're not there. And that's because the physics of water is Really Really Hard."
Who cares about that level of Water physics? I mean seriously. It adds practically nothing to gameplay, the physics in many games are merely gimmicks, take Bioshock for instance. The water EFFECTS are very beautiful and interesting but you don't need to model the physics of water for it to be immersive. The art, color and sound, and atmosphere built up by the game is far more important then the physics of water.
I mean really... look at all of the best games we've ever played... Take God of war for instance, basically an action-platform game, one of the best games I've ever played who's AI was not overly complicated, but a game that allowed a lot of freedom of action and had well built combat system, tight control, withimmersive camera angles while fighting, all set in a fantastic 'plastic' world in which you were for the most part 'on rails'.
The game EXPERIENCE is most important, what kind of emotions and stimulation you're giving the player matter more then something totally irrelevant to the experience...
When we think of the best games we've ever played, we're not thinking about -- woah, the water physics in that game was 'awesome', we're thinking about something ENTIRELY different, unless the game is based around someone having figured out a way to make it integral to the FUN and enjoyment someone gets out of a game.
I know people have said this every year since 0, but graphics are fine these days. The biggest improvement in my view would be animation/lack of rag doll-ness on the rag-dolls. Animation tends to be either really over-the-top or really static, sometimes varying between the two in a single game. Even some films have this problem (I'm looking at you, Spiderman).
But AI is surely where it's at. Current AI is terrible, and I think most people confuse "hard" with good AI. They just up the accuracy people! That doesn't mean it's more intelligent, it means it's less crippled. Sports games can be even worse - I just started playing a copy of Pro Evo 4 (okay, not new...) and it's shocking. The basic strategy of the computer players is to stand still and look at the ball. Or if there's really a risk of the opponent scoring, run away. And that's a game with a good reputation. Infuriating.
sam brightman