World Exploration...Yay or Nay, Big or Small

[galsiah]

...but adding up lots of details like these can make gameplay a much more memorable experience.

And adding lots of details NOT like these can make it memorable for the right reasons (in an RPG). Better to spend time on details which matter, and on mechanics which aren't needless nonsense.

...this is exactly what I've been talking about with food; its not about realism, but an abstract and more a way to take something seemingly mundane and make it have value more so...

And I'm still not seeing how reward for mundane-activity-X makes a game any better unless that activity is intrinsically entertaining, or is the only way to introduce a mechanic which creates entertaining gameplay.
This simply is not the case for any "bathroom break" system. It sucks in any game that doesn't either consider toilet humour an asset (fair enough in some games), or is based in a setting where a "bathroom break" is a significant event (I've never seen one).

In a semi-serious RPG, it's a pointless, incoherent conduit for an unremarkable mechanic which could be introduced in more effective ways.

Procedural Content - Works, but like a computer painting a painting or composing music, its only as good as the initial input, and never as good as something handcrafted. Obvious statment I know

No - it's not "obvious".
Get a child of 3 to make your art, and it will suck. Get a trained artist to do it, and it'll be good. Write a trivial algorithm, and it will suck. Write a brilliant one, and train it with a huge amount of information, and it might well not. [it's also as good as the initial input combined with the algorithm: simple input will only work well with an exremely specialized algorithm - but who said we can't have complex, thoughtful input, or highly specialized algorithms?]

By all means say that current algorithms aren't great (for some purposes), or that developing good ones is impractical with current technology (for some purposes). Don't maintain that humans are simply better as some kind of "obvious" statement, since you have no way to back that up.
I'd probably agree if you said that humans were clearly more versatile. I don't see that changing any time soon. However, even creative computer science problems can be very specific. There's no need to create an "artist" algorithm - just one that can create good art within very specific bounds (after perhaps months/years training).

In some problem areas, this is probably already a reality - particularly those where a human artist/designer simply doesn't have time to give all areas of the creation enough attention. You could argue that the human would be better if he had the time, just as I could say that he wouldn't (for any purposes) with a good enough algorithm. If we're talking practicality though (as I'm sure you're eager to do on the anti-perfect-algorithms front), then time is a vital element.
In any endeavour where huge scale and fine detail are both important, a computer can already wipe the floor with a human, however "artistic" the problem. It can also introduce any amount of run-time variety desired - where a human is stuck at zero (at run-time). Computer entertainment is primarily about interactivity. A computer can make art interactive. A human can't.

Though I'd dispute the claim that a human can "obviously" paint/compose better than any computer [primarily since the playing field isn't level - at least try a 3-year-old vs a computer, or train a computer intensively for 20 years], this isn't the important point. Computer entertainment is not the same as art/music. The greatest thing it has in its favour is interactivity. The more a game is pre-written, pre-drawn/rendered, pre-scripted, the more it loses that advantage.
I'm not saying that's never a trade-off worth making, but it is a trade-off. The less you do procedurally, the more you set in stone, and the less potential for interactivity you get.

but perhaps the most successful adopting of this system is Spore, and even though I'm intrigued from a technical standpoint, it loses a lot for me knowing that there won't be much thought put behing the procedural worlds other than a seed value.

That is just nonsense.
I'm not sure what information Spore uses as inputs to its algorithms (have you any info? I'd be interested.), but I'd bet a whole lot it's not simply one seed value. A procedural world (or anything) can be created using as many seed values as you like, with as much thought behind them as you like. Similarly, an algorithm to choose such seed values can be arbitrarily complex, with as many inputs as you want.

Effectively, hand-crafting a world is doing this:
Enter in a huge amount of seed values, affecting very local areas.
Put them them through a trivial algorithm.
Get an output.

Procedural generation does this:
Enter in a fair number of seed values, affecting areas on various scales.
Put them through a non-trivial algorithm.
Get an output.

The difference is in the user's ability to predict the effects of a seed value change (trivial for "hand-crafting", more difficult for a procedural method), and in the user's ability to express his intentions by manipulating seed values (very hard for true "hand-crafting" [most art tools are just simple procedures where you pick seed values], much easier with a wide variety of powerful algorithms).
The most effective method depends on the scale of the problem. To create an 8x8 texture (for whatever odd purpose), you might set each pixel manually (i.e. true hand-crafting). To create a cave interior (or various models for part of one), you're not going to set each vertex individually - you're going to rely on various (procedural) tools to move groups of vertices. To produce an Oblivion scale landscape, you're going to use larger scale procedural tools. To produce a Daggerfall scale landscape, still larger.

There's nothing to stop designers hand-crafting things at the top level. The only difference is where this top level is set. You'd never want to set it at a vertex/pixel level, since computers can interpolate/extrapolate much better than a human for simple curves. I don't see why there's an automatic assumption that computers must be worse than a human at higher levels of complexity (A computer draws a horrible spline curve if you get the algorithm wrong too - higher levels of complexity just require more difficult algorithms).
The lower the level at which human design comes into things, the more interactivity and variety you lose. That's a bad thing. The level should be as high as practically possible. Perhaps that means hand-crafting most things right now: but this is an unfortunate sacrifice - not an ideal situation.

For ancillary things like forests, using Lsystems and, dare I say, soil erosion and such is perfectly fine...but once you get to places of interest, it can never be.

That's just BS, plain and simple. You might as well say that there aren't natural "places of interest" in the real world (since there was no human designer to make them interesting). The places of interest in the natural world arise out of combinations of natural systems. Thinking that combinations of "artificial" (i.e. designed, hand-crafted) systems are less capable of creating places of interest, seems rather odd.
Certainly creating good systems which produce interesting results is hard. It's absolutely not impossible though, and increases in processing power / memory are opening a lot more doors in this area (which is why it's such a shame so few are walking through them).

Note that even if all this were realized, it wouldn't leave content designers out of a job (necessarily). It'd just mean they got to create interesting countries, worlds and universes (including defining - in any amount of detail - the range of artistic styles present), rather than interesting doormats, dining-rooms and cottages.
Wouldn't that be nice?

Since its been agreed that theres only so much tolerance for an open world and that there need be a great deal of unique content to make exploration worth while...

...what is the Tolerable Minimum Frequency of unique objects to open-world? How far apart can unique content be...or...how far would you travel before 'something' needed to happen or be found?

Would knowing that anything you did find would be unique allow for even greater distances since the reward would equal the effort?

Do other elements such as combat or essential item discovery or even a survival mechanic make the interim travel interesting enough to allow for greater distances between 'set pieces'?

