In the end its a simple economic calculaton:

Artstyle-To-Use must be choosen by:

(AverageTimePerAsset / NumberOfArtists) * NumberOfAssetsNeeded <= TimeAvailable

with: NumberOfArtists * MonthlyPay * TimeAvailable(Month) < Overall Budget

-> AverageTimePerAsset (smaller for Comic/Abstract look)(longer for realistic / detailed look)
-> NumberOfAssetsNeeded (should be estimated early by the Gamedesign planning)

-----

Thus, since Indy Devs have many more limits (often just one Artist, and also a limited Time - before Frustration stops the project)
they should either Choose a game where less Assets are needed
(Casual Games, Games where Assets can be repetatly used - like in a tilebased gameworld)
or use a quick-to-develop Artstyle (Comic / Stylized look with low details)

But I am afraid that many of you have a wrong idea about the costs of art.

I can say from my own experience that todays realistic art is not more expensive than a few years ago. It is the other way around. Tools like Modo and ZBrush allow to create in hours what we had to built in days or even weeks with old-school polygon modellers like Max, Lightwave and Co.

I can model a texture faster than I painted it in the past.

The tools are just better, we use less time. Graphic automatically became more detailed this way.
I would even go exactly the same route for a cartoon asset. So I don't see any difference here. And cartoon models can look way more rounded and cartoony when I use some proper normalmaps and modern technology.

a good point was made here. concepts and coherent style.
thats what i miss with most indie games. you need to have a cocnept, an idea how your game should look.

also, another good point wwas level design. i can clearly tell through the level design of some games if a pro was at hand or a trainee. a good level designer creates a level that consists of unique rooms. none of them look the same, and backtracking is reduced to the minimum the game requires. i find it often annoying in indie games that you have to run through the level a gazillion times to lengthen the game. this just sucks. reduce backtracking. its bad level layout and bad gameplay, and gameplay is frankly, all we have...

I'm not primarily an artist, but I do have some experience with sculpting, UV mapping, baking normal maps, texture creation and so on. I remember first hearing of normal-mapping technology (many years back) and being dismayed at the inevitable increase in difficulty of asset creation.

To my surprise and relief, such a natural and real way to add per-pixel details (as well as continual advances in software such as Blender) actually made it easier and easier as the years went by.

So yeah, I'm not an experienced artist like you, Frank, but it's nice to see my impressions of the art side of things reflected in your experience.

Jibb

I am not a Blender user but I saw some videos and it looks indeed very good. The tools for sculpting and re-topology are looking fast and convenient. I hope they finish this release anytime soon. I would like to check it and I am also interested in the animation tools. I could imagine that they became better with this new open video project.

speaking about metroid prime, aren't we?

metroid prime is a "prime" example. not only do they have a uniqueness, they also tell the whole story through level design. unlike HALO which has to have the worst level design of the last generation (right after the PS2 castlevanias)

As I first saw this game I instantly thought this is a rip-off of the good old Turrican that I played on C64. So I did not find it unique, but in the end it was well made.

It depends on the type of game if "going back" in a level
should be used.
In a free Gameworld such as Stalker, its part of making
it an exploration world.
Very liniar Games (like Fear, Halflife) should not
send the player back.
But I dont really like liniar games.
The players choice is limited down to follow the
path the gamedesigner defines stricly, and maybe kill
one monster more or less.

I thought the level design in Metal Gear Solid 1 was brilliant. Stealth was very important, and you'd have to go back to the same areas from different directions, and this often had a big effect on how you'd need to go about choosing hiding spots and sneaking past enemies.

Also, whenever you were forced to go back, elements changed to reflect progress in the story.

For many games, each level feels disconnected from the plot. For MGS1 the level design felt like it was part of the story, and vice versa.

I've ashamedly never played a Metroid game, but I imagine that's the kind of thing sP|Ke means by telling the story through the level design.

Jibb