The pros, cons, and techniques of procedural asset creation (renamed)

[Laa-Yosh]

Not really... What this means is that Epic and Crytek can't do much about including procedural tools because every single client of them will have vastly different needs and requirements for such tools, depending on the type and specific attributes of the content they want to create. So it would be a waste of their resources and development time.

For example, I don't think that Mass Effect, Bulletstorm, or Borderlands could use the same face generator tool because of the vast differences in art style and what the characters are required to do in the game. So what kind of a tool should Epic then try to include and who would end up using it? And that's still a relatively small and self-contained piece of tech...

But yes, in general, procedural content does mean that you want to shift the workload, that is true - however it's not an answer to this particular question.

 

[Prophecy2k]

Just because a vocal segment of the customers criticized ME2's approach it doesn't automatically mean that they are the a majority and they are also right in their criticism.

The only objective facts are that ME2 got better reviews and sold more copies.
I'm sure Bioware also has statistics on how many of the random missions in ME1 were completed and how many of the more hand-made ones in ME2, so they can draw more detailed conclusions and we'll see if procedural missions are back in ME3 or not.

Oh, and throwing in planet scanning as an argument is a low point, as it's completely unrelated...

Well of course ME2 sold more copies when it was on three platforms instead of two, plus the fact that superior sequels generally tend to sell better as their prequels have already built a nominal fanbase. That still takes nothing away from the fact that removing the explorational elements from ME1 to ME2 was one of many reccuring criticism of ME2 across the board, dispite the game scoring higher. Planet scanning also isn't unrelated because it was a direct replacement to the planet exploration elements in the first game. In terms of gameplay mechanic for the discovery of side-missions, they both served exactly the same purpose, so of course they are directly comparable.

 

[Laa-Yosh]

ME2 sold more copies even before the PS3 version was released, this is really a lame argument BTW.

Most criticism comes from a very vocal minority, by the way, which is why I don't see this entire discussion going anywhere. You'll always just say "it's like this and that" and there's no objective way to argue with such points.
Bring up some rock hard data on how many people liked procedural content vs. handmade side missions and then maybe we can talk about this, otherwise it's like a religious debate and ultimately pointless.

 

[Prophecy2k]

ME2 sold more copies even before the PS3 version was released, this is really a lame argument BTW.

Most criticism comes from a very vocal minority, by the way, which is why I don't see this entire discussion going anywhere. You'll always just say "it's like this and that" and there's no objective way to argue with such points.
Bring up some rock hard data on how many people liked procedural content vs. handmade side missions and then maybe we can talk about this, otherwise it's like a religious debate and ultimately pointless.

The point about sales was one you raised BTW and tbh is entirely irrelevant to the topic at hand. From the rest of your post it's clear you didn't really understand my original post and the point i was trying to make by bringing up ME1&2 in the first place.

I'm not arguing procedural content vs. handmade. ME isn't a good example of that because there was even more handmade content in ME1 (vanilla) than ME2 (vanilla), so there's nothing to argue there because that's objective fact. My point wasn't even about fan reception of the two games, again you brought up that point, which is again irrelevant to my original point.

My original point was to argue procedural content PLUS handmade content (=ME1) vs. just handmade content but drastically less content overall (=ME2). I'm not making any inference at all on the overall quality of either ME game as which one you prefer is subjective and of course a fools errand to try to argue. My point was to use ME1 & 2 as an example for the use of procedural content to ADD extra side content to a game, over not doing so at all and having a smaller game content-wise.

The fact is that the sentiment in this thread from some people (including yourself) seems to be that procedurally generated content isn't good enough, or is too hard to do, so let's not bother at all. Whereas i'm arguing that whilst procedurally generated content won't ever be as good as handcrafted stuff created by artists, if tools for procedural content generation can be improved to a point where games can be done bigger in a more meaningful way that won't have to require the hiring of 100 new artists on a dev team, then why not explore it as an option? (I also made the supposition that others in the industry may already be exploring these lines).

 

[Laa-Yosh]

Here's a super cool detailed blog about procedural world building:

http://procworld.blogspot.com

 

[Laa-Yosh]

Just found another link, this time on character/weapon customization in Brink:
http://eat3d.com/blog/metalliandy/s...g-above-your-weight-small-art-teams-big-games

 

[archangelmorph]

See this for a specific example of what I'd consider to be useful parameterisation of content.

It wouldn't then be too difficult to build an algorithm that sits on top of a system like this & generates permutations of content procedurally based on your parameterised data sets (i.e. nodes representing assets + relationships between them)

I could certainly see a system like this being capable of extending to support the parameterisation of worlds/levels or any other component of the game.

& then once you have a data set you're happy with you can just bake it all down into fixed native data sets & run with in-game.
Also the great thing about a system like this is it provides effective built-in semantics of authored content (defined directly by the relationships between nodes) that can be used for all sorts of things from environment destructibility to character dismemberment etc. It would all be a matter of how you choose to define your transformed fixed data formats & what information you want to take.

 

[patsu]

May be more interesting to steer the discussion towards the 'hows' rather than 'whether' (How good, how terrible, how do they work).

For auto-gen vegetation, do the tools follow fibonacci sequence when generating the branches and petals ? Or just fractal math that looks mostly correct ?

Off topic but I came in to post this in the name of science... just because it's so cool !

Mathematically generated trees work in real life too:
http://www.amnh.org/nationalcenter/youngnaturalistawards/2011/aidan.html

My conclusions suggest that the Fibonacci pattern in trees makes an evolutionary difference. This is probably why the Fibonacci pattern is found in deciduous trees living in higher latitudes. The Fibonacci pattern gives plants like the oak tree a competitive edge while collecting sunlight when the Sun moves through the sky.

I postulated the same idea to my son when visiting Pisa about a month ago (Fibonacci was buried there, that's why the topic cropped up), but this 13 year old went ahead and confirmed the findings. He gave me much hope about future generations.

 

[Laa-Yosh]

Interview about the making of Deus Ex HR, with a few notes about generic characters
http://www.pixologic.com/interview/deusex3/1/

 

[Fafalada]

Aside from generating clouds and random mazes they don't seem to have gotten very far with it, certainly no where near the ability to replace an artist.

I generally agree(especially considering how AAA production lines typically go through many iterations of content, leaving a lot on the cutting floor to get the high-quality bar).
That said, it's a moving goalpost.

Eg. last generation it was arguably easier to replace one with the other - Spiderman2 used mostly procedural buildings and I doubt anyone was wiser for it (at least mass media never picked up on it), and it certainly didn't look much different from other open-world city games (like GTAs, which at least to my knowledge were all artist made) on PS2 either.