Are we still on "unique" = "hand-placed/hand-crafted"? It's getting a little tiresome if we are. Apart perhaps from writing/dialogue, there's no need for hand-crafting in order to achieve uniqueness (even then, a pre-written NPC/book etc. doesn't have to be hand-placed - it just needs to get placed somewhere appropriate [perhaps somewhere very rare/unique], and not be put in two/more places -unless that made sense).

What's necessary is variety in every sense. Variety within areas, variety between areas, variety between times, variety due to world events, variety due to player character changes....
Variety in NPC interactions probably needs hand-crafting, as does variety in books/scrolls/involved descriptions [though descriptions / books /NPCs can have a few Variable_Name_Here entires (naturally cross-referenced to keep all world information consistent [or intentionally incorrectly cross-referenced to allow different books/NPCs to be wrong each time]) - e.g. separate playthroughs can have subtly different histories / NPCs with mildly different backstories / outlooks...].
However, these don't need to be "set pieces" as in "previously hand placed to be in exactly spot X". So long as you have enough hand-crafted stuff, you can stick it in the player's path (only when the situation meets relevant criteria). Of course you'd eventually run out of hand-crafted stuff (at least for each context), but hopefully there'd be enough to keep the player interested for ages.

I don't think that overall world size needs to limit the density of interaction too much (so long as you don't use all your unique encounters too quickly).

If you chose a large world, why do you prefer the larger world and greater distances over the smaller, tightly packed worlds other than the 'real-world scale' rationale? In other words is there an emotional satisfaction of the journey? Does it feel more epic to travel to far off lands?

With a non-open large world (e.g. Fallout, Arcanum), I just prefer the coherence it brings to the setting. They feel like convincing worlds (in this respect at least).
With a large open world (Morrowind...[or preferably a better version of it]) I think I like the feel of adventure. I want travel to be unpredictable, enable exploration, and be a significant challenge in itself. I also prefer a setting to seem like a coherent world. Having everything packed together like a fairground doesn't give that feel. [though I'd be fine with an RPG in a very small location, so long as it doesn't attempt to pass itself off as the known world]

So long as wilderness travel/exploration/combat/survival... were handled very well, I wouldn't be too bothered by not bumping into a pre-written quest every two minutes. I would want some over-arching motivation(s) though. If my journey from A to B leads me on a long voyage through dangerous, uninhabited wilderness, there had better be some motivation to get to B (real motivation - i.e. consequences (preferably interesting ones) for not getting there). Just as I want wilderness exploration to provide immediate information (i.e. rich, informative environment) and consequence (non-trivial survival requirements), I want it to be a journey made on the basis of information (faction goals, character goals...) and consequence.

That doesn't mean a time-limit or similar that's very difficult to fulfil - there's no use penalizing exploration by making travel a sprint. So long as a long journey is challenging, it doesn't need to be a race against time. In large scale world, it'll be convincing that getting to B at all is a challenge: there's no need for a "This message must get to B at once!!1!" quest (or equivalent), but rather a "In such perilous times, a journey to B will be hard indeed. I trust you have the skills to make it with this message. It's vital that no-one learns of your purpose." (or whatever). [clearly there doesn't actually need to be a formal quest - just a convincing reason that it's helpful to be in B to further certain aims]
What I'd want is a world that responds to my heroic/epic journeys enough to make them seem worthwhile. It needn't (preferably) be any urgent/fed-ex purpose, but I do want purpose.

 

[galsiah]

Presumably you could regenerate the same terrain with the same seed.

Probably true. I mean, there would have to be significant randomness in the area generation, once other basic parameters are specified, but areas that have already been visited could have those random values fixed and would therefore be re-generated in exactly the same manner.

Exactly. There's really no need for any real randomness at all - any decision you'd make on a random basis should raise a question: "What criteria could I use to make this decision better than random?". Perhaps at some low level you'd want to introduce real randomness (though preserving any seed of course), but the lower you make this level, the more control you'd maintain at higher levels.
Very little that appears random actually is random. Most of it emerges from combinations of complex systems. Quite often you'd want the appearance of randomness, but you'd rarely want true randomness. Nearly all of the time, you'd want all the control you can possibly take. (much of the beauty of nature lies in the complex patterns behind perceived randomness - you don't want to throw that away)

Each time you use randomness, you're effectively saying "I can't think of a better way to make this decision than rolling a dice.". Usually that's not something to be happy about - on anything but the lowest levels.

 

[suibhne]

Agreed in general, but stuff like leaf litter on the forest floor and rocks will need to be fairly random. Ideally, tho, I'd love to see a system that could generate a world based on input values like rainfall, altitude, ecotype, etc.

For me, I'm not as interested in seeing a world like this generated on the fly. I'm more interested in using this as a design tool so a world doesn't have to be hand-created; a designer could then hand-tune aspects of the generated world. Either way, a lot more attention could be paid to this, but I suspect most developers aren't that interested in real "world-creation".

 

[galsiah]

Agreed in general, but stuff like leaf litter on the forest floor and rocks will need to be fairly random.

They'll need to appear fairly random, and will need some level of randomness at a low level, That doesn't mean they need to be randomly placed using a similar/identical algorithm. There absolutely should be something unique about patterns of leaves, and something separately unique about patterns of rocks. That uniqueness will come from having a unique process behind their placing. That's going to be more interesting, the more non-random factors are involved (though on a small scale it will probably appear random).
There's not a great benefit to one such algorithm (who really cares about leaves, after all?), but each sort of algorithm would be extremely cheap to construct, and could take seed data from a huge variety of existing data points (combined in a unique way). Spending a few hours to come up with a reasonable leaf placing algorithm is going to take much less effort than clicking a Leaves_Here button thousands of times. It's also going to appear consistent (whether that'll be consciously noticed, or just a small part of the world looking right).

Ideally, tho, I'd love to see a system that could generate a world based on input values like rainfall, altitude, ecotype, etc.

Sure - I'd want a huge "etc." there though. I'm not advocating only using natural phenomena as seeds. The important thing is to get an end product that you're happy with. Forming it using natural processes is one reasonable way to keep things from going crazy, but I don't think it's the best way to make things interesting.
The natural world shows that processes are capable of creating beauty, but it also shows they're capable of huge swaths of boredom. This is why (with current tools) you need to create a basic landscape first, then add most areas of interest by hand. Natural processes are rarely going to provide the density of interest desired.

I'd want a boatload of extra input values, precisely to ensure that things were kept interesting. That'd make it a bit harder to keep things sane, but that's just as true for hand placed areas of interest (which might well make even less sense climatically/situationally...). The target would be an interesting world, coherent as far as possible, rather than a realistic world - with all the boring mundanity that entails.

For me, I'm not as interested in seeing a world like this generated on the fly.

I'm not interested in the "on-the-fly" part for the sake of it, but rather due to its implications. It would allow worlds on a much grander scale (much reduced storage), more continuous variety (no pre-defined lego set), as well as allowing each playthrough to have a unique world (with all the high-level similarity required by the setting).

