Puzzle Design Methods
Quote from Test Subject #54067 on July 29, 2014, 11:05 amA lot of the maps that I download on the workshop have very good puzzles, which involve repeated backtracking and preserving your portals with really devious solutions. Examples are BEN 77's map Testing, Wheatley test 1, and Bridge; BanJaxe's Decay part 2; and josepezdj's und3CIDE. I have never managed to make such a puzzle on my own.
What strategies do you guys use to come up with puzzles?
A lot of the maps that I download on the workshop have very good puzzles, which involve repeated backtracking and preserving your portals with really devious solutions. Examples are BEN 77's map Testing, Wheatley test 1, and Bridge; BanJaxe's Decay part 2; and josepezdj's und3CIDE. I have never managed to make such a puzzle on my own.
What strategies do you guys use to come up with puzzles?
Quote from Fluppy on July 29, 2014, 1:09 pmBasically you can use one of these two ways of making a puzzle:
1st approach: You can make a large chain of little puzzles where you basically let the player to solve one part and then he is required to reuse everything which can be reused (all spare cubes and elements) in the other part of the chain and so on. You are usually required to leave a portal in the new section and then go through the portal to return to the previous part and take the elements with you to the next section. By stacking the parts like this you can get very devious puzzles because you have to think about the every move you make.
The problem with this approach is that the layout of the map gets all messy very easily and the puzzle becomes frustrating. As a player you usually don't know where to start and you have to explore the test chamber which is not always possible because of how the chaining of the areas works. Sometimes you don't even know where the exit is so you don't have a visible goal in front of you. The worst thing which can happen is if you for example destroy a cube and have to return all the way back and start over. That's why I don't like a lot of maps made in this way, I think it artificially increases the difficulty.
2nd approach: This is a way which I prefer and you see it in puzzles by people like Mevious and Gig - you have a compact test chamber which you can easily explore. There are usually 2-4 testing elements (it depends on what you count as a testing element) and the difficulty of the puzzle comes from finding the correct combination of the elements.
Puzzles like this usually have a clear goal and you see all the tools you can use. This usually leads to that great feeling that you know what you have to do to solve a puzzle but you don't know exactly how to do it.
So how to make a map like this?
- Start by choosing a couple of testing elements which you are going to use. The most commonly used testing elements are funnel and laser, because they can be used in many ways.
- Try to come up with a cool way of combining these elements (this will be the 'core step' of your map).
- Find a few more uses for the elements (it's good if each element has at least 2 different uses in the test).
- Playtest a lot and make sure there isn't any unintended solution which would skip the 'core step' of the map.
Ideally the puzzle designed like that will lead to a state in which player thinks "Oh I have to do this, but in order to do that I have to do that - but how am I supposed to do both things without ruining my current state?". The solution to this problem will then be the 'core step' from the step 2.
I'm not sure if my explanation makes sense, feel free to ask me if you want to clarify anything!
Basically you can use one of these two ways of making a puzzle:
1st approach: You can make a large chain of little puzzles where you basically let the player to solve one part and then he is required to reuse everything which can be reused (all spare cubes and elements) in the other part of the chain and so on. You are usually required to leave a portal in the new section and then go through the portal to return to the previous part and take the elements with you to the next section. By stacking the parts like this you can get very devious puzzles because you have to think about the every move you make.
The problem with this approach is that the layout of the map gets all messy very easily and the puzzle becomes frustrating. As a player you usually don't know where to start and you have to explore the test chamber which is not always possible because of how the chaining of the areas works. Sometimes you don't even know where the exit is so you don't have a visible goal in front of you. The worst thing which can happen is if you for example destroy a cube and have to return all the way back and start over. That's why I don't like a lot of maps made in this way, I think it artificially increases the difficulty.
2nd approach: This is a way which I prefer and you see it in puzzles by people like Mevious and Gig - you have a compact test chamber which you can easily explore. There are usually 2-4 testing elements (it depends on what you count as a testing element) and the difficulty of the puzzle comes from finding the correct combination of the elements.
Puzzles like this usually have a clear goal and you see all the tools you can use. This usually leads to that great feeling that you know what you have to do to solve a puzzle but you don't know exactly how to do it.
So how to make a map like this?
- Start by choosing a couple of testing elements which you are going to use. The most commonly used testing elements are funnel and laser, because they can be used in many ways.
- Try to come up with a cool way of combining these elements (this will be the 'core step' of your map).
- Find a few more uses for the elements (it's good if each element has at least 2 different uses in the test).
- Playtest a lot and make sure there isn't any unintended solution which would skip the 'core step' of the map.
Ideally the puzzle designed like that will lead to a state in which player thinks "Oh I have to do this, but in order to do that I have to do that - but how am I supposed to do both things without ruining my current state?". The solution to this problem will then be the 'core step' from the step 2.
I'm not sure if my explanation makes sense, feel free to ask me if you want to clarify anything!
My workshop
How to control other people
[spoiler]LOL, I've just forced you to click on the 'SHOW' button! Now click here: http://soundcloud.com/vojtasiman/[/spoiler]
Quote from Test Subject #54067 on July 29, 2014, 1:46 pmThanks, I'll try that soon.
P.S. I really liked your test "Multipurpose".
Thanks, I'll try that soon.
P.S. I really liked your test "Multipurpose".
Quote from Fluppy on July 29, 2014, 2:06 pmMultipurpose is example of the 1st method - it's a chain of events you have to perform and you have to reuse the elements multiple times. I tried to fix everything I didn't like about that 1st approach in the map (I tried to make everything visible and accessible).
Let us know about the results after you try these methods!
Multipurpose is example of the 1st method - it's a chain of events you have to perform and you have to reuse the elements multiple times. I tried to fix everything I didn't like about that 1st approach in the map (I tried to make everything visible and accessible).
Let us know about the results after you try these methods!
My workshop
How to control other people
[spoiler]LOL, I've just forced you to click on the 'SHOW' button! Now click here: http://soundcloud.com/vojtasiman/[/spoiler]