I'm only saying that it's a very impressive Engine that COULD be put to good use.

 

I know what you're saying. I'm getting a little too argumentative in this thread and I apologize to everyone here. My thing is that I appreciate artistry in video games. I appreciate video games as arts, as human expression. I've played plenty of games that were made based on incredible technical merit, but lacked artistic merit. And it shows to me. It's why I don't like CryTek's games. This particular engine is incredibly technical, but threatens artistry.

 

I aspire to your sentiments. However, I do not believe that partial automation has to destroy artistry. Given the very real constraints of time and resource on the company, part automation could in fact free up artistic time, rather than forcing creative genies to slog through minutiae.

You won't find me arguing against that point. But nobody here has so far suggested automation for minutiae. Tilesets, game engines, they all feature some level of automation for minutiae. Even the painter leaves some of it to the brush.

 

But there's a difference between using a tool to make something easier and letting the tool be the creator.

Edited December 20, 200718 yr by Tale

To be honest, the "randomisation" is nice awesome, but what I really think could be useful for Aliens or any other space game is the seamless transition between space and atmosphere... sure it's not that hard, compared to the creation engine, but it removes loading screens, and that's a good thing.

Edited December 21, 200718 yr by WILL THE ALMIGHTY

To be honest, the "randomisation" is nice awesome, but what I really think could be useful for Aliens or any other space game is the seamless transition between space and atmosphere... sure it's not that hard, compared to the creation engine, but it removes loading screens, and that's a good thing.

 

I have no problem with loading screens.

 

They did not detract from the amtosphere or Bloodlines, Deus Ex, System Shock 2, etc. Loading screens give the play a reprise and help to "regroup" the focus and tension of the player - instead of the player going on and on and eventually tiring of the area. Being tired of something certainly kills the horror atmosphere.

I believe it was Tycho of Penny-Arcade who brought this up in a way that has stuck with me most. Emergent gameplay and procedural content are an attempt to create fun with an equation. And I think there's a problem with that. My disagreement with it isn't simply of principle, but of experience, I've yet to see a procedural work come out half as beautiful or engrossing as one designed by an artist who knows how to properly use colors, forms, and space.

 

Amen sister.

 

Anyone here played Hellgate London? I didn't play much but oh god the levels were pure ****.

I believe it was Tycho of Penny-Arcade who brought this up in a way that has stuck with me most. Emergent gameplay and procedural content are an attempt to create fun with an equation. And I think there's a problem with that. My disagreement with it isn't simply of principle, but of experience, I've yet to see a procedural work come out half as beautiful or engrossing as one designed by an artist who knows how to properly use colors, forms, and space.

 

Amen sister.

 

Anyone here played Hellgate London? I didn't play much but oh god the levels were pure ****.

 

Once again, you are talking about boring handmade-levels, not procedural.

 

I know that if an engine create billions of solar sistems and planets (like infinity engine does) a lot of planets and moons are gonna be very similar, like in real life. The 90% of extra-solar planets discovered out there, are gass-gigants. Mars its a red desert, mercury its like our moon etc, etc. The big difference its the gravity, the atmosphere pressure and composition, temperature etc...(the infinity engine could make realistic planets whit realistic phisics values in spite off its position from its sun)

The mass efect planets are very, very similar. Hand work dont guarantee you diversity.

Anyone here played Hellgate London? I didn't play much but oh god the levels were pure ****.

 

Once again, you are talking about boring handmade-levels, not procedural.

 

No I'm not, Hellgate London has randomly generated levels. Incredibly boring randomly generated levels that could never hope to be anywhere near as interesting as half-decent handmade levels.

 

The mass efect planets are very, very similar. Hand work dont guarantee you diversity.

 

Compare Mass Effects story planets to the sidequest ones. The sidequest planets can hardly be considered "hard work", and they're closer to what this infinity thing would produce. Prettier blandness.