I'm more interested in using this as a design tool so a world doesn't have to be hand-created; a designer could then hand-tune aspects of the generated world.

Sure - that's something. But again, the fact that you're aiming to "hand-tune aspects", pretty much implies that you're planning to fail with 90% of the world. I'd rather aim to produce algorithms that do the job properly, then transfer design to a higher level.
For example, you don't design a house, but rather an architectural style; you don't design individual natural caves, but rather varieties; you don't make an individual town, but rather a style of town planning....

Sure it'd be horribly difficult, but isn't it something to aim at? Isn't it more interesting to create a prototype/style, and see a world full on variations on that theme, than to create one object?
This would also give individual artists/designers the chance to explore and appreciate the results of their work, without a "seen it all before" feeling. Each change would inspire a virtual artistic revolution in the game world. I don't see how that'd do anything but encourage creativity - the act of creation becomes an interesting game in itself.

Either way, a lot more attention could be paid to this, but I suspect most developers aren't that interested in real "world-creation".

I wouldn't say it needs to be "real", just that the more creative expression and control that's possible, the better. A world should be more than the sum of its parts, so designers ought to be aiming at expression and control over the world - not over one wall-hanging (in my oh-so-humble opinion, naturally :D).

 

[EvoG]

Shit, I thought we were past this brand of 'discusion', and with all due respect to the common ground I thought we were finding days ago, this is becoming terribly frustrating. I can't tell now if its my posts that elicit your confrontational responses of if its just your personality, but I'm not enjoying this and I read your responses with increasing anxiety simply because I want to talk but I have to wade through paragraphs of how much nonesense I spew. Do you not give me ANY credit for perhaps knowing a little bit about what I'm talking about, or at the very least have given EVERYTHING you've mentioned research and thought? You ideas frankly are not that 'unique', but their implementation simply has not been successful.

I'd hope that we were beyond the micro analysis of every minutiae of every concept and that you'd stop taking everything is say so damned literally.

And I'm still not seeing how reward for mundane-activity-X makes a game any better unless that activity is intrinsically entertaining, or is the only way to introduce a mechanic which creates entertaining gameplay.
This simply is not the case for any "bathroom break" system. It sucks in any game that doesn't either consider toilet humour an asset (fair enough in some games), or is based in a setting where a "bathroom break" is a significant event (I've never seen one).

For the last time (I promise you I wont address this any more), from a game theory perspective, its a trivial interaction that can generates positive reinforcment to an action, arguably no different than Pavlov's experiments. For someone who likens himself to a broad thinker, I'm baffled you are having difficulty with this concept. For fucks sake not everything NEEDS to make the game "better" for whatever the fuck that even means. What if ADDING trivial interactions IS what makes a game better? Obivously people enjoyed talking about flushing the toilets in the game. It not only does NOT detract from the game, it in fact enhances the experience simply because an expected reaction resulted from interaction, or else people wouldn't have discovered it by 'trying' to flush the toilet. Plus they get 10 HP increase...and I dont know, but in a shooter, any HP you can get is a good thing...but I'm sure this is bullshit too.

No - it's not "obvious".

No? Its not obivous statement to make? You're kidding right? You find me ONE piece of computer generated art that is superior to high quality hand crafted work and I will forever proclaim computers are artistically superior to professional artists...

Don't maintain that humans are simply better as some kind of "obvious" statement, since you have no way to back that up.

Um, its called ANYTHING made by artists since the dawn of man versus the BEST examples of algorythmic creation...but I'm sure this is probably nonsense

Though I'd dispute the claim that a human can "obviously" paint/compose better than any computer [primarily since the playing field isn't level - at least try a 3-year-old vs a computer, or train a computer intensively for 20 years], this isn't the important point. Computer entertainment is not the same as art/music. The greatest thing it has in its favour is interactivity.

but perhaps the most successful adopting of this system is Spore, and even though I'm intrigued from a technical standpoint, it loses a lot for me knowing that there won't be much thought put behing the procedural worlds other than a seed value.

That is just nonsense.
I'm not sure what information Spore uses as inputs to its algorithms (have you any info? I'd be interested.), but I'd bet a whole lot it's not simply one seed value. A procedural world (or anything) can be created using as many seed values as you like, with as much thought behind them as you like. Similarly, an algorithm to choose such seed values can be arbitrarily complex, with as many inputs as you want.

This is where you are taking me far too literally, perhaps in an easy attempt to refute my statments but I can't see why.

I didnt' say ANY human...I was referring to HUMANS that are also ARTISTS that are also VERY TALENTED...

I didn't imply that Spore uses a SINGLE SEED for its creation, thats rather silly, no? But whats perhaps more silly is that you assumed thats EXACTLY what I meant. Please dont tell me I have to explicitly spell everything out...but its more likely I'm just failing basic communication skills.

That's just BS, plain and simple. You might as well say that there aren't natural "places of interest" in the real world (since there was no human designer to make them interesting). The places of interest in the natural world arise out of combinations of natural systems. Thinking that combinations of "artificial" (i.e. designed, hand-crafted) systems are less capable of creating places of interest, seems rather odd.

I dont even know how to start this repsonse...dear god I thought the term 'unique content' has been established by now over six pages.

"Places you visit in an RPG that serve as a set piece to the narrative." or some shit like that. You have a forest...then you have a town...and a dungeon...and a ruin...and a caravan...and ANYTHING else other than just terrain...but I'm certain that I'm full of shit.

Are we still on "unique" = "hand-placed/hand-crafted"? It's getting a little tiresome if we are.

...and your lengthy retorts on how just utterly clueless I am isn't tiresome? Yes the thread is indirectly ABOUT unique content with regards to the creation of a game world. I started the thread, I know what I was asking and this hasn't deviated at all. Narrative games since the begining of narrative gaming have and still have "hand-placed/hand-crafted" content, and it always will. Honestly if you want to have a discussion of procedural worlds then, by all means start a thread for that, because you clearly will be disappointed the next time I mention "unique content"...and if I do I'm pretty sure I'll be called on it.


I know this repsonse is rather heated, and I apologize and dont want to 'go there', but I dont see any other way to impress upon you that this is getting nowhere and again, instead of say something is NONSENSE, please offer evidence, or information, or ideas rather than just saying several different ways why I just MUST be completely wrong galsiah.

 

[Dire Roach]

And adding lots of details NOT like these can make it memorable for the right reasons (in an RPG). Better to spend time on details which matter, and on mechanics which aren't needless nonsense.

Unless a game developer is working under extremely tight time constraints (which is a sure recipe for disaster anyway), I don't see why one should purposefully avoid adding extra details that complement basic gameplay. The way I see it, these unnecessary details are what makes a game much more memorable and endearing. They make the "fishermen's tales" Section8 mentioned earlier all the more interesting to tell and listen to.

