Level Design


Level design is a lot like training a pet. It may seem like a stretch at first, but good level design will build mechanic tutorials into the world seamlessly. Beyond mechanic tutorials, you are building a perception of your game to your players. Likewise, how you train your pets will determine how they think of you. Before you begin training your pets, you need to keep their attention, this may be easier with some pets than others. You generally either train pets with positive reinforcement or negative punishments after they complete an action.

Keeping player attention is often more challenging than keeping a pets attention. Players will grow bored of tedious repetitive tasks or scenery that doesn’t change. It’s really tempting to just create a few assets and copy them all over a level. This is perfectly fine for prototyping, but your final product will need more variation. There are many tricks you can apply to stretch the use of your objects while keeping down the art workload. In 3D games, stretching, squashing, and rotating objects is very easy to do and will often be enough to make a lot of your assets feel unique. You will see this done with trees and other flora in countless games.

Another very useful technique is to have your artists model parts of objects that can be combined by the level designer to create unique variations. An example here would be modeling different roofs to go with different sets of walls to create unique combinations of houses. One of the most commonly applied technique to reuse assets is recoloring. The original Mario clouds and bushes are a perfect example of this. Applying these techniques when needed will help combat the scenery repetition fatigue your players experience.

Even with varying scenery, it can still be easy to lose the attention of your players by making them feel like they are doing repetitive tasks. There are two parts to this that I consider when designing levels. First, is the introduction of new mechanics. It’s hard to find a better example of this than the game Portal. The player begins with basic movement controls, then learns how to interact with the world, how portals work, how momentum works with portals, and ultimately, how to think with portals. All of the new concepts are introduced in small conceptual chunks. Many games either fail to keep players interested, or overwhelm them to the point that the game is difficult to learn. Finding the perfect balance is easier if you have access to play testers for your game and iterate on your design frequently.

The second part to reducing the repetitive feeling of your game is the flow of a level. I consider flow in level design to be the desired path for the player to follow to complete a level. It’s really easy to fall into a pattern and create a bunch of levels that have the same flow. There are many platformers that do this and all levels are designed to only move to the right. Adding some vertical movement and moving to left in the levels will greatly break up the tedium. In the case of my latest game, the levels are all the size of a single screen. I often started designing my levels by drawing a path that the player will be required to follow to complete the level before filling in any details. This also can help bring notice to excessive backtracking or tedious traveling. By applying all of these techniques, you can greatly improve your odds of keeping a player’s attention.

Now that you have their attention, you need to provide motivation for players to perform different tasks. Exploration can be excellent source of enjoyment for many players. However, many games fail to reward players for exploring. When a player takes the time to climb a hill or check in corners, they will find satisfaction for finding something there. It’s common for games to hide items to promote exploration, but even if your game doesn’t have items, hidden little easter eggs can be just as rewarding. They tell the player’s that you put thought into where they would explore. Now, it may be tempting to throw rewards around everywhere, but this can actually reduce the enjoyment of finding something new as it becomes mundane. Like when training a pet, it’s good practice to provide them rewards frequently early on, and then spread them out later so that they can learn the behavior without expecting an explicit reward for every action. Rewarding players reaches beyond just exploration. Make sure you reward players for completing challenging tasks, such as defeating a large group of enemies or evading some dangerous traps. You still need to be careful about frequent rewards, as abundance will reduce their value.

Positive reinforcement is usually more effective than punishment. You might be tempted to create really cruel levels, that only true masters of your game can complete. There is nothing wrong with that, but be warned that you will have a very small target audience if that’s your main focus. The normal players will feel like you are punishing them for trying to play your game. Providing many checkpoints is one way to reduce the punishment of difficult areas. Many players will quit playing if they need to redo 20 minutes of gameplay to get to a section that they cannot complete. Providing visual cues, such as making background objects point in the direction of the level flow, can help reduce the punishment that players experience. However, sometimes players do need hints that what they are doing is not intended. Telling a player subtlety can often convey the needed information without feeling oppressive.

By grabbing your players attention, rewarding them for their actions, and avoiding excessive punishment, you can create truly captivating levels. Do not expect to create the perfect designs on your first attempt. It often takes many iterations to find a design that works well. With hard work and a little luck, your players will think fondly of your design, just as a well trained pet will have affection for its trainer.