Edited December 21, 200718 yr by Hell Kitty

Compare Mass Effects story planets to the sidequest ones. The sidequest planets can hardly be considered "hard work", and they're closer to what this infinity thing would produce. Prettier blandness.

The blandness of the Mass Effect worlds had nothing to do with the way the areas were mapped so much as the complete lack of variety re: encounters. You had 2-3 mineral deposits, a crashed probe that provides equipment upgrades, and an "anomaly" that contained one of the collection quest items (data disc / matriarch's writings / IDs / whatev) and whatever facility or beacon / ambush the planet's there to provide. It's safe to assume that a randomizer would include a little more than that.

 

If you really want an idea of what randomized gameworlds would probably yield, you'd probably want to take a good long look at Diablo 2 or even STALKER, a game which the tru roleplayers around here are all too willing to rub down with scented oils and gingerly tongue. If you took away all the annoying bugs and the intrinsically impossible "secure this area for group X!" quests you'd have a good idea of what a randomized gameworld would look like. And it ain't that dreadful.

 

And as far as I'm concerned, the "you need a true artist for environments" reasoning seems to have more or less died out with 2d gaming. Where you want to impress, have somebody design a level. Surround it with randomized environs, and you're looking at a decent synthesis. I'd wager as the years wear on the weird inconsistencies in random generation could get smoothed out and a palatable product could be made, provided the interest and effort is there.

Edited December 22, 200718 yr by Pop

I think the Infinity engine could be used to produce a large number of good looking solar systems and planets. But it would still be necessary to add that human touch to each and every interesting spot on those planets. Just look at Daggerfall: the only interesting dungeons/cities were the ones made by hand by a human artist. The same applies to everything in every game. Random generation never produces anything interesting, but it can be a great asset to ease the work for the artists as they don't have to do the menial task of trying to come up with different cool-looking planets and instead can fixate on the micro-managing of colonies and detailed stuff.

 

A balanced mix between hard work and procedural generation would be the best.

What about a mix? The game Chaos Gate had main story missions and subsidiary random gen missions.

Whatever the use for random generation might be, it won't make the game a hit regardless of how well made it is. The true essence of CRPGs (say, Torment and BG2, for example) can probably never be randomly generated. I mean, come on. Why do you like a RPG? Because it has infinitely many random- generated areas in which bland random- generated side quests take place? No, I don't think so. Let's leave random generation to action games like Diablo 2 (although I admit it has some RPG- ish parts), games whose point does not go beyond anything which a computer could randomly generate.

 

I do not necessarily disapprove of a limited use of it, but it is important to be aware of that the fact that an area is randomly generated also means that it must be equivalent storywise to any of the other iterations of it the player could possibly encounter. Of course you could have a game whose story would be dependent on which area was randomly generated, but that would limit the game to a smaller number of randomly generated areas, which would basically make it pointless to have random generation at all.

I don't think it's as clear cut as you're saying, Rostere. To use a slightly different (and hopefully illustrative) example, consider NPCs. Generic NPCs are all very well, but they're a bit like wallpaper. instead, what I use in my pen and paper games is a quick NPC generation sheet, with some character quirks and a couple of problems. This makes them partially random, and occassionally more interestng than the campaign NPCs.

 

The analogy would be potatoes. A soup consisting entirely of potatoes would be pretty Russian boring. But add potatoes to a soup in moderation and you make it hearty and filling. I'd suggest that a judicious quantity of randomly generated and hand finished material could bulk outa game very very nicely.

True enough, Wals, but most of hand-crafted areas are left as that 'soup' anyway.

 

I really think the way to go in most games is randomly generated areas heavily crafted by hand, so you don't have to start from a flat grass ground and make the entire map just slightly bumpy with trees and weeds.

Back when I DMed I used to use a fractal computer program to randomly generate terrain and then go in by hand and insert whatever I needed or whatever striked me as useful at the time. Of course, that was a pretty simple process, and all I needed to render visuals were words, but I think you could do something akin to what I did on a larger scale for a game.