...what is the Tolerable Minimum Frequency of unique objects to open-world? How far apart can unique content be...or...how far would you travel before 'something' needed to happen or be found?

Personally, I expect to find at least one thing of interest in a given "landmark area". If I arrive at a forest, a small lake, a valley, a peninsula, etc., I will expect that there is at least one reason why this place was added to the game, and not just for sightseeing. The player should receive an appropriate reward for carefully exploring such an area.

Would knowing that anything you did find would be unique allow for even greater distances since the reward would equal the effort?

As long as the player perceives that the rewards are indeed equal to the effort, then yes. Distance shoudn't be the main factor when determining challenge, though.

Do other elements such as combat or essential item discovery or even a survival mechanic make the interim travel interesting enough to allow for greater distances between 'set pieces'?

Combat and item discovery, no; Oblivion is a good example of this. A fun survival mechanic, probably. Either way, sometimes I just want to walk around and absorb the scenery at a slow, uninterrupted pace while I'm exploring.

...why do you prefer the larger world and greater distances over the smaller, tightly packed worlds other than the 'real-world scale' rationale?

A larger-scaled world definitely gives the illusion of a greater sense freedom of and accomplishment when you travel from one area to another.

 

[galsiah]

Shit, I thought we were past this brand of 'discussion'

Ok - I'll try to be more reasonable.

if its my posts that elicit your confrontational responses of if its just your personality

I'd say four things:
(1) I'm hopelessly arrogant (I think we covered this).
(2) I care about these issues pretty passionately.
(3) I strongly disagree with some of your standpoints (particularly in light of your aim of creating "open worlds").
(4) I think you're a fairly reasonable, intelligent, informed person, who therefore ought to see the light of truth. :D
(Oh and I'm hopelessly arrogant.)

If I thought you were dumb I wouldn't be continuing the discussion. What I find frustrating is the fact that you're clever, yet don't seem to see my points (AKA The Truth In All It's Glory).

Certainly my presentation is poor. I'll aim to do better.

Do you not give me ANY credit for perhaps knowing a little bit about what I'm talking about...

Certainly - that only makes things worse (see item (4)).

or at the very least have given EVERYTHING you've mentioned research and thought?

Pretty much. Bear in mind I'm mostly taking a "X is theoretically possible" standpoint - which doesn't need much research. I'm against dismissing procedural ideas out-of-hand - not against saying they're very hard to do well, or impractical in some particular context.

You ideas frankly are not that 'unique',

Absolutely. They're not new or unique - hence my frustration at the lack of widespread support.

...but their implementation simply has not been successful.

Sure - that's evidence that they're hard to get right. Also, how often have they been tried with real backing in game development? What kind of resources were available for such attempts?

Again - I'm not saying these things are easy or right for every project. I am saying they're worth a thought if you're aiming to construct a large world.

I'd hope that we were beyond the micro analysis of every minutiae of every concept and that you'd stop taking everything is say so damned literally.

I'll try to do better on this.

What if ADDING trivial interactions IS what makes a game better?

That's possible - fine.

Plus they get 10 HP increase...and I dont know, but in a shooter, any HP you can get is a good thing...

This is not good (IMO). The enjoyment of the trivial interactions comes from the ability/freedom to do it, and the coherence that adds to the world. This is true of using taps/flushing toilets in e.g. Deus Ex.
However, incentivizing the use of such trivial items doesn't make them more entertaining - it merely pushes the player into jumping through uninteresting hoops for the sake of a game mechanic. Had using taps / flushing toilets in Deus Ex given bonus health, that would have made most people's games a little more dull, as they're pushed towards engaging in mundane activity. [[EDIT: actually, thinking about it, drinking fountains did give 1 health per click until they ran out - and yea, verily, it was dull (anecdotal, but still true)]]
Where it does make more sense (e.g. gaining health from food), it's a little more satisfying, but even then I'm not sure it improves the game. Was Deus Ex a better game by virtue of Soda cans and chocolate bars that gave 1 point health increases? I'm not sure it was.

Either way, if you're incentivizing fairly mundane behaviour at all (which isn't necessarily terrible), it makes sense to do it in such a way that you can control the frequency/distribution of such items independently of level aethetics/layout. Incentivizing urinal use means you can't place a urinal without thinking "Am I harming gameplay by pushing the player towards a near constant urinal usage?".

If you want a funny little bonus, it makes sense to do it in such a way that it's coherent, and you have the freedom to control frequency without messing with other level-design criteria. I don't think urinals are a good candidate on either score. Will they ruin a game horribly? No. I just don't think they're a good use of effort.

and I dont know, but in a shooter, any HP you can get is a good thing

That's not a great argument. [makes effort not to take too literally] [check failed] A similar argument could apply to getting health bonuses for anything - every ten seconds, every time you look up, pressing a mouse button four times.... An advantage for the player character does not necessarily constitute a good thing for the player's overall entertainment - even if he's happy at the moment he gets it.
I'm sure you realize this anyway.

No? Its not obivous statement to make? You're kidding right? You find me ONE piece of computer generated art that is superior to high quality hand crafted work

If you'd read what I said carefully (and perhaps my tone didn't help there), you'd realize that this is not what I suggested.
First, I'm not saying that "computer generated art is superior". I'm suggesting that it's not "obvious" to think that it will always be inferior in every context.
Second, I'm saying that you're making comparisons between highly trained, experienced artists, and fairly simple, inexperienced programs. I'm not saying you need to compare any human - I'm saying that if you don't, you ought not to compare just any computer art program: rather only well-supported, purpose-built, highly trained ones.

and I will forever proclaim computers are artistically superior to professional artists...

Did I make such a statement?
We're not in black-and-white territory here. You made an absolute statement. I'm just saying it's not "obvious" that the situation is absolute. You're dismissing the potential of computer algorithms to create "art" as well as human artists without regard to context or improvement.
That's not very fair.

Um, its called ANYTHING made by artists since the dawn of man versus the BEST examples of algorithmic creation...

First, see my above comments on the absence of a level playing-field.
Second, it's just not true to say that a computer hasn't done better than artists at some artistic tasks (e.g. creating huge landscapes in a reasonable time). You just probably wouldn't call them "art" (but the boundaries are far from clear - particularly in game related terms where anything displayed on the screen tends to get called "artwork").
Not all computer "artwork" needs great artistic virtue - it just needs to look right.

Also, I'd say some computer generated abstract art can be pretty interesting/"good". Perhaps it's not to your liking, and perhaps it isn't that relevant (I'd probably agree), but it's not bad for what it is.

I didnt' say ANY human...I was referring to HUMANS that are also ARTISTS that are also VERY TALENTED...

