Okay, so. On this guide I'll show you how to make a simple, challenging map.

This will be the final map.

It is recommended that you play it before reading the guide. You can attempt to complete it, or just look at the mechanics and see how the chamber flows.
http://cs2bus.com/sharedfiles/filedetails/?id=115039114
Well, that's out of the way, so let's get started.

Okay, gentlemen. So, when you want to make a good map, the first logical step is getting inspiration. Here are some things I recommend:

First, play the whole campaign. This will ensure that you have basic knowledge of the game mechanics, and will also show you valve's-- I mean, Cave Johnson and GLaDOS' map making skills.

Then, go play some random maps. Play them and analyze them. Look for what they did right, and what they did wrong. You can usually learn more from bad maps, considering you're trying to be original here, not make a copy. Here are some examples I snatched from quickplay:

http://cs2bus.com/sharedfiles/filedetails/?id=115009290
Classic maze map, except with one extra dimension of frustration. This kind of puzzles offer no mental challenge, just frustrating walk around and find the exit after five hours. Very poorly lit, I even had to use portals to let me know where the walls were sometimes.


http://cs2bus.com/sharedfiles/filedetails/?id=115009828
This map has literally no portalable surfaces. It was a complete no brainer and offered no challenge except on a part that involved turrets, and the challenge was mostly about me attempting to jump in front of a moving platform at the right time and walking a bunch of time synced up with it. The concept itself is good however, but it needs more thinking and less timing.
Oh, and when I had all of the buttons pressed and laser recievers recieving, the door did not open. I have not yet found out what left is there to do, considering the door shows no additional connections.


http://cs2bus.com/sharedfiles/filedetails/?id=115013278
This map is probably one of the few out there where you can actually place portals. However, the puzzle is overly long and relies mostly on timing and speed, and has some unneeded light bridges and gels.

Okay, so you've got inspiration now. But where to start your map?

Well, for starters, you're going to think of one or more mechanics you'd like your map to have.

In our case, let's go for the classic and versatile buttons and lasers puzzle. These usually offer a wide range of ways to make your map, and can provide for simple, yet challenging and mind stimulating puzzles. I usually use this combination.

I made a basic, solveable test chamber. But it doesn't quite have the difficulty that we're looking for.

Let's just scrap this and start over. Most puzzles look nothing like they started as, unless you have a very concrete idea of what you're going to do. Wich we don't.

Let's ramp up the difficulty a bit by adding an emancipation grid. We also need to move stuff around.
Remember, challenge doesn't neccessairly equal danger of death.

After moving some stuff around, I end up with this.

Okay, we got a good, finished test chamber now... Right?

Not at all. I tested the chamber and found out the test chamber is impossible! Thus, I move some stuff around again.

I made the laser emitter activate with a new button on the top. The button is on an infnite timer so people can take their time. This button, along with the laser reciever triggers a cube dropper.

I also move the button back, because it was getting frustrating to adjust the cube without line of sight to where the laser was pointing.

I decide the button isn't doing anything there but complicating the antline maze, so I remove it. I also make the cube's location not need you to portal anywhere, so that you don't need to displace your pretty placed portals to reach it.

I procceed to playtest it again.

However, the weighted cube proves to be there for pretty much nothing, so I take a different approach.


I think this has finally become a worthy challenge. Or, well, even if not much of a challenge, still a good puzzle.

Okay, we have our chamber set up, a nice easy but enjoyable chamber... Nothing left, right?

Wrong.

For starters, this chamber is full of shadows. One light source doesn't cut it.
Add more lights and observation rooms.
Use the latter sparingly though. A good idea would be to use it as a replacement for unportalable walls, as they can ruin the aesthetics of the room sometimes.

Then there was another, slight problem...


After noticing these problems, I fixed them by adding lights and moving the exit door.

It's time to publish!

This part is mostly up to you, but I usually like to make a title that gives a good idea of what the chamber is, and make a related(ish) GLaDOS or Cave quote on the description. That's just me. Be original, I guess.

Okay, so. You've made your chamber. You've refined it. You've lit it up. You've completed it. You've published it.

Now just answer some comments and forget all about it, right? I mean, even if you, say, get comments like...



This can't be good. I can't ignore them, now can I?

So, since the main complain seems to be the way you have to adjust the box until it's just right, let's think a bit. What could replace moving the box around...
What about placing portals in different places?

With this in mind, I procceed to move the observation room.



I realize, however, that it is still incredibly hard to actually hit the small target. So I go ahead and decide to place the button in a place where it would be easier to actually see what's going on.



I use a faith plate so that you do not need portals to get up there, and make use of light strips as to avoid portal placement.

I find a bug, and fix it by swapping the dropper.[/screenshot]

I think this is good to go, I ship it, and I hope that people will no longer have trouble fiddling with the box, considering it's so much easier to do so with portals.

Okay so, to finish this guide off, here are some tips and tricks:

• Players almost never look up.
  
• Players don't like not knowing what a button does.
  
• Harder does not always mean better
  
• Don't bother portal bumping proofing your map; if players want to cheat let them cheat: they might as well just turn on noclip.
  
• Make your test chambers possible without glitching or contraptions like cube stacking.
  
• Don't allow players to be stuck in your chamber: kill them if you must.
  
• Make the player think with portals: if your chamber relies mostly on other mechanics, that makes the whole main mechanic of the game redundant
  
• Remember to have the chamber well lit, but not excessively lit. Use light strips for small corners and observation rooms for big halls.
  
• Don't place observation rooms on ground level: it not only makes the floor near it overly lit, but it also simply looks bad. Place observation rooms in areas where one would be logical.
  
• Have your chamber be only one chamber, don't have a chamber with several separated rooms or a bunch of cubes you have to collect. If you feel like having a bunch of several puzzles, just make them separate chambers or find a way to connect them

If you have any more tips or tricks or any feedback on this guide please post it in the comments, hope this will help you make better chambers in the future.

Credit to Inspirata for some of the tips.