 

It would beat the S-shaped levels that are in so many games these days, at least. Donkey Kong is far too influential.

Edited January 8, 200818 yr by Pop

Back when I DMed I used to use a fractal computer program to randomly generate terrain and then go in by hand and insert whatever I needed or whatever striked me as useful at the time. Of course, that was a pretty simple process, and all I needed to render visuals were words, but I think you could do something akin to what I did on a larger scale for a game.

 

It would beat the S-shaped levels that are in so many games these days, at least. Donkey Kong is far too influential.

 

 

They're S-shaped to conserve memory.

You're S-shaped to conserve memory.

I don't think it's as clear cut as you're saying, Rostere. To use a slightly different (and hopefully illustrative) example, consider NPCs. Generic NPCs are all very well, but they're a bit like wallpaper. instead, what I use in my pen and paper games is a quick NPC generation sheet, with some character quirks and a couple of problems. This makes them partially random, and occassionally more interestng than the campaign NPCs.

 

The analogy would be potatoes. A soup consisting entirely of potatoes would be pretty Russian boring. But add potatoes to a soup in moderation and you make it hearty and filling. I'd suggest that a judicious quantity of randomly generated and hand finished material could bulk outa game very very nicely.

 

Sure, random generation could assist in creating interesting content. Thing is, when you're a DM you will probably improvise dialogue, simulating the personality of the random- generated NPC. Computers however, can't do such things yet. In your adventure, this NPC takes as large place in your story as you want it to, but in a RPG of today he/ she would probably give you a quest like "Go to (X; Y), kill the monsters, fetch the item", and for the rest of the game only offer the player one meaningless sentence about mudcrabs. Now, you say, couldn't you make a more complex random generator that allows for more interesting NPCs and quests? Sure, but then it becomes a question of prioritizing. You can't have it all, and why waste time doing a such a complex random generating engine when you can make handmade content which is relevant to the story? I really think that all so- called "side quests" should add something to the game. Perhaps a section with a different mood than the rest of the game, or another aspect to the main story. No game should need dull side quests only for the purpose of leveling and farming gold. At least that is not what I find exciting with RPG games.

 

I agree you can use random generation to create realistic trees, to rapidly create "generic" environment and to get inspiration for NPCs, but it only helps to create the background... For example, if the colour of the clothes belonging to the 13258785 inhabitants of Sigil is completely irrelevant to the story, you can use random generation. It will look more natural than if everyone was to wear clothes with the same colour. But could you random generate Torment's story? I don't think so. Now, the point of my last post: which is the reason for buying a game? A realistic but shallow scenery? Or a well written story?

 

Ideally, a game should of course have both. But you can never have it all. And of course it's also a question of which kind of games you like. If you prefer Morrowind over Torment for example, I guess we simply see things differently.

Filthy creatures.

Filthy creatures.

 

Aliens shouldnt be random because this isn't diablo 3! Who agress with me?

Good comeback, Rostere. Thought provoking, particularly re priorities. I guess my own comeback would be that for one game you'd be right. But if you are a games making company, then having a plot/mission generator could be a good investment that would pay off in the long run. What you'd need is a realy first class literature professor to give you the various subclasses of plot, a few soupy twists, and some tactical boffins to give you the sub-classes of bad guys*, plug in a level generator and you'd get something which would - as has been said - need editing rather than composing from scratch.

 

The analogy, which i can't recall using yet, would be the TRIZ toolset in engineering. It doesn't replace a good engineer but it massively speeds up teh job of being creative and makes you consider more stuff than you would otherwise.

 

*Not in just random race x, but in terms of what it means to the fight, so ranged, stealthy, ambush etc.