Sure - but then don't include a load of terrible attempts at computer generation in the comparison. Compare the best with the best in a given context. I freely admit that computers aren't great at this yet (apart from in a few very specific areas), but the point isn't whether computers are better artists than human artists.
It's whether suitably designed, highly specialized programs can create some types of material to display on a computer screen, faster/with more variety/better than human artists in some contexts. You can probably argue against that (though I'd disagree), but it's a bit much to think that it's "obvious" - i.e. that the problem doesn't even deserve consideration.

I didn't imply that Spore uses a SINGLE SEED for its creation, thats rather silly, no? But whats perhaps more silly is that you assumed thats EXACTLY what I meant.

That was in the context of your assumption that "there won't be much thought put behind the procedural worlds". What basis do you have for that statement? (again, if you do have info, I'd be interested).
It's entirely incorrect to go from something's being procedurally generated from seed values, to thinking that there "won't be much thought...". If you had meant it was one seed value, that'd be fair enough (hence my assumption). As it is, there are presumably loads. Why can't there be huge amounts of thought behind each choice? (again, perhaps they're just random, but I'd be disappointed if that were the case)

Please dont tell me I have to explicitly spell everything out...

You don't, but I'll tend to make assumptions based on what would have your argument make sense to me. To my mind it doesn't make sense to think there wouldn't be much thought involved, simply due to procedural generation. Therefore I'd either presume that you thought they'd be randomly selected seed values (let me know they are, and Ill shut up on this one), or that there was only a single / trivially few seed values.
Your statement seemed to imply the latter, so I assumed that.

If there are loads of carefully chosen seed values, I don't think it's fair to say that the worlds aren't thoughtfully/carefully constructed. They'd just be thoughtfully/carefully indirectly constructed.

I dont even know how to start this repsonse...dear god I thought the term 'unique content' has been established by now over six pages.

Did you say "unique content"? Perhaps you did....[checks]... right - you did, but you also said "unique objects". I didn't take it from that that you meant "unique main quest items", or similar. My mistake.

Anyway - all my "It doesn't have to be tied to a specific position - just a specific set of circumstances..." stuff still applies. You can have a huge world, but make sure that relevant events happen without too much delay, just by moving them into the player's path (so long as this is consistent with everything).

"Places you visit in an RPG that serve as a set piece to the narrative." or some shit like that. You have a forest...then you have a town...and a dungeon...and a ruin...and a caravan...and ANYTHING else other than just terrain...

I'm not sure quite what you mean here (was it something I said :D). However, I'm certainly not saying that you can't generate towns/forests/dungeons/ruins..., then populate them with unique NPCs/objects as necessary. Perhaps you're not interested in doing that sort of thing, but it's certainly possible. (Clearly just generating terrain would be a little dull)

The main point (even if you're totally against general procedural terrain etc.) is that you can move these things around to make sure the player bumps into them. E.g. a simple NPC X in a hut in the woods. You don't need to worry about the player spending an age without bumping into that guy, since you can move him.
Probably this makes most sense in a very large world (so most in a huge procedural one), but it applies anywhere.

...and your lengthy retorts on how just utterly clueless I am isn't tiresome?

When did I say you were utterly clueless? The last time I wrote that something I, myself had written was "nonsense" was yesterday.
If I say that something is nonsense, it's because I think that [my interpretation of what was meant] is nonsense. That implies nothing about the writer of such a remark. If I follow that up with a lengthy retort, that's to explain what I mean. If my meaning misses the point, then my remark is addressed at a MISinterpretation of something you said.
By all means set me straight, but there's no need to think I'm calling you "utterly clueless", when I'm merely misinterpreting your remark.

Narrative games since the begining of narrative gaming have and still have "hand-placed/hand-crafted" content, and it always will.

Hand-crafted yes. Hand-placed is much less necessary. You can't easily (or at all) write a program to write you a convincing story; you absolutely can write a trivial program to shuffle a few NPCs around (hand-placing them is certainly not necessary). In a overview-map-based game (e.g. Fallout), you could do the same with towns (within certain limits); in a Morrowind style world, that'd be harder (so probably impractical for your purposes - fine).
Even in a Morrowind-style game, you could move individual huts / cave entrances etc. to different locations without much trouble.

You don't need anything close to a "procedural world" for this sort of thing. You only need a few fairly trivial algorithms to mix things up and make replays a little fresher. If you hold off on placing stuff for a while, it can also help with your "interactivity density" - since you can exert more control over pacing.

Procedural design is the most relevant in large, open worlds - so I mention it in this context. If you don't want to go off on a procedural bender, that's fine. I do think it can help both with pacing (i.e. density), and with freshness between replays though - all fairly trivially.

I know this repsonse is rather heated, and I apologize and don't want to 'go there'

That's fine. Again - I'm not wholly unaware of my own (occasional) arrogance, and you're being pretty reasonable. I just happen to disagree quite strongly in some respects.

instead of say something is NONSENSE, please offer evidence, or information, or ideas

I usually do, to be fair (my "lengthy retorts"). I should probably just lose the initial "NONSENSE!!!!!", since it doesn't set the most constructive tone.
Usually I stick to ideas (rather than tangible evidence), since I'm rather more interested in thinking about what might be possible, than going over what was done badly a few years ago.

rather than just saying several different ways why I just MUST be completely wrong galsiah.

You do disagree with me in some areas though. If that doesn't make you wrong, it's hard to see what would :D.

 

[EvoG]

Nice reply galsiah :D

I'll hit this in just a moment...

...I'm watching Modern Marvels on Hist Chann about Stealth Aircraft and I'm tired of sitting here as it is (for work)...

...but I see where we crossed our wires, so I think I can help here with a productive response. :D

 

[suibhne]

...what is the Tolerable Minimum Frequency of unique objects to open-world? How far apart can unique content be...or...how far would you travel before 'something' needed to happen or be found?

Personally, I expect to find at least one thing of interest in a given "landmark area". If I arrive at a forest, a small lake, a valley, a peninsula, etc., I will expect that there is at least one reason why this place was added to the game, and not just for sightseeing. The player should receive an appropriate reward for carefully exploring such an area.

I disagree. I hate arriving at something (like "a forest, a small lake", etc. that you mention above) and being able to safely assume that the developer wouldn't have created it without stashing something there for me - a dungeon, loot, w/e. This approach ultimately makes me feel as if the entire world was designed around me, and I want the opposite feeling - that I'm stepping into a world that exists independently. I'd like nothing better than to be able to explore a world and experience interesting, even extraordinary landscapes without feeling like the devs are tossing me little rewards for exploring every area.

(As we discussed above, of course, the flipside is that exploration itself should be interesting in some way. But using loot, dungeons, etc. as the reward everywhere is not a good long-term approach, even tho it's obviously appropriate to place unique loot, dungeons, and characters in some areas.)

