MasterLagger wrote:
When I map, I try to keep the puzzle simple to figure out, but a bit challenging to actually solve. One of those, "Easier said than done" moments.

Interesting. I take the exact opposite approch when I map. I try to make it a challenging puzzle, but once you know what to do, it's easy. The challenge should be in figuring out a puzzle, not actually doing it. This is what makes portal fun; it isn't about having fast reflexes or perfect aim. It's about thinking, and problem solving. (Some people like rediculous jumping puzzles and things of the nature, where there is little thought involved, but it personally appauls me when I know exactly what to do, but have to try 50 times to do it.)

The puzzle should be challenging. The solution, though challenging to figure out, should be medium difficulty at most to carry out.

A word on dieing. In singleplayer, death-obstacles like crushers and slime pits should be easily avoided, because loading a save is required if the player dies. The player should only die if they do something really, obviously wrong. In coop, however, death isn't as hard to recover from, so it can be applied more liberally. The same is said on the valve developer wiki, and MOST map makers abide by it.

So if a faith plate launched a player in an unexpected direction, then death should not be the consequence of not making an action, unless it is really obvious what the player should do, and they have a few moments to do it. In coop, you could have a faith plate launch you into a grinder, because it takes seconds to respawn, and then you could work out a solution from there.

BTW: There would be an auto-save before this event would happen.

(As for the auto-save, I'd make it trigger just moments after the player gets airborne, so they know that they didn't fuck up by stepping on the faith plate, but still have plenty of time to react the second time, should they have to reload.)

Killerblonde wrote:
Yep, I agree with the above. I do have a question for the mapping community though. How do you think of puzzles? The valve developer wiki has a section on it, but it's not very helpful. Right now I just throw random stuff together, start playing to see how I can solve it, then put I/O's together without a solution in mind, try to solve it, and then reconfigure I/O's to make it possible. It's unefficiant, and usually I don't get that great puzzles. How do you make your puzzles?

You work from either one of two approaches. You write up puzzles in a logical step by step order such as 1. Put cube on button - activates funnel 2. jump in funnel and redirect etc. with a little diagram of locations of the puzzles. OR you just experiment with a variety of different game mechanics and see how a puzzle forms from it. I usually do the latter inadvertantly because my original puzzle either turns out to be not as fun as envisioned or I just prefer my spur-of-the-moment puzzle ideas.

Just think of something that Valve has done, and corrupt it into an idea that has seen little or no usage in the game or other custom maps. For example, in my next map one player has to be in a funnel heading towards several laser fences while the other one uses a laser to turn off each laser field in order. It takes the awesome idea from the single player level with the laser field and laser switchon/switchoff thing and changes it up a bit. Alternatively, come up with a completely revolutionary game mechanic which is often massively appreciated by the community (or fails from impracticality).

.. bleh.

Killerblonde wrote:
Yep, I agree with the above. I do have a question for the mapping community though. How do you think of puzzles? The valve developer wiki has a section on it, but it's not very helpful. Right now I just throw random stuff together, start playing to see how I can solve it, then put I/O's together without a solution in mind, try to solve it, and then reconfigure I/O's to make it possible. It's unefficiant, and usually I don't get that great puzzles. How do you make your puzzles?

I think the most fun way I've recently come up with a puzzle design was for a BTS area. I knew the entrance and exit, and a logical idea of how the area should look (it's BTS, so it should look like it's an area intended for function, not testing). I designed the area, made all of the surfaces portal surfaces, and then just tried anything. I ended up solving it about 3 different ways (not including 'portal at feet, portal at exit wall, done), I then picked my favorite, most creative way, and made the room work like that, removing extra whitespace/tweaking elements until I had my puzzle.

HMW wrote:
I like that idea. If you also make so it that the player would survive if they decided to jump off the panel, then you will have created a moment that will startle most players momentarily, but almost every player will then manage to avoid the death trap.

(As for the auto-save, I'd make it trigger just moments after the player gets airborne, so they know that they didn't fuck up by stepping on the faith plate, but still have plenty of time to react the second time, should they have to reload.)

This particular trap would be used in order to reach the exit. Since the player must fling themselves to reach the exit, the momentum from falling towards the trap area panel would allow the player to fling themselves to the exit. Unless the exit portal was in the wrong place in which case the trap would be reset itself after the player escapes.

Okay, I need to start another discussion. There are ceartain test elements to use in ceartain themes. Lasers are good for destroyed aperture and clean, funnels for clean and wheatly, (although technically wheatly only made 1 chamber,) appropriate boxes are for anything, catapults are for destroyed, clean, and wheatly. EXCEPTION I saw a custom catapult made just for underground style maps. This fit nicely. Gels are for underground, and maybe wheatly or a large clean chamber. Projector bridges are best in destroyed, sometimes clean/wheatly. Does anyone have differant opitions?

tl;dr bump

sp_a4_speed_tb_catch has 4 elements (cube/buttons, funnel, orange paint, and a fizzler that is part of the puzzle), and is my favourite map from the single player campaign, gameplay-wise.

(My favourite ones aesthetics-wise are sp_a1_intro5 and sp_a2_ricochet.)