Quote from Idolon on August 9, 2011, 11:46 am

These are some basic level design guidelines I wrote up for Portal 2. I'm not saying you should live by these, but these should give you a general idea of ways to keep the player happy in your level. As a mapper, you are putting on a performance for the player: your main goal is to keep them happy and entertained.

Keep in mind while reading this that I'm not trying to control how everyone makes maps. I tried to write these guidelines as objectively as possible, but I guarantee some of my opinion is in this article. Take this advice and use it however you want to, make your maps however you want to make your maps, and just do what you want to! Mapping is a wilde world out there, just waiting for us all to explore.

~Idolon

1: Difficulty

Difficulty is one of the more touchy subjects in level design: You want to challenge players, but not frustrate them. Your puzzles should be guiding, but not obvious. The best way to do this is to leave as many walls white as possible. If your puzzle is still too hard, start laying down some metal to guide the player to certain conclusions.

Also, you need to watch out for "artificial difficulty." Artificial difficulty is the result of something being harder than it should because one single, annoying aspect, when a simple rearrangement or modification would make the puzzle/challenge much more enjoyable.

For example: let's say you have a fling maneuver. At the end of the fling, there is a nest of turrets lying in wait for an ambush. An example of artificial difficulty in this case would be hiding the turrets so they are only visible after you land right in front of them. To eliminate the artificial difficulty here, you could put the turrets behind glass walls, or even better, make the turrets able to be killed (lasers, blue gel, etc.) before the player flings over.

Now, this is not to say you should eliminate any surprises! Surprises keep the player on their feet. However, the player needs time to analyze the situation and determine the best plan of action. For example, turret ambushes work when the turrets are revealed one by one, or revealed when the player is in a safe zone. It may seem like a bit of a tone-down in your map, but when done right, you can get a more enjoyable, yet exciting sequence.

2: Expectations

One very, very good idea is to test from the perspective of the player. Better yet, get a friend to test it for you! This helps you assess expectations.

Expectations are what the player thinks will happen in your level. You need to remember expectations when designing your map. Expectations, however can be dangerous. Remember that players only know what information you tell them. They can't foretell what will happen later in the level unless you let them see it in action. Players won't expect a platform to save them from a fall when they fling into a toxic goo pit. As a result, if you expect them to do so, they will probably become very stuck. Players never expect the unexpected, because, well, they're unexpected.

Another aspect of expectations is teaching concepts. If you are introducing a new concept or puzzle element, it is best to explain by example. Teaching works best by telling players why reactions happen as opposed to just that they happen.

To clarify, let's say you're introducing the Aperture Science Companion Cube Retriever 2000ZX. Instead of telling them "If you use the ASCCR2000ZX on that box you will solve the puzzle," tell them "If you use the ASCCR2000ZX on that box it will cause it to move over somewhere else, which will solve the puzzle." Make sense? Teach cause and effect, not just cause or effect.

3: Guidance

Guide the player through the levels. This is just another example of difficulty, but this needs more exploring. I stress this heavily: The player should always have all of the information in front of them. A player will probably become frustrated when they can't solve a puzzle, and they realize the solution was one little switch hidden behind a wall that falls right into place. This is just another example of artificial difficulty.

Now, I'm not going to name names here, but there is one particular level that doesn't use this tool effectively. Instead of having one puzzle will all of its tools that leads to the exit, it spreads all of the tools to open the exit all over a giant maze. The main problem this causes is that the player doesn't know when they have all of the tools needed to solve the level, causing them to try and solve the puzzle when they can't. This is essentially hiding the switch behind the wall: Hiding information to make the puzzle more difficult.

(I know some of you liked this level, but I thought it was a good example of not using this guideline. I'm not trying to put this level down, it was good in its own respects! Everyone is entitled to their opinion. Instead of getting mad, make a level that uses this effectively! That's what experiments are for: testing new ideas.)

Guidance is also a double-edged sword, however. Don't baby the player into the solutions, just nudge them in the right direction. Give them an idea of the solution, but don't give it away. This is a very delicate balance, and once we went over in the difficulty section.