 

[Dire Roach]

I guess the trick is to give players the illusion that the world was not designed around them... otherwise, how can exploration be interesting for its own sake?

 

[Damned Registrations]

I disagree. I hate arriving at something (like "a forest, a small lake", etc. that you mention above) and being able to safely assume that the developer wouldn't have created it without stashing something there for me - a dungeon, loot, w/e. This approach ultimately makes me feel as if the entire world was designed around me, and I want the opposite feeling - that I'm stepping into a world that exists independently. I'd like nothing better than to be able to explore a world and experience interesting, even extraordinary landscapes without feeling like the devs are tossing me little rewards for exploring every area.

(As we discussed above, of course, the flipside is that exploration itself should be interesting in some way. But using loot, dungeons, etc. as the reward everywhere is not a good long-term approach, even tho it's obviously appropriate to place unique loot, dungeons, and characters in some areas.)

I second this, with this added though;

If every damned lake, rock, pillar, wall and hill in the game has treasure in it, then I'd feel like I had to scour every damned one of them till I found it's goodie. Otherwise I'd probably end up missing out on the one out of 50 of them whose reward was actually special. Few things piss me off in a game more than knowing I missed some rare loot. Half the reason I play RPGs is to attempt to create the most imbalanced, overpowered monster possible within that framework. Knowing I missed out (Or might be missing, until I search all 3000 lakes) on the only pair of gauntlets +7 in the game and being stuck with the gaunlets +6 would annoy me to no end. If I have no idea where they might be I won't mind as much; I didn't fret over trying to find all the bits of daedric armor in MW because I knew they were scattered SO well that actively searching for them was futile, they were essentially random rewards. If I knew that they were all in dungeons just sitting in a chest, I'd have felt like I needed to go to every damned dungeon in the game, which would have been irritating.

Having treasure stashed in the hollow log in the middle of nowhere is only cool if thats a special occurence. If theres treasure in every hollow log it just becomes a hunt for hollow logs, which isn't very interesting.

If something is obviously, truly unique (The only giant white marble shrine in the wilderness) THEN it should have some sort of cool (And unique) reward, even if it's just some minor blessing like the wayshrines in Morrowind.

 

[Dire Roach]

If something is obviously, truly unique (The only giant white marble shrine in the wilderness) THEN it should have some sort of cool (And unique) reward, even if it's just some minor blessing like the wayshrines in Morrowind.

This is what I'm referring to when I was talking about minimum tolerable frequency, not having every single rock and log be a container for random uber loot. When I reach a place called "The Valley of Giants," I'm hoping that the place has something of interest for me other than it just being a pretty place where I can imagine that giants once roamed. Maybe there's nothing of interest but a rare flower that grows exclusively in that valley, which might have some possible use to me, and that would be enough. What I don't like is having a swamp in the middle of nowhere populated by generic monsters. I would explore it once, just to see if I would find something of interest, but then I'd feel like I wasted my time going there.

The only exception I can think of to this is if the player is told beforehand that a certain location is meant primarily for sightseeing. In that case, it's more about the game's artists showing off their mad grafix skillz than anything, unless it's somehow a part of a puzzle like the Mural of Enlightenment in Arcanum.

 

[Damned Registrations]

But if you don't have that swamp and places like it, there won't be any middle of nowhere. It'll be like playing Diablo 2, where every 400 feet is another tomb, another cave, another whatever with something 'unique' in it. The valley of giants won't feel special if every single valley is 'The Valley of Blank'. You need a nameless swamp filled with pointless fluff to fill up the space between those places. If you don't like fighting trolls, run past it. As long as you aren't forced to stop and explore every boring inch of the map, having boring areas isn't a problem.

I just remembered something else about world exploration I generally dislike. Having the entire world either be at the same level of difficulty or become progressively more difficult the further out you go. I remember when playing Everquest one of the most interesting things to do in the game was simply try to travel from one major city to another without being killed. The areas right next to the cities were always newbie zones filled with weaklings for the new players to farm, but a zone or two past that you could run into pretty much anything. I rather liked having to sneak around through the mountain pass filled with deadly gnolls to get to another city. Both because the high powered gnolls and people around fighting them were new and cool looking to me, and because when I got to the other side (No small feat if done alone.) I could fight with things of my own level again; only they'd be entirely different enemies than what I was used to, and all the other players there would be of a different race than me. There was also a forest, which was relatively safe during the day, but at night was populated by undead horrors that would scare off even high level (level 40 out of 60 levels) from travelling through alone. That makes for a hell of a lot more interesting world than one where you find the difficult enemies and then stop exploring because the only thing beyond them is a set of even stronger enemies, ad nauseum.

 

[suibhne]

But if you don't have that swamp and places like it, there won't be any middle of nowhere.

That's my basic concern with the prioritization of "open exploration". When you turn world exploration into a "real-time" activity (and yeah, I know time is compressed in G3, Oblivion, etc., but it's still real-time in game terms; there's no time differential between world exploration, haggling in town, fighting in the arena, etc.), you're forced to make your world unrealistic. Players can't go for too long without finding some distraction, something to do, whereas any sort of real-life travel would take hours if not days. In real-life travel, most of the areas between your origin and your goal will be "middle of nowhere".

 

[Mr Happy]

Maybe I'll grace this thread with a full read-thru tommorow, but right now I'll just throw in my prefences on world design.

I do like exploring, looking for nooks and crannies etc, and this would suggest a small hand crafted world. This, of course, comes into conflict with my dislike of scaled down worlds (I'm also not a fan of sped up day/night cycles) in most cases. I dont really get much out of walking from one town to the next real time. Small worlds are fine, but I'd prefer they be in a (moderately) realistic scale.

For massive worlds, a good travel system, at least for me, is usually a good substitute for the ability to explore (this is the main reason I'm trying to get a hold of RoA ). It certainly should just be clicking on a location and arriving there, and it shouldnt feel easy. Strangely, I am a fan of the Daggerfall way: the land is there, but it is to big to travel between locations in most cases in a reasonable amount of time. I like the openness and the fact that its not just location based, but I also like the details of a good travel system. Things like encampment options, deciding in what way to travel, options to stop at inns etc. would not be all that fun if done in realtime, outside of an abstracted system, tho I like just being able to "walk into the wastes".

