Getting back to the topic at hand, we should also acknowledge the most self-evident drawback of using modular art: the fact that resused art can become repetitive. This leads to what we commonly refer to as “art fatigue”. The median play time of Skyrim on Steam has peaked at well over 100 hours, which is a huge compliment . With that kind of time spent, however, players are bound to notice the same rock or farmhouse or tapestry used again and again. And another two dozen times after. Art fatigue sets in where this repetition becomes obvious and erodes the authenticity of the world.
It wasn't art fatigue that annoyed me. I don't care about seeing the same tapestry or window over and over.
What annoyed me was that after playing for ~10 hours, I started to recognize the entirety of sections (chunks) of levels that were reused. When you're in a cave, and the cave takes a turn to the right and immediately descends down to a rock bridge over a stream - just like the ten caves before it - it kills replayability. For a game that is supposed to reward exploration, designing levels in that manner seems like an odd design choice. It's probably why I played for ~40 hours and then gave up before even finishing the game.
I understand the challenges of scope and the need to find an efficient system to design the world and its levels. But they should have gone with a level randomization similar to Diablo 2. That game also reused art assets, but levels never became repetitive because, except for very specific parts, they were truly random all the time.
Diablo was designed to be repetitive (at least it seemed to me, it's a hack-and-slash game), and the levels are quite frankly bland in comparison to Skyrim, not to mention it's played on third-person view-from-above tiles.
Is anyone aware of randomized level design in a first-person game? It seems like a much more challenging task to get right. If you want to avoid repetitiveness at the scale of Skyrim I'd guess you would need reusable assets on the scale of pieces of a tapestry, and creating unique dungeon layouts would be extremely difficult, as they need to be designed to hold the attention of the player. I believe it's difficult to create artistically interesting and attractive layouts randomly without reuse of large pieces of dungeons (the main pain point in Bethesda's modular style).
I suppose the long-term solution to this problem will be to computer-generate the layouts based on basic level design theory, and then have artists glance at them to check for errors.
Minecraft can auto-generate some pretty interesting cave formations, maybe with a real budget, a team and the advantages of pregeneration you could do a lot better.
The overly similar caves was compounded with the problem of so many quests being oriented around them. Everyone and their mother seems to have lost a valuable family relic or heard of a treasure in some nearby cave. Each one is as predictable as the next.
It's probably why I spent so much time exploring above ground. At least with dragon fights you could enjoy the greater variety in terrain and wildlife.
I've noticed that by creating visually appealing levels that are also random, you pretty much cannot avoid art fatigue.
Diablo 2 had pretty bland levels, so all the details of them, including how copy pasted they are, just get blurred together.
However, if you try to make everything look good, like in the Torchlight games, then you start noticing the world around you and with it all the repetition.
I don't feel like there's a good middle ground here, sadly.
I agree. Despite the fact they said they removed copy-paste elements from the process they used in Oblivion, I found there were much fewer dungeons in Skyrim and that they all felt more similar than those in previous Elder Scrolls games.
I found it interesting that a lot of towers look the same, like ah - the people of Skyrim learned that this tower design works for then, hadn't thought they'd use architecture patterns. Interesting. Rather than, oh I'm annoyed the developers reused a tower (and it didn't happen to me with caves)
"There are 90 people in the photograph below. With the exception of a small number of outsourced assets, you’re looking at the entirety of the dev team across all disciplines. We've resisted the temptation to grow into a multi-studio team of hundreds, such as you often find behind games of similar scope to those we make."
I find it most astounding that Skyrim was made by a team of less than 100 people. It seems that these days it's not unheard of that hundreds or even more than 1,000 people work on a game. An example of the latter category is Grand Theft Auto IV.
There weren't 1000 developers working on GTA IV, certainly not full-time. More like 1000 people who once had at least one small task related to the development, sales or promotion of the title.
Having said that, 8 level designers making 300+ dungeons is extremely impressive.
Most of the time there are contractors, and lots of the art gets outsourced to other companies, so you won't have all the people there. Was QA present?