I believe it was Tycho of Penny-Arcade who brought this up in a way that has stuck with me most. Emergent gameplay and procedural content are an attempt to create fun with an equation. And I think there's a problem with that. My disagreement with it isn't simply of principle, but of experience, I've yet to see a procedural work come out half as beautiful or engrossing as one designed by an artist who knows how to properly use colors, forms, and space.

 

 

I enjoy procedural animations. The tech demo I saw for Indiana Jones at AIIDE looked fantastic.

Procedural world content is great... Though it only really works in games which are basically sandboxes.

 

I've spent many many hours researching procedural content generation, it's mostly fractal based if not 100% based upon fractal geometry. It's excellent for storing lots of information on a tiny amount of disk space, the trade off is that everything needs to be generated, so the load times and memory requirements can be impractical.

 

At the end of the day it simply makes sense to me to generate certain features procedurally, tree's, rocks etc... Naturally occuring data, also AI driven procedural animations are very very cool.

I don't know what Tycho said, exactly, but the idea that procedural graphics can never match real artists is simply wrong. Two points:

 

1. Real artists don't all know what they're doing. Given the choice between mediocre artists who produce bland and uninspiring "art," and a procedural graphics engine that is capable of at least the illusion of beauty, I hope devs are choosing the latter. It's not just about saving costs - it's about ensuring a minimum quality to the visuals while allowing you to focus on the things that you're better at - such as story, gameplay, etc. Sure, if every studio had ace artists who create amazing art and don't just do what everyone else does, I'd be more hesitant to make this claim, but that's just not the case. Humans are not necessarily better than machines, and this will only become more true as technology improves.

 

2. Nature is procedural. Is anyone going to argue that nature is not inspiring, or capable of taking our breath away?

 

As for other procedural forms of other content, it depends. One could argue that in order to create compelling procedural content vis-a-vis gameplay, the computer would have to understand enough of human psychology to know what is entertaining for us. That could be hard, though it's not really been attempted. Audio generation (sound, music, even voice), on the other hand, does not seem exceedingly difficult, particularly given what passes as "good music" in most games. Natural language generation and storytelling (the latter requiring a in-depth understanding of aesthetics and dramatics) are, on the other hand, decades away.

 

Perhaps I'm a bit biased, but the truth is, I don't see any reason why procedural content, in principle, cannot work. As games become more complex, it is only prudent that more of the grunt work will be offloaded to machines - they have to be. Afterwards, it's only a matter of time before the higher level work is automated, too. In the end, we may be responsible for nothing more than the story and character concepts, and perhaps, before long, not even that.

Edited February 22, 200817 yr by Azarkon

Azarkon, You're absolutely correct!

I'm not trying to start a fight, but isn't he contradicting you?

I don't know what Tycho said, exactly, but the idea that procedural graphics can never match real artists is simply wrong. Two points:

[stuff about nature.]

An important question though is whether the real world could be fun (remember that was what Tycho was attacking) if imported into a computer game or whether the human nack for artistry is important.

Also simply calling nature procedural

Real artists don't all know what they're doing. Given the choice between mediocre artists who produce bland and uninspiring "art," and a procedural graphics engine that is capable of at least the illusion of beauty, I hope devs are choosing the latter.

Perhaps a better more realistic option might be to try and get some better artists? Unless of course your definition of mediocre artists are the average game-artist of todays world.

... the computer would have to understand enough of human psychology to know what is entertaining for us...

Do we even really know that ourselves?

As games become more complex, it is only prudent that more of the grunt work will be offloaded to machines

Noone argued that grunt work or minutiae shouldn't be done by machines, it is the important stuff (however one might define that) that needs artists.

 

Somehow your post seems very concerned about what could be possible in the future, but lots of stuff can be possible in the future, and since the future is uncertain the procedural future might never come.

 

Procedural world content is great... Though it only really works in games which are basically sandboxes.

Very true, and people should remember that this is not a sandbox like some call oblivion (or any other recent elder scrolls game) but stuff like daggerfall which is horribly bland compared to modern games.