The great thing about a location based world is that places can be "unlocked'. This seems great for roleplaying, because it allows the game to let you figure the location out through information or stumble upon the place based on the skill of youre character. This is also great for replay value: the second time through you cant just say "oh, I remember Necropolis is over here, time to get that waterchip!!!" In this case, you arent going to find a dungeon just by walking youre charater into the woods and noticing it tucked in a crevice. Of course that's fun as well, so for me, there really isnt one perfectly designed rpg world. Both systems suit different styles of RPGs. If I had to choose one system, it would have the daggerfall massive open world with a detailed travel system that was a challenge in itself, had plenty of random encouters/environmental hazards, and took character skill into account

About getting lost, well, usually. Being lost out of your mind in a daggerfall dungeon is all good with a recall spell. Looking for an obscure dungeon in Morrowind's cliffs in a dust storm is a bit of a pain in the ass. However, I'd like to see getting lost modeled in a travel system. Poor travel skill, faulty compass, whatever, and using clues to find your bearings, maybe noticing a city in the distance or something ( I thought Fallout modeled seeing a city out in the distance decently, the problem being that it was a little "easy" and Vault 15 was just as easy to spot as LA).

Disconnected thought: As a way of getting around the...non handcraftedness of a massive procedurally generated world, hand crafted "pockets" around cities, important locations, or certain regions/islands would be interesting. Time consuming to, probably, but even so.

 

[Amateur]

Do you like large worlds, where locations are farther apart and there were a lot of natural filler terrain?

Yes, a lot. It makes the world much believable. But of course there should be some hidden unique rewards to make the travel meaningful. And of course a "smart" implementation of fast travel is necessary(not the stupid implementation of travel in OB). Actually I liked the travel in Fallout series a lot. When completing a quest just simulate the travel like fallout and alert when an event occurs. As improvements for example an implementation of drawing the path on a map might be a nice improvement...



Do you like small worlds where its quick to get from location to location, with little to no filler, where every area was unique?

I like unique areas but definitely don't like small worlds. With an intelligent implementation you can make travel faster also in massive worlds.


Do you like to get lost?

Not really. I don't like to get lost while completing a simple quest. But when exploring the environment it is actually nice.


Do you like to find hidden areas based on distance or based on simply being hidden by occlusion of large objects or general object density?

Yes, very much.

 

[Klinn]

I love exploring, both in games and real life. It’s a series of mini-mysteries -- what's over that hill? what’s around the bend? what’s that strange half-hidden structure silhouetted on the hilltop? This is distinct from what I guess you could call target-based “pathfinding”, where you are trying to find a particular goal and are using tools/skills/deduction to reach it. That’s also “exploring” but not quite the same activity, and the player’s expectations and potential frustrations are different.

This discussion of game world size is interesting since games can now have such long view distances. In games like Morrowind, there was always the “distance fog” that you could count on to make a world seem bigger than it was. Now in Oblivion and other games, that trick is stripped away. As somebody alluded to earlier in this thread, I remember exploring through the Ob’s Great Forest for what was supposed to be a long period of game time (about 1-1/2 days) then climbing a hill, turning around and seeing the Imperial City tower clearly visible just over there, maybe a kilometer away at the most. The world suddenly felt very small.

Gothic3 had an interesting approach with its pseudo depth-of-field effect. It allowed some of the long view distance, but made it fuzzy and indistinct, kind of heading back to the traditional ‘distance fog’ but not as arbitrary. It felt a bit like a fishbowl at first but I think that’s just because of the specific implementation in G3, the method itself has promise.

Anyway, now that it’s possible, I don’t want to give up long view distances. It’s so enjoyable from an exploring point of view to come across these grand vistas. But to compensate, the game world has to become bigger so places are actually further apart. This means the time compression should be lowered so a day’s hike still takes you a ‘realistic’ distance. The problem with this approach is that we are getting closer and closer to a 1:1 hiking simulator. We don’t want to spend an hour of actual time in the game to get over the next hill.

How to fix this? There’s the traditional ‘instant transport’ method, e.g. via a map or asking an NPC to take you somewhere on her boat/slitstrider/portal/etc. Usually these are only to specific locations. I would love to see a situation where you could point out any reasonable location on your map and ask the NPC to take you there. By reasonable, I mean something like “drop me off halfway along the Rambling Road on your way to Thrillville” not “take me to this mountaintop in the middle of the frozen wasteland”. In other words, something that suits the NPC’s usual behaviour.

The instant transport method has the disadvantage of taking the player out of the game environment for a moment, and it's not really 'exploring'. Another way to avoid travel drudgery is to somehow increase his speed. Use horses/tame guars/machines/magic boots (preferably not blinding) so the player can zip through long distances within a reasonable length of time, but not so fast that they can’t notice something cool and slow down to change direction and examine it. Make it the inverse of the adjustment to the game’s time scaling, e.g. have time pass only half as fast as before but give the player a way to move twice as quickly. Yup, this means developers will have to stream world content more quickly and consistently without annoying pauses.

Another way to cut out the travel time is using smaller episodic zones and just rip out the distance in between, like Fable. Personally I felt that was too artificial -- ah, I’m in the lake zone, now I’m in the farm zone, where’s the next gateway?

Any other ways of taking the drudgery out of long distance travel?

Heh, back to your initial question… As you can tell, I prefer larger game worlds. Basically for two reasons, both of which are tied to our expectations that the world should work in a manner somewhat similar to our own. First, it strains credibility when the giant glacier is right next door to the bubbling swamp. An exaggeration, but there needs to be a buffer or transition between different areas. OK, it’s a magic world, anything is possible, but what about reasonable NPC behaviour? As in “You want me to try and find the bandit raiders? What, has nobody from the village ever wandered over that hill? They’re right there sitting around a campfire. You can even see the glow at night, yah dumb stumps!” Need buffer zones again to make the villagers’ ignorance seem plausible.

I’m OK with some ‘filler’ terrain, the quiet lulls between places of interest, but if the size of it increases as I think it will, there needs to be a method to travel more quickly than just running.

 

[suibhne]

@Evo: I thought about this topic again while walking about four miles to work today, from Chicago to Evanston. I'd fired up Oblivion last night to test out some graphics settings, loading up an old savegame from Sancre Tor and looking all the way down to the Imperial City, and I couldn't help but compare that to my morning walk: I can see my workplace from miles off, since it's on the lakefront, and that measly 4-mile walk looked no closer than the Imperial City to Sancre Tor.

...Which reminded me of another memory, a solo backpacking trip I did in the Rockies a few years back. My last morning started on the Continental Divide and descended all the way to the foot of the valley, where I was parked. I could see most of the way down the valley from my campsite that morning, and the distance was impressive - but, as it turns out, I hiked it in about 3 hours. It was only about 9 miles, a single morning's hike, but it represented far more scale (subjectively and objectively) than almost anything available in Oblivion. It was unquestionably vaster than the distance from any point on the Empire's border to the Imperial City at its center.