The post discusses assets and creating asset modules for constructing their virtual worlds. It would be interesting if they discussed hardware constraints, and how lifting those constraints effects the creative process. The release status of an unreleased console is inconsequential to the information. It has been announced that the new playstation console has 8gb's of ram. In my experience playing Bethesda's games on console and PC, consoles take a significantly longer time to load. Even if a current-generation console is more powerful than a PC, it will be much slower than a PC at loading assets. This has nothing to do with hard-disk speed, as running the game off of equally slow HDD's of a console and a PC, the PC will still significantly quicker. Someone voted me down for my comment. I was not expecting them to cover this in a post about an already-released game. I just said it would be interesting (in future).
We all have to be really careful when commenting on the internet, especially Hacker News, that we are clear. For that I apologise, even if I am quite a bit annoyed, though I realise it is my fault. I have been toying with the idea of building an online community, but I think I will pass on that idea as I don't like excluding people. What did I offer this discussion? = nothing. What did most other people? = nothing. Why are we even discussing this?
15 comments
[ 4.9 ms ] story [ 42.6 ms ] threadIt wasn't art fatigue that annoyed me. I don't care about seeing the same tapestry or window over and over.
What annoyed me was that after playing for ~10 hours, I started to recognize the entirety of sections (chunks) of levels that were reused. When you're in a cave, and the cave takes a turn to the right and immediately descends down to a rock bridge over a stream - just like the ten caves before it - it kills replayability. For a game that is supposed to reward exploration, designing levels in that manner seems like an odd design choice. It's probably why I played for ~40 hours and then gave up before even finishing the game.
I understand the challenges of scope and the need to find an efficient system to design the world and its levels. But they should have gone with a level randomization similar to Diablo 2. That game also reused art assets, but levels never became repetitive because, except for very specific parts, they were truly random all the time.
Diablo was designed to be repetitive (at least it seemed to me, it's a hack-and-slash game), and the levels are quite frankly bland in comparison to Skyrim, not to mention it's played on third-person view-from-above tiles.
Is anyone aware of randomized level design in a first-person game? It seems like a much more challenging task to get right. If you want to avoid repetitiveness at the scale of Skyrim I'd guess you would need reusable assets on the scale of pieces of a tapestry, and creating unique dungeon layouts would be extremely difficult, as they need to be designed to hold the attention of the player. I believe it's difficult to create artistically interesting and attractive layouts randomly without reuse of large pieces of dungeons (the main pain point in Bethesda's modular style).
Minecraft can auto-generate some pretty interesting cave formations, maybe with a real budget, a team and the advantages of pregeneration you could do a lot better.
It's probably why I spent so much time exploring above ground. At least with dragon fights you could enjoy the greater variety in terrain and wildlife.
Diablo 2 had pretty bland levels, so all the details of them, including how copy pasted they are, just get blurred together.
However, if you try to make everything look good, like in the Torchlight games, then you start noticing the world around you and with it all the repetition.
I don't feel like there's a good middle ground here, sadly.
"There are 90 people in the photograph below. With the exception of a small number of outsourced assets, you’re looking at the entirety of the dev team across all disciplines. We've resisted the temptation to grow into a multi-studio team of hundreds, such as you often find behind games of similar scope to those we make."
I find it most astounding that Skyrim was made by a team of less than 100 people. It seems that these days it's not unheard of that hundreds or even more than 1,000 people work on a game. An example of the latter category is Grand Theft Auto IV.
Having said that, 8 level designers making 300+ dungeons is extremely impressive.
Here are the credits (Xbox 360) - http://www.mobygames.com/game/xbox360/elder-scrolls-v-skyrim...
That's ridiculous.
We all have to be really careful when commenting on the internet, especially Hacker News, that we are clear. For that I apologise, even if I am quite a bit annoyed, though I realise it is my fault. I have been toying with the idea of building an online community, but I think I will pass on that idea as I don't like excluding people. What did I offer this discussion? = nothing. What did most other people? = nothing. Why are we even discussing this?