A general look at how the design focus shifts when transitioning from linear/multi-path level design patterns to open world design.
What is 360 approach and how do you use it correctly to get the most out of your sandbox level?
In order to understand what we are dealing with, we need to go deep into the concept of actual linearity in video game levels.
Video game levels generally fall under 3 categories:
You can read about each of them bellow:
- The player travels from point A to point B using a predefined path.
- Since the representation is molecular we can assume A and B can be anything. For the sake of simplicity they represent the start and end of a linear progression.
- Between A and B the players will find obstacles (marked with the red cross). As the design develops we can turn such obstacles into actual encounter spots, that will have spaces dedicated to them.
- We can go further and assign thematics to each space, hinting at some sort of progression.
- It’s always a good idea to try to use a node structure to illustrate how locations tie to each other. It’s rather important to keep this in mind since it makes your design easier to read. It gets much harder in later stages when things start to get complicated.
If we were to take the molecular structure and turn into in a level sketch we might get something like this:
Red areas represent game-play spaces, green areas represent quiet time spaces, for quiet reflection while moving on to the next game-play space.
As a side note when designing a level one thing that comes up a lot is the idea of having an escalation curve running trough your level.
It may look something like this:
This means that the difficulty graph goes up and then at some point reaches a peak and the goes down and then the level stops.
However this would mean that if we did exactly that, the player would end up having a heart attack before reaching the end of the level.
In reality it might looks like this:
The peaks represent the loud time (where the challenge is the highest) while the valleys represent where the quiet time is supposed to be:
You can find out about the importance of quiet time from Super Bunny Hops video on this exact subject: https://www.youtube.com/watch?v=rCxR__N0_is:
So in-spite of the actual linearity of the molecular graph the intensity and pacing of the level could be very different.
Why this is important to remember:
- Because you will forget (unless you are Mitja Roskaric, Mitja never forgets)
- Because you can easily get sidetrack and start over-scoping or underscoring your level
- Because it’s useful when you are trying to show case things in a easy to read, non convoluted format to shareholders that might pay your checks. Not everybody understands what you want to do with your level, in fact you might be the only person that knows that.
A multi-path level implies that the level is designed to facilitate multiple routes trough an environment, paths that tie objectives/important areas together.
This raises a bunch of issues:
- The player has multiple ways of reaching an area/objective
- How do you make sure the player doesn’t get lost or confused?
- How do you make sure that the progression chart stays the same regardless of what the path the player takes to reach the objective?
- How do I make sure the player does what I want him to do?
One of the solution is to simplify the problem:
We can assign different locations and the game-play associated with them to a specif beat.
If we look at how this beat structure looks like we can immediately notice how what the next step could possible be.
This way we can ensure that:
- The player progression stays the same on all paths
- Each path has it’s own distinct flavor
- We know where the player “might” be at any given moment
- We allow the player to choose what path he wants to take
- We make sure each area is distinct enough from the others so the player won’t get lost
It does however mean that as a level designer you need to give up some control and let the player decide how he wants to play your level. to a certain extent.
But in a nutshell it comes down to designing multiple linear levels that are stacked on top of each other, intersect each other and give more game-play options to the player.
If the open world is empty, but we scatter a bunch of spawn points around what we get is a lack of directionality. The player can go anywhere but there isn’t really anywhere “worthwhile” to go.
So regardless of how many spawn points you have and how big the map is , from a molecular stand point, everything can be simplified like this:
If we introduce an answer to that question mark and allow the player to notice it, it becomes a point of interest:
In simplified form it will look like this
In the absence of any kind of direction from the designer the player will simply explore what he finds interesting. If the orange circle is the most interesting bit in the landscape that’s where he will go.
If you concentrate your game-play beat distribution in and around this area that’s how the player is going to experience that content.
This happens because every point of interest inside a map (and this is valid in level design in general) exerts a magnetic pull on the player. The player is going to be drawn to that POI and will eventually get there.
If you introduce other locations in the map things tend to get a little more complicated:
POI’s tend to compete with each other and they sort of fight for the players attention.
The players will weight in a bunch of factors before visiting a POI:
- How far is it?
- Does it look more interesting then the other ones?
- What do I know about?
- Have I already been there?
We can tweak the magnetism of a location by tweaking all these variables and making the locations more interesting and distinct from each other
This leads to a bunch of varied approaches to playing this map:
These approaches are all based on player preferred play styles and typologies.
- Some players might like to go for the hard locations first
- Some players might like to explore the other side locations
All the POI’s need to be compelling however, since player will only visit them and engage with the game-play beats if they see a point in the effort to get there.
If you throw more variables into the mix the equation becomes even more complicated and harder to predict.
This is when distribution and simplification needs to be heavily reinforced:
In this example it’s extremely hard to predict where the player is going to go, so it becomes very difficult to ensure that a progression curve might work in this scenario at all.
So what do we do?
We apply the simplification method that we talked about earlier.
We split the world in regions and try to layer the player progression using an onion structure to ensure that they, in the best case scenario, will travel the world from bubble to bubble experiencing the content based on what our progression curve might actually be.
So… this leads us to the very core of the discussion. Open world Levels
Levels in open world scenarios.
We have already established that the player can reach these POIs from multiple angles based on the complexities of the world POI network.
So let’s look at how a POI should be composed.
From a molecular stand point the diagram should look like this:
We could try to apply the multi-path solution to this and it would not be a bad way of approaching this situation:
Another way to illustrative this is using the diagram bellow
- Approach angles from open world — infinite
- Approach angles from Layer 1 to Layer 2–4 entry points
- Approach angles from Layer 2 to Objective layer — 2 entry points
- Location size
- Number of Objective layers
- Number of Layer 2 areas
- Location difficulty
Arguments for trying to keep it simple:
- They have so many components in them that are very hard to keep track of
- Feature creep will happen.
- The intent is that if you manage to identify any good preferred routes trough the location then the focus should to polish them. If they are too many you won’t have enough time to cover all of them.
- Locations that are too complex will end up confusing the player.
I will cover the subject of challenge distribution and how to make things easier or harder in each layer in a different article.
Also stay tuned to see this applied in practice rather then just theory.