So, again, I'm confirmed in my extreme skepticism about any game being able to handle "open exploration" without A) a lot of dead time, or B) much abuse of realism. Sure, the gameworld can be compressed as in G3 or Oblivion, and time can be compressed, and monster population density can go through the roof; this is a perfectly acceptable approach to the challenge (or can be acceptable if handled well). But it's intensely unrealistic. If the game is immersive enough - i.e., if its internal logic is consistent enough - then I'm not especially bothered...but this design approach still misses one of the things I'd love to achieve in a game, viz. a sense of true scale. Maybe it's just the necessary tradeoff, but, even after this long thread, I can't imagine an "open exploration" game coming close in that department.

 

[elander_]

No? Its not obivous statement to make? You're kidding right? You find me ONE piece of computer generated art that is superior to high quality hand crafted work and I will forever proclaim computers are artistically superior to professional artists...

I have a similar issue with 2d in isometric or platform games art versus 3d art. It's obvious that 2d hand drawn can be much better than 3d environments. It's like a painting where the artists can control every aspect of color, brightness, texture, contrast and so on. Yet everyone uses 3d art and if check some very well drawn and older 2d games you will notice something was lost when they switched to 3d. Eventualy they will switch to a mix of procedural content and hand-made 3d art and that is as much unavoidable as the 2d to 3d switch.

The advantages are many. You don't have to pay a lousy or average artists what an algorithm can do for free and much faster, and the algorithms for procedural content generation are getting better.

Then rpgs worlds are geting bigger. People (myself and other Daggerfall players included :p ) are complaining that worlds don't have a more realistic scale. Daggerfall cities suffered from a lot of repetition but they look like real cities with more than 500 houses each. Most of these houses could not be entered and were there just for the scenario but the feeling was great.

Now imagine this done with a modern algorithm that can create cities like Speedtree creates forests and at the same time allow for hand-placement of special content. Once someone does this everyone will want to do the same thing.

Theres, of course, a problem when trying to copy the exploration model from games that use a simplified world. Games like Daggerfall solve this problem by having the player serach for quest maps that will point the location of treasure sites on the players map. Fallout used a system where the player could explore the world using a campaign map. This was a sort of campaign map similar to those we see in older rpgs, but not as complete as a in strategy games like Civilization.

There's also a problem with open-endness when the game doesn't have barriers like the Gothics. In a huge world the player would never find a secret location just around the corner of every road and every little mountain like in Morrowind. This would require new game mechanics for a new type of world. With an advanced LOD system (similar to the one used in Oblivion, fair is fair and their graphics engine deserves the credit) the player could spot, for example, that imponent wizrad tower of the mighty xerces miles away, and so with proper directions exploration would only have to be done in the imideate vicinity of the tower.

 

[Veracity]

It's obvious that 2d hand drawn can be much better than 3d environments. It's like a painting where the artists can control every aspect of color, brightness, texture, contrast and so on.

Indeed, because unlike nasty old 3d, sprite-based games on 2d backgrounds (more typically tile-based than hand-drawn, but sometimes the latter) gave art teams complete control over each and every end user's monitor. I actually agree that the rush to 3d-ify anything and everything around the era of 3dfx dominance and the rise of the original Sony console resulted in a slew of visual atrocities from which we're still recovering, but proponents of the virtues of 2d don't half talk some bollocks.

EvoG: Darwinia looked nice. It's obviously easy to argue that the appearance of it involved human artistry somewhere in the process, but its landscapes were generated mostly procedurally. Would that be a meaningful example in the current context? Introversion's also been banging on at some length about procedural generation of cities for its next project - not much in the way of detail released, but possibly of interest. And Spore's an obvious current high-profile potential drinking game (down a shot every time Will Wright says the 'P' word). I'll stay out of the main topic, since I suspect I don't like exploration much outside more 4X-oriented games.

 

[crufty]

Take:

1 part shadows of the Colossus overland travel/combat
1 part random city generation ala daggerfall
1 part random dungeon/item generation ala nethack
1 part item generateor ala nethack/diablo
1 part random quest generator ala daggerfall/gearhead
1 part detail ala Ultima VII

get: best rpg of all time
also: commercially unviable *cries*


in hand-crafted vs random. If developers/designers can create a great hand-crafted scenario, then they should be able to create a great set of samples for a random scenario generator that not only pops out fantastic scenario W, but scenario W+1,W+2 as well as X Y and Z

 

[crufty]

Any other ways of taking the drudgery out of long distance travel?

I think you hit the nail on the head: increase player speed. For example, you might have walk, run, run fast, run fast *time compression*. So lets say you are going from point a to point B. You decided to hoof it. So you start your character running. After a few seconds, the game engine engages in time compression, with suitable blur and camera zoom/focal point change, so that one second of running is no longer one second in game time, but say 1 minute.

Now the countryside is flying by at a blistering rate. The secret would be to have the game engine recognize that you are on a road, and more or less restrict you to that road....

 

[LCJr.]

I confess to not reading the whole six pages so forgive me if this has already been covered.

I like world exploration in the sense you're talking about if I find it interesting. But frankly in many games it's boring to fill in your minimap looking for that hidden item the developer may have stashed in some far off corner.

The kind of exploration I much prefer is what you had in the first Geneforge. You're presented with a completely alien world and your exploration is more about finding what happened and what's going on then trying to map the world.

 

[AlanC9]

Exactly. Zone mowing like in BG1 just annoys me. I feel obligated to un-black every map screen there is so I can extract all the content from the game. But what I'm doing when I'm doing that is not role-playing in any meaningful sense.

OTOH, exploration which is in character is very welcome indeed.

Though I wonder if I'd feel the same way about procedurally generated content. Exploring a BG1 map felt fake because I knew that the designers wouldn't just put an empty map in the game. If there were no designers for some areas, I think I wouldn't feel the same way about it.

 

[elander_]

Though I wonder if I'd feel the same way about procedurally generated content. Exploring a BG1 map felt fake because I knew that the designers wouldn't just put an empty map in the game. If there were no designers for some areas, I think I wouldn't feel the same way about it.

That's one of the issues i have with BG1 and clones like NWN. It's a good game alright but the exploration part is not so satisfying. I feel that i have to unlock every door and visit every hidden part of the map because i will miss collecting some xp i will need later to finish some quests.

Daggerfall was very good with exploration even if it was more limited in terms of character interaction. The player has to join guilds and ask for jobs to the commoners to receive the most interesting quests that will link content dynamically to the world. There are also special occasions and events that will link new content to the world.

Without taking quests the Daggerfall world is just an endless repetition of the some blocks and the same generic npcs, except for some special places (that are impossible to find without directions because of the huge size of the world) and certain npcs the player will stumble upon in the courts of the 3 major kingdoms, which the player is advised to investigate.

With the rogue features this game has, exploration is even better and it's a great sense of freedom not to have to worry if we missed any spot in the map since we can always roleplay our mistakes and get new opportunities.