[Tip]Making a Hard Puzzle
It can be used effectively to make an area you can't pass, but it should be rarely used as punishment for executing a manoeuvre incorrectly. The same with deadly lasers.
Well put. If I die in a map I want it to be because I did something stupid, not because I failed to pull off the thing that I actually was meant to do.
On a more general note, I don't think 'complex' necessarily equates to 'hard'. A puzzle can have a lot of elements to it but still be completely straightforward, and I think a lot of mappers (certainly on the Workshop) tend to make puzzles that are more confusing than they are hard. As a general rule of thumb, I like to always give the player some idea of what they're trying to do next, in terms of 'get that cube over here' or 'get the lightbridge up there'. Exploring is fine for a short while, but spending 20 minutes wandering around a huge map trying to work out which bit you're supposed to attack first is not fun, it's a chore.
I've nothing to do right now, so I might just explain the puzzle-making process for my latest map on the off-chance anyone's interested (Gymnasium, it's in my sig). If you're going to read all this you might want to play it first as a) I'm totally about to spoil the whole thing and b) it'll be really difficult to understand what I'm talking about:
Brace yourself:
a) directly, with portals beneath the dropper, or
b) by carrying the paint over with a funnel and dropping it from there.
I wanted to find a third way, so I decided to fling the paint. The problem is, it's actually very difficult - putting it in an infinifling just fires it at the wall like a bullet - and the best I could think of was to hold it in a vertical funnel before dropping it through a portal to ensure a nice even spread.
Fine, so I do need a funnel, but I need to prevent simply carrying the paint over in it. The solution to that was a timer. I also need a goal for the puzzle, and so I could double this up as the door timer! Sweet. Now I have to set about making things a little less obvious....
My usual schtick is to add a small preliminary puzzle to introduce the test elements. For my paint fling to work I need quite a tall chamber, so a big hall with a fling seems the obvious choice. I start off with a fling from dropping out of a funnel, but the part of me that wants to re-use everything starts thinking about that paint... I end up making a chamber that requires a run-up through a portal and a couple of high-speed flings (was actually a major pain in the arse to get all the distances right). I make it so that the funnel is used both to gain access to the paint in the first place and to spread it on the runway. I also have a fizzler to stop the player getting to the second part without the fling, but this needs to be switched off so they can use the funnel for the solution.
Now my puzzle's pretty much complete, but after nixing all the unintended solutions I'm left with very little portalable wall left, and I'm worried this'll make everything too obvious. I really want to have a load of useless walls around the runway to the exit, but this presents a problem - it would be possible (albeit arduous) to keep sending paint through the funnel and painting it in increments. I need some way of stopping the funnel from ever reaching this room, though I'd like to be able to use paint in there (even though it's useless) to distract the player....
After a lot of thinking and many stupid ideas, I come up with the double switchable fizzler idea. There are two requirements:
- The switch must be behind the second fizzler to prevent the funnel being used in the room
- The gel must also be behind the second fizzler (or the whole thing is pointless anyway and it might as well be a permanent fizzler or black walls).
The problem with this is that I still need to bring the gel to the funnel for the solution, which is impossible now! I was tempted to throw the idea away, but I kept thinking a bit more. I needed a way to activate the switch and be on the other side of the fizzler. A timed button would do the trick, but that would be dull and obvious. A little more thinking and I came up with what I have now, which is actually my favourite part of the map!
I guess the point of this long and no doubt quite dull post is that you shouldn't throw away ideas just because they don't work right away. Often my workarounds for various problems end up being the best parts of my maps. I've frequently thrown away my original idea because an exploit was more fun, and indeed a lot of the puzzles in my next map will revolve around ways I found to break this one. My favourite thing about mapping is that it involves at least as much problem solving as actually playing the game!
My maps: Double Bill / Jam / Oval Window / Gymnasium / Gymnasium Part 2 / Minimalism (collection) / Resolution
All feedback welcome!
---------------------------
Wow El! That story is so damn ilustrative!! It REALLY reminds me of what happened to me TWICE with my mappack! Now I've got 2-3 parts in 2 of my maps that I truly love (humbly). After some beta-testers did a few ways that was not necessarily intended, and also after me mmyself thought that a couple of parts were too easy or obvvious, I left the maps right there and continue mapping other maps for the mappack. Well then one day I suddenly opened one of those old maps (that I was indeed about to discard!) and the mmiracle happened. I thought this way (just in case some else could take advantage of this story):
- So I have to get the cube out of here but building simply a couple of fizzlers that alternate their status (while one opens, the other closes) is too easy, I only make the player sit down and wait for 2 secs... Hmmm that's stupid!
- Ok, which elements do I still have in the map which are reacheable in any way? a funnel and a white floor panel... Great!
- So I'll make the player have to think out of the box and figure out that he has to bring here the funnel and funnel the cube out!
The point is: it's sometimes easy to make a map harder by simply forcing the player to use some puzzle elements that, under certain circumstances could be reacheable and useable. It gets harder as further those elements are or you make the player believes he is all done with them!
useful tools and stuff here on TWP 
[spoiler]
[/spoiler]
Totally agree with Jose!
Using every puzzle element at least twice is my basic concept for every map I make.
I was thinkig about making a video tutorial on how to make challenging puzzles.
There are a lot of tutorial on using hammer or the new puzzle creator but I couldn't find any that actually shows how to create good puzzles. Since mapping became very handy and easily accessible for everyone why not sharing the experience I already made and helping others to make well conceived puzzles. There are a lot of things that can be tought. Like elaborating a basic idea, combining puzzle elements and using them in a different way or simply the desing of your map. For exaple I noticed that a symmetric design takes away a lot of that confusion and disorientation el farmerino was talking about.
I think I'll give it a try next week since that will require some time. I'll let you know if I make progress and will post here.
Marc, I'm SO looking forward to it!!!! It'll be very handy to all of us! 
useful tools and stuff here on TWP [spoiler]
[/spoiler]
To the OP: thanks for sharing! Nice thread...
Since I have relatively more experience now, I'll give my two cents' worth:
1. One can take a logical puzzle and implement it using buttons+cubes, fizzlers, death fields and logic gates. So based on the limit of applicability, one may take as hard a logical puzzle as one may please (for example the standard grid puzzles).
Translating and implementing such puzzles in terms of a Portal 2 map are quite hard though.
(Logic gates can be made using combinations of lasers, catchers, flip panels, etc. There are already good examples featured in the workshop)
2. Properly disguising the solution is really important in my opinion. I don't exactly know any fool-proof method of doing so, but mainly this should vary from map to map.
3. Finally there are "tricks" or unconventional use of various testing elements. There are even tricks that only involve portals and using the structure of the map to one's advantage, which are even harder to see. These are generally not easy to figure out, but not hard to come up with. With this in mind, one can then build a map centered on these "tricks".
I think this has already been mentioned, but I'll repeat that whatever be the puzzle, it should have a "fair-play" solution - no hidden things, and preferably no hard-to-notice aspects (by this I mean there should be no such things as very hard-to-see portalable surface, etc at a large distances). Of course, good puzzles will have hard-to-notice aspects, but they will still be in plain sight. All relevant portions of the map should be clearly indicated.
One can also (and generally should) combine the above 3 points to increase the difficulty.
Hopefully this will help those who wish to make challenging maps...