I believe a Portal 2™ map is divided into 3 main parts. First is the concept, the core of the map; that creative thing that makes your map special. Second is the puzzle, the sequence of events the player must go through. And third is the chamber, the room itself; the aesthetics, the existence of the puzzle not as an idea but as something interactive. Realizing these 3 parts is important for me when creating a puzzle, and so will it be for you to understand this guide.

Hope you enjoy reading this guide as much as I had writing it.
Mr. P. Kiwi

As we all know, this is the hardest part when designing a puzzle. Where do we start?

First thing we should do is to think of a concept. A small idea to be the main challenge. To know if your idea is good, try answering these questions:

• Why is your puzzle different?
  
• What makes your puzzle creative?
  
• What is the core of the chamber you're designing?

Answering these questions will help you define your puzzle. If you can't answer these questions you might need to re-think your concept. You want to create something new and original, not the same puzzle someone else has made. I for example designed: Diversity Gels. I wanted to create a map, something original no one had ever seen before. It's a bit ambitious, I know. But I was confident.

To come up with such a concept, first you must play a lot of maps. Gather ideas, see what works and what doesn't. Then google a lot of concepts for Portal. Maybe you'll find a concept for something that was not shipped with the game and you'll find inspiration in it. Now ideas of all kinds of weird concepts should be flowing.
I, for example, was fascinated by the un-shipped pneumatic diversity vent. However I felt they were lacking something - if they can suck walls and turrets why not players? So here began my quest to make a puzzle that revolves around the player being able to get sucked by the vent. A concept was formed.

Your concept can come in all sorts of ways: an unintended solution, a problem you had in someone else's concept or even through fan art. So keep your mind open to the possibilities. If I hadn't over thought the diversity vents promotional video I might have never got the chance to make Diversity Gels. Also shipped mechanics might work well. The addition of gels to a diversity vent map came from one level in Portal 2's single player campaign:


Now, with your new concept at hand, try answering those questions again. If you have a clever concept idea: great! Planning before starting to map is important in making great chambers.
If your concept uses only entities found in the PeTI editor - you can go ahead and work on it there. However, if your concept is something truly unseen before, something you can't make in the in-game editor Hammer should be able to make your dream a reality. There's a great community of mappers out there who would be happy to help you get your concept to work in-game.

With a concept in mind we should now think how to make it into a puzzle. A concept alone can't be a puzzle; elements must be used in conjunction with your concept to create a puzzle. Details of the finer workings of the puzzle must come as it evolves. After settling on a concept, it's time to think what elements will help make that concept an interesting puzzle. You'll need of course to take your concept one step further and understand what you can do with it.

The concept for Diversity Gels meant that not only cubes and turrets can get sucked by the vents, everything can. With this realization, it was obvious I had to make more of the idea, and so transporting gel through the vents was formed into a puzzle idea: the vent will suck the gel to a room where the player is unable to use it to exit the chamber. To get the gel to the correct chamber he must deactivate the vent by solving a puzzle. Now I was getting somewhere!

After deciding what elements you want to use, and what are your rough "mini-puzzles" start thinking how to solve your puzzle from the end to the start.

• How do I reach the exit door? - I place the gel near the raised platform.
  
• How to I get the gel to the main room? - I shut down the vent.
  
• How do I shut down the vent? - I block the laser that's activating it.
  
• How to block the laser? - I should probably use the concept as a means to do that.

And so forth. Just like you would solving a puzzle.

This is called the "top-down design". It is well suited to introducing a new mechanic, a new concept, and everything else follows from it. The most important thing is not losing your mental build of the map, which is easy to do when there are several steps involved in putting the idea into practice. There are many, many steps in Portal level design.

Note: You want to expand your map by adding these mini-puzzles. If you think your puzzle is too simple or short, just try adding more steps to the solution. The best use of these mini-puzzles is if they are built one atop of each other, that solving one is essential to solve the next, but timing of when to solve each one and how to do it is the real challenge.

Tip: You want to challenge the player, not frustrate them. Try to make the exit and all the information one needs to solve the puzzle clearly visible. The player should understand what he must do to exit, but be challenged when thinking how.

What I did was to design other mini-puzzles to make the most out of each element. You might notice the gel must be used in both rooms and so must the cube be used in both rooms at some point. Each element is used more than once, and every time slightly differently. You should keep in mind that your concept must still remain the core of your puzzle, as it's the original bit in it that makes it different. And so the mini-puzzles should be designed to complement it. In my map to solve the puzzle you must realize what the vent does (by observing the gel) and send the cube through the vent as well.

Now you have a concept, and a puzzle around it.

Well, to be honest, we don't have much of a plan now. But the hard part is behind us. After we've settled on a puzzle all it means is that you know what the player should do. Now before you forget your idea I suggest grabbing a piece of paper and a pencil and sketch down a rough draft of your puzzle. So even if you do forget you'll have somewhat of a reminder. You need to think where each element should be for the test to work, and sketch according to that. If you're designing the map in hammer, now would be a good time to consider optimization and aesthetics.

Still, our chamber is incomplete. Even with an idea at hand and a rough draft of your puzzle we still need to see if our test is possible. This is when we start to map! You need to make something that's called a "proof of concept" (POC, for short). It is, in our case, a small map showing that your concept is feasible, that it can be made with the tools the engine gives you. A POC is usually small and may or may not be complete. This will require some problem solving and a lot of creativity. In my case the proof of concept didn't need any work from me. The gel running through the tubes in Valve's official maps were enough for me to know it is possible to make such a thing. All I had to do is to decompile that map and check how they did it there.

Tip: There are a lot of tutorials and mappers[www.thinkingwithportals.com] that will be willing to help. So if you are stuck you should ask them.

After getting your concept to work on a small scale you know that it won't be a waste to build the whole chamber. I like drawing my chambers before I build them. It helps me stay focused on mapping and not designing while I'm in hammer, and so it saves time. I suggest getting the best version you can on paper before starting to map. You might also want to "solve" your puzzle a few times to see there are no unintended solutions.

Now you've got a plan, and all that's left is to build, test and publish. Let the community tell you how awesome your concept is and if any fixes are needed, don't hesitate. Fix them and re-publish. The Workshop is great for things like that. I hope this helped you get on a track to make a new chamber.

This is not "my idea". Many other mappers had their hand on such methods for a long time. This merely describes my interpretation of the technique from tutorials and overall personal experience mapping. I wrote this guide from scratch and did not meant to offend anyone who wrote similar ideas. I would also like to stress this is not an effort to say that this is the only way to design Portal chambers, but it is rather a useful one.

For your comfort I had searched the web for you. This is a list of more articles on the subject I found that you might also find interesting:

• Valve Developer Community
  
• Gamedev - tuts+[gamedev.tutsplus.com]
  
• relinquish152 - Blog

Looking forward for your maps,
Mr. P. Kiwi