Show Posts
1351
Level Design / Re: Opinion on designing levels
« on: February 16, 2016, 12:06:54 AM »
As you already can see from the previous posts, everyone has a different approach to level creation. My own workflow is usually like this:
1) Get level idea
More often than not, I get level ideas while playing other levels. I try out something, but it doesn't work. But the lemmings interact in an interesting way, skills get combined in an unusual way, ... in other words: stuff that cries "turn me into a new level".
2) Strip the level idea of anything unnecessary
When first encountering some level idea as above, I usually made many skill assignments before, there is lots of terrain around, ... But I want that players of my level to see the level idea in the clearest possible way. So I ask myself the questions: What are the key ingredients contained in the level idea? What is the minimal terrain layout where this can work?
To answer them, I step completely away from the computer and "play" the level purely in my mind. For me this is the fastest way to modify the imaginary level, strip away parts and see what happens, add some piece of terrain to see how this influences the level, let the lemmings wander around without having to care about their precise timing, ... For very complex ideas, I make rough sketches on paper once the broad layout is fixed.
This step might not be absolutely necessary. But I found it useful to reduce the amount of unintended solutions. In the following steps, the levels usually gains quite a bit of complexity again, and the more complex, the more potential for backroutes.
3) First rough design in the editor
Now I return to the editor and put terrain where I imagined them. I do not worry about the looks at all and in this state the levels usually look like an unholy mess. It is meant to check, whether my imaginary level can be turned real. If some lemmings have to meet at a certain point, I have to adjust the lengths of paths they walk; if a miner should reach point X, I have to worry that this is actually possible, ...
4) Do I see easy alternative solutions?
Changing the final level with all the terrain carefully fitted together is quite a bit of work. Therefore I found it useful to start worrying about alternative approaches/solutions quite early. So I sit down (usually after taking a break) and try my best to backroute my own level. Whenever I find one, some rough modification is made to the level to remove it.
5) Prettify the level
Now that I have a level that more or less works as intended, but looks totally unappealing, it is time to prettify it. Often I find, that the style I chose at first is not really suitable to turn the level's terrain into some pretty terrain. So I start another editor and recreate the level from the beginning, but now properly fitting the terrain pieces together.
6) More backrouting
Prettifying a level usually changes lots of details, so backroutes may creep in again. I found it useful to have a whole day break before starting backrouting again.
7) Make intended solution easier to execute
As the level designer I already know, where to start the builder, miner, ... but all other players don't know. So I go through the whole solution and always ask myself: Does the solution still work if I assign the skill a few frames earlier or later? Quite often changing a few pixels can make a huge difference to this question.
8) Let other people play the level and encourage feedback
Other players might see problems, ways to improve the level, ... that I missed. Not to mention that they will almost certainly find more backroutes that needs to be fixed
.
Some final advice:
- Take your time! People are much more likely to play your levels, if they see that you put much effort into it. I am probably one of the slower level creators here and even for a one-screen level I rarely need less than 4-5 hours (not counted any time I spend backrouting my level).
- Do not be discouraged if your first levels have not quite the quality you wish for. Of the first 20 levels I made, only 5 were ever seen by other people (and then only after heavy editing once I got more experienced). Of course there are other people, who create excellent levels right away.
1) Get level idea
More often than not, I get level ideas while playing other levels. I try out something, but it doesn't work. But the lemmings interact in an interesting way, skills get combined in an unusual way, ... in other words: stuff that cries "turn me into a new level".
2) Strip the level idea of anything unnecessary
When first encountering some level idea as above, I usually made many skill assignments before, there is lots of terrain around, ... But I want that players of my level to see the level idea in the clearest possible way. So I ask myself the questions: What are the key ingredients contained in the level idea? What is the minimal terrain layout where this can work?
To answer them, I step completely away from the computer and "play" the level purely in my mind. For me this is the fastest way to modify the imaginary level, strip away parts and see what happens, add some piece of terrain to see how this influences the level, let the lemmings wander around without having to care about their precise timing, ... For very complex ideas, I make rough sketches on paper once the broad layout is fixed.
This step might not be absolutely necessary. But I found it useful to reduce the amount of unintended solutions. In the following steps, the levels usually gains quite a bit of complexity again, and the more complex, the more potential for backroutes.
3) First rough design in the editor
Now I return to the editor and put terrain where I imagined them. I do not worry about the looks at all and in this state the levels usually look like an unholy mess. It is meant to check, whether my imaginary level can be turned real. If some lemmings have to meet at a certain point, I have to adjust the lengths of paths they walk; if a miner should reach point X, I have to worry that this is actually possible, ...
4) Do I see easy alternative solutions?
Changing the final level with all the terrain carefully fitted together is quite a bit of work. Therefore I found it useful to start worrying about alternative approaches/solutions quite early. So I sit down (usually after taking a break) and try my best to backroute my own level. Whenever I find one, some rough modification is made to the level to remove it.
5) Prettify the level
Now that I have a level that more or less works as intended, but looks totally unappealing, it is time to prettify it. Often I find, that the style I chose at first is not really suitable to turn the level's terrain into some pretty terrain. So I start another editor and recreate the level from the beginning, but now properly fitting the terrain pieces together.
6) More backrouting
Prettifying a level usually changes lots of details, so backroutes may creep in again. I found it useful to have a whole day break before starting backrouting again.
7) Make intended solution easier to execute
As the level designer I already know, where to start the builder, miner, ... but all other players don't know. So I go through the whole solution and always ask myself: Does the solution still work if I assign the skill a few frames earlier or later? Quite often changing a few pixels can make a huge difference to this question.
8) Let other people play the level and encourage feedback
Other players might see problems, ways to improve the level, ... that I missed. Not to mention that they will almost certainly find more backroutes that needs to be fixed
.Some final advice:
- Take your time! People are much more likely to play your levels, if they see that you put much effort into it. I am probably one of the slower level creators here and even for a one-screen level I rarely need less than 4-5 hours (not counted any time I spend backrouting my level).
- Do not be discouraged if your first levels have not quite the quality you wish for. Of the first 20 levels I made, only 5 were ever seen by other people (and then only after heavy editing once I got more experienced). Of course there are other people, who create excellent levels right away.
