Monday, 14 May 2012

Portal 2 Perpetual Testing Initiative (and Level Design Guide)


Over the last few days, I have been creating levels for the PC version of Portal 2 using the new game editor that was released last week as part of an update.

See my efforts for yourselves - Skelitor77's Steam Workshop Page

My First Level Attempt (How the Chicken Crossed the Road)
- A fairly easy puzzle which got me used to the editors possibilities.

The new editor is very easy to use and works on a click-and-drag basis for most areas but by right-clicking an object, you can edit its properties and connect it to buttons or objects. And although I have experience with other editors such as the UDK, it would be pretty straight-forward for anybody that is new to level editing to pick up.

With over 30,000+ levels being created and 1.2mil custom level downloads since its release, it is also proving to be very popular amongst Portal 2 fans. There have been a few issues with publishing and playing levels during peak time, but this is something Valve are looking into as it appears that this has proved more popular than they predicted. Hopefully they will also looking at ways to improve the exposure of new levels and good but low-rated levels, but it is still early days.
 

Lab Escape - Keeping the player focused on reaching the central room.

Having played some of the levels currently available, while most are fun and enjoyable, there a few issues and problems people seem to have when it comes to creating the levels. Obviously it is difficult trying to work out what other people are going to do when they enter your level, and they are likely to come across solutions and bypasses that you never intended. I’ve done a quick list though of the things you should be doing to create a better experience.
 
Here's my guide on creating great levels:
  • Test your Levels – You’d be surprised at the amount of levels that people will play once decide it works then publish it. And you not only have to play it a number of times, but test it to its limits. Look for puzzles you can bypass, try to approach the level from a different perspective. But best of all, get your friends and other people to play it. You’ll be surprised at what you can miss. 

Puzzle Machine - This required a lot of testing...

  • Don’t make it completely Obvious – As a Puzzle game, a lot of the enjoyment you get from Portal 2 will be from solving the puzzle and then executing the solution. The amount of levels where there are only a handful of places where you can put a portal is ridiculous, and so many good concepts are ruined by the lack of thought to this area. Nobody wants to play a puzzle game on-rail. 

The Flooded Room - Although some Levels can just work first time.

  • Let the Experiences Flow – All great games have a flow. One experience, leading to another, to another, and then on to the next. The player doesn’t want to be trying to find out what they’re supposed to do, especially if they enter a giant room with hardly anything in it. Everything should be compact, whilst still having the room it needs to work. Flow will keep a player in the game, whilst lapses give the player the chance to leave, and they can do that at the end of the level.

Puzzle Machine - A Big experience that needed the player to know where to go next.

  • Make Good use of Light – If a player walks into a room, more likely than not they are going to look towards the better lit area, Its Human-nature. Don’t ignore dark areas though, these add atmosphere, just don’t expect people to find the solution in the dark. I have play-tested my own levels with minimal lighting but then I know how to solve it, after all I built it. But trying to spot the difference between a white wall and a grey wall when they are both near black is just about impossible.
  • 
    Lab Escape - The Big Observation window is great for shadow effects
    
  • Don’t Force Players to Restart – This all comes back to Testing and Flow, but if a player has to manually restart a level, then the level has failed. If you don’t want the player to get into a certain area, or want to have a trap the player can fall into. You need to allow them to escape, or at least kill them. If you don’t, you are basically giving the player the opportunity to quit.
    Broken Lifts - The bright lights of the lifts gets the players to look upwards and their attention.
    
  • Dying for the Hundredth Time isn’t Fun – An Player Error executing a manoeuvre can be funny, just don’t expect the player to be able to always be able to perfectly execute a jump/manoeuvre. Challenging manoeuvres add to a puzzles difficulty. But when they have to perfectly execute it to avoid falling to your death; they are going to die a lot, no matter how good they are. If a player keeps dying they are eventually going to quit, and that will not reflect well if they end up rating the level too.
Shooting Gallery - It's a tough challenge trying to face up to a number of Turrets.
  • Bad Trolling – The point of designing levels is not to beat the player, but to pose a challenge for the player to overcome. Much like the Evil Genius from Spy Thrillers and Action films that keeps the Hero at arm’s length until he overcomes the odds to achieve victory. The player should always wins in the end, as they are the hero. You can scare the player with impending death, only to save them at the last minute, but don’t kill the player because you can.

  • Finally, make it Look Appealing – Imagine a big grey room, four walls a floor and a roof, with a few objects in the middle of it. Boring isn’t it. Once you’ve got the basics in place, decorate it. Use the contrast between White and Grey walls, have undulating walls, place light features, get some glass and iron grating, raise columns and beams. Having something pretty to look at is never a chore, but flat, square, monotone rooms are. Games are about an experience, and the look of it is included in that, make the player think about it after they’ve stopped playing.

Puzzle Machine - All white walls here are for effect.

If you’re looking for ideas, the best thing to do is to play around with the tools on offer. But you can also play other peoples levels to look for good ideas. The link to my levels is at the top of this post, and the screenshots are all from my levels. Always try to leave feedback, good or bad, as with feedback and critique is how we improve as designers. You can only go so far by yourself. My levels are certainly not perfect, with myself and others spotting things I had overlooked. Some things I had even considered before publishing, but either decided I had done enough to combat it, or that the solution I had put in place was not effective enough.

Shooting Gallery - Could you get though that gap in the glass? Maybe you should try the hard version.
I hope this helps any wannabe level designers and puzzle-makers, and a lot of these rules apply to many other genres and games. After all, games are about using the mechanics you are given to work past the obstacles, and achieve victory.