Wednesday, September 26, 2012

Portal 2 - Ways to design a level

Portal 2 is a rather simple game with pretty basic mechanics but the concept of having two portals to play with via portal gun opens itself to limitless numbers of puzzles. With Portal 2, Valve later released a level editor for the community to share. The level editor itself is very easy to use and with basic knowledge of Portal 2 of even just experimenting with the level allows you to create all sorts of puzzles. For Game Design our goal was to create a Portal 2 level with our group with some requirements.

I’m not going to talk about what I did to fulfill requirements for the assignment, instead I'm going to go through what I learned in designing a level. Essentially what I should do and what I should not do for a puzzle. These dos and don’ts are up for debate but I feel these are things that really help in level design.

My designed level (With testing and additional content from teammates)


What to do!
Make it obvious what a button does

When triggering a button, you need to make sure that you see it affects the environment around you. This means if you press a button, a cube should drop down for you to use right in front of you or near you. It shouldn’t be all the way across the map where you can’t see or hear it.

Laser gate on

The reason we do this is because that way it won’t confuse the player needlessly. Have a button be in viewing distance of what the result is going to do, or at least keep it close enough so that players will see the result without having to search too much. Having an obvious audio queue can also work but isn’t as effective. We need to provide the player some feedback on their actions otherwise it’s needlessly harder. If we want to make a puzzle harder, it should be because they need to think about how to solve a puzzle, not more difficult because they need to find out what your button did and where it dropped the cube.

Laser gate off. The button is in plain site of the laser field so it's easy to tell what the button did.

In my level, I’ve positioned all my buttons in such a way as if you look close to them, you will see the result immediately or soon enough. For example, a button will drop the cube right next to you, or will raise a bridge right in front of you.

Lead a path for the player

In other words, don’t make a giant room that doesn’t look like it has any particular order or requires a lot of back tracking. The reason we would want to lead the player is because we want them to get to our puzzles, this isn’t an adventure game. Now there is a limit as to how much we should lead the player, don’t point out everything for them, but point enough out for them to get somewhat of an idea of where to go. Don’t have three paths and in no sequential order scattered about.

The purpose for this game is for players to solve puzzles, not wander around like an adventure game. Though designers could attempt making a level like this, the main draw to playing portal is puzzles. Any moving around without doing any actual puzzle solving could be considered boring. Walking around to get to the end of the hall just isn’t fun.

The goal is in sight, all you need to do is figure out how to disable the threats in your way

For my level I lit up the path (This was actually a requirement for our levels) but I also made it very linear, you can actually see your final destination right away. I made it obvious where you need to be, you just have to figure out how to get past the obstacles in your way. Some key areas are also highlighted by lights though you still have to figure out how to solve the puzzle once you get there.

Allow some mistakes

The player is going to make mistakes and we shouldn’t always punish them too terribly with them. One bad thing to do is make it so if players pass a certain point, but forget something they become permanently stuck and have to restart the level. This is especially terrible if there isn’t even much of a clue that continuing without a key object will end up this. I made a jump in one level to test and found out I became stuck at the bottom because I couldn’t portal to anywhere else.

For my level, it is possible to return all the way to the beginning if you dropped a cube

One debatable thing is the use of precision mistakes. For example I played a level where I had to use a reflection cube to aim a laser to a laser catcher to cross a bridge. If at any time the laser slipped off, the bridge would fall under me. Now this is difficult to cross and it’s a neat idea but for some players it can be frustrating because they know exactly what to do but one small slip up leads to restarting a level. This can lead to even more frustration if this is right at the end of a level meaning you have to do everything all over again because of that. If I were to use this in a level, I would just force them to restart the bridge, not restart the level. Either that or have a checkpoint right at that spot.

On the topic of checkpoints, one thing to do for a level is that if you plan to have a very dangerous area where players will die, try to keep it closer to the beginning rather than the end. This way if they make the mistake of dying, they will be able to retry as soon as possible rather than having to go through the same beginning puzzles over and over. Eventually that just becomes a chore because you already know how to solve everything at the beginning and have to keep doing it over and over.

Precision mistakes are minimal as the player is given the whole length of the level to try aiming this laser

How I used this in my level, you ride across a tractor beam throughout the level, spotting obstacles in the distance. You can also see these obstacles before riding the tractor beam too. I ensured that players could return to the beginning of the level by raising a bridge if they forgot anything important. To solve my puzzle, I also give you ample time and alternative ways to get to the finish so as to avoid players blaming precision mistakes. At the end of the level, you have to destroy some turrets with a laser beam and using a reflection cube. The farther back you start on the tractor beam the more time you have to target the turrets in your path. There is also a fairly wide window you have to destroy the turrets even if you start as far forward as possible.

Don’t have “Troll” buttons

Try not to have buttons that will screw over the player, especially into death. Don’t have a button that immediately kills a player for pressing it, unless it’s obvious that’s what the result will be. Don’t have buttons that randomly cause effects like that or put you at the beginning of a level unless you make it clear pressing it will have that effect. It will only serve to frustrate players and make the experience worse.

These two button (left) in my level used to be unhelpful. This has been rectified.

For my own level, I did in fact have a “troll” button, where activating it would not really help you. It wouldn’t kill you or set you back, but it wasn’t helpful either. Players in general will want to press the buttons and have them do an effect of some sort that will help. My buttons simply disabled the laser that would allow you to kill the turrets blocking you from the end of the level. Now I have it so pressing the buttons will instead activate the laser so that you can proceed. I made the button actually give some benefit for the player.

Get rid of anything useless

This one should be fairly obvious and it’s to make sure not to have any areas that look like they mean something, but they don’t actually do anything. So don’t have a bunch of buttons in an area but pressing them actually does nothing. It will only confuse the player even more, similar to the problem I mentioned earlier of having buttons who’s effects are obvious. Don’t have corridors that lead to nowhere either, they should be either a means to complete a puzzle or a way to get to another puzzle. This is similar to the troll button issue except these serve to waste time rather than frustrate the player.

The button behind this cube is given use to flip the turret in the distance. It may be a one time use, but it always benefits the player.

For my level, I got rid of any extra areas that weren’t needed. I had a small corridor that had a reflection cube but I got rid of the cube because I had already provided a reflection cube at the beginning of the level. Hence I made sure to get rid of the corridor later because it wouldn’t make the puzzle any better. I also made sure every one of my buttons did something meaningful. This is also part of the reason I changed the buttons that disabled the laser (noted in the trolling section), because they didn’t actually help you at all. You could get by without even touching them, which confuses a player even more.

Difficulty vs. Accesibility
Level beaten!

Everything I have said is relating to making everything easier to understand for players. It is my belief that it’s a very good practice. There is a reason that Portal & Portal 2 did so well in their campaigns despite them being brief. It’s because they were all very accessible, pretty much anyone could go through the game and enjoy it. They pretty much adhered to all the “What to do”s that I mentioned. Everything was clear and straight forward, the difficulty was in figuring out the puzzle, with no other issues in the way.

Think about it, if you suddenly had useless buttons, buttons that changed things in the level far away, way out of your sight then that’s just going to confuse and frustrate you right? That’s not confusion by logic, it’s just confusion to frustrate, it does not benefit the player experience in anyway. Sure it can promote observation but the key to making levels that will appeal to players is to make it accessible and it’s easily possible to keep observation too.

That’s the point of good level design to draw in players, to make them want to play even if it’s difficult. You want to make players like your levels and play through the game. If word gets out your game is broken, punishes players and is confusing in its level design, people don’t want to play. So making accessible, but difficult levels is really the best way to go. It’s simply good level design.

No comments:

Post a Comment