Level Design
The scope of this project was to create a playable section of open world to be taken to conventions and played by prospective customers and investors.
Sinbad is a 3D action platformer with inspiration in games like ‘Super Mario: Sunshine’ and ‘Banjo Kazooie’.
It needed to do the following:
Provide a playground for the various platforming mechanics.
Convey the narrative of the location.
Illustrate a vertical slice of the full game in terms of visuals.
Planning
Initially, I was offered a file containing some geometry and a slew of feedback from early testers. They had found the level:
Too claustrophobic.
The camera was getting stuck on walls and pillars.
Difficult to traverse.
There was little to no sense of flow around the level. Paths lead to dead ends.
Too linear.
The level was longer than it was wide. Players got to the top of the map and felt like they had “finished”, with relatively few deciding to turn back into the level and continue playing.
This also disincentivised exploration and expression.
Doesn’t feel like a real place/culture.
This was owed largely to the disparate assets that had been used.
Felt like it belonged to a different character.
This one required some decoding. I took it to mean that Sinbad’s jumps and dives didn’t map well onto the environment. It was too easy to over/overshoot a jump.
Metrics
As with many of my levels, I began with establishing some key metrics (jump height, jump length, run speed, stopping distance, etc.) and codifying this into geometry. This way I could know, among other things, how wide my streets should be, how tall my buildings should be and how big my level should be - in terms of square footage.
The clip here shows me testing some of these metrics in a small gymnasium.
With this in a workable state, I could begin blocking out the new level.
Blockout
I wanted the finished level to contain varied areas like markets, slums and gardens. I started with building a basic market area to give me an idea for how big a full level might be. I liked how this layout suggests a parkour route around the rooftops, as well as presenting areas to jump across using market stalls.
Slotted into a larger context, the market appears to “open” to seaward visitors. I’m please with how this reinforces the friendly nature of the town’s inhabitants as well as presenting pleasing visual lines to the player (who would spawn facing it).
Not wishing to neglect gameplay on the ground, I added a small amount of verticality and visual interest. The placement of an empty fountain in the centre suggests the market exists on top of a historical square, as opposed to a dedicated market area. I wanted to imply history to the space.
Buildings like this illustrate my approach to this problem. The are low, wide and multi-levelled. The structures are easy to climb and leap between.
The final blockout was a good size for a demo, traversable from end to end in around one minute.
In general, structures needed to invite vertical movement. The two-to-three storey buildings from the initial file would need to either go, or be redesigned to include balconies, mezzanines and vertical pathing (like the stacked crates in Assassin’s Creed).
With more time, I’d like to have broken off the grid a little more. Some better diagonal lines in my buildings’ footprints as well as more incremental verticality would have gone a long way to improving the feel of exploring this level - though I am thrilled with what we achieved.
Overview
Please enjoy this WIP video of me speaking about the level for UK Games Fund.