Kristala
Kristala is a dark fantasy soulslike with an emphasis on acrobatic traversal, fluid combat, and deep exploration.
Engine: Unreal Engine 5.5
Genre: 3rd Person Soulslike ARPG
Team size: 40
Role: Level Designer
Tools: Blueprints, Git, Miro, ClickUp, Trello, Modeling Tools.
Contributions:
-Designed 2D level layout & player flow
-Blocked out level geometry
-Organized regular playtests & iterated based on feedback
-Scripted combat encounters and quests.
-Requested modular 3D assets from the art team and collaborated with level artists to set dress the level.
Case Study: Tandara's Forest
The problem: Kristala players were quitting early levels after only a few deaths, in particular they would quit just before the level's climactic boss fight.
Approach: I used Alex Brazie's First Principles of Fun framework to identify problems with anticipation and motivation.
​
Solution: I used techniques from Cameron William's 9 Ways to Encourage Open World Engagement, such as Landmarking, Zones of Influence, Content Diffusion Patterns and Dual-Process Theory to keep players engaged.
Result: ​​​​​​
"I didn't want to stop playing."
-Playtester
Landmarking
I built the level around one central POI: a giant, cursed manor house. This setup solved a lot of problems.
1.Players are curious about the manor. They want to know why it has been corrupted and discover what secrets it holds inside.
2. Players can easily anticipate how close they are to reaching the manor. Despite using a tease and denial structure, players remain motivated as they feel the manner is just at their finger tips, even if they are occasionally misdirected.
3. Players can easily orient themselves using the manor, preventing them from becoming lost. Very useful in a game with no map!
Iteration
I started by collecting references and researching examples of medievel manor architecture.
I originally built the manor to a somewhat realistic scale, but it failed to make an impression on playtesters. Interior spaces felt cramped and it wasn't a strong enough landmark for navigational or motivational purposes.
For the manor to make an impact, I decided to go big and doubled its scale.
At this point, we realized we had something special on our hands. I imported the manor into a new level and rapidly iterated on a variety of different layouts, prioritizing ways we could interconnect spaces likes balconies, windows, walkways, and rooves in interesting ways.
The result was something iconic that felt straight of Bloodborne... It anchored the level and players loved it!
The manor easily allows players to understand where they are in relation to it.
Zones of Influence
In order pique players' curiosity, I assigned each POI to a zone of influence. The purpose of this is that each POI's environmental storytelling or narrative drops should be used to develop the intrigue of their master POI.
​
In the chart above, you can see that all POIs before the manor develop the manor's lore. The red POIs develop the lore of the wendigo, the manor's boss. Green POIs start alluding to the next level, the siege of Tandar.
This system ensures that the player always has a hook to the next major piece of content.
​​
Environmental Storytelling & Context Connections
Using the relationships defined by the zones of influence, I tell stories at each of the POIs using narrative drops or environmental storytelling techniques. This builds anticipation for the major POIs and gives them meaning.
​​
The silk farmer was eaten by his spiders because the spider food deliveries stopped arriving.​​
Trade caravans were ambushed because the manor's garrison stopped patrolling the roads.​
Scouts had to camp in the snow because the manor became inhospitable.
Villagers were dragged to the well for a blood sacrifice.
Content Forecasting
The path of bloody footprints to the right lets the player know that path leads to more wendigo related content. The path with log barricades and torches to the left suggests that this road leads towards civilization (and perhaps the city, the player's primary objective).
Dual Process Theory
One technique to keep soulslike players engaged is to have emergent solutions to difficult encounters that players discover through play. This way, if they die, they see a tactical solution that will make their next attempt easier, and therefor they want to try again. It's another take on dual-process theory where players will first attempt an encounter using familiar, automatic play patters, but after failure, they will take a more engaged, analytic approach.
A systemic way to create these encounters is to place attractors and affordances in the encounter space, but to not signify their use before the encounter. Here's an example:
​​
The entrance to this camp doesn't offer any affordances that there might be a better approach.
​​
However, as players progress through the camp, they see indications that there is another way to access this camp.
​​
Players can follow the trail of traversal elements to discover another entrance to the camp. This path allows players to bypass many of the enemies and get an aerial takedown on the camp's elite enemy, making the encounter much more forgiving.​​
Content Diffusion Patterns & Pacing
I divided the level into 30 activity blocks, each with a dominant activity type. Each block encompasses 2.5-4 minutes of gameplay. We expect players to die 3-5 times per session, and each death will set players back 2-4 minutes. This results in 1.5-2.5 hours of total gameplay for this level depending on playstyle.
Level flow chart broken down by activity type
In the level flow chart, notice how each activity type is always followed by a different activity type to keep player engaged. Exception: When the player first enters the manor house, we hit them with 2 back to back combat blocks to spike intensity. Also, when the player first exits the manor, we hit them with 2 back to back exploration blocks to give them some relief as they've just survived the level's most intense moments.
I also follow AAA engagement principles by requiring players to make micro decisions, such as button presses or path selections at least every 15 seconds.
Surprise and Delight
Surprise & Delight is one of the guiding principles of quality hospitality experiences. When I designed a walking tour that earned a 5 star rating on TripAdvisor, our surprise and delight moment was when we walked guests by an excellent chocolate factory and surprised them with free samples. I try to put one of these moments in all of my levels.
There is a secret passage hidden behind this wall. It is accessible at this moment, but no playtesters found it.
Later, when players clear this corruption nodule, the passage's entrance is revealed.
Playtesters would audibly say "Wait, what is that?"
The passageway was inspired by Elden Ring's Siofra River, on a miniature scale. It teases to players that there are magical discoveries waiting to be uncovered all throughout Kristala's world, and it contains a key narrative drop that confirms that the manor's lord has succumbed to the curse of the Wendigo.
Teaching Design Language
I force players to use this log to reach this checkpoint at the start of my level. This teaches them that logs are for traversal.
Observant players may notice this log above them which is a foreshadowed route.
Which players recognize as a means to cross over the gully later on.
Playtesting
I started a regular playtesting program as soon as a roughout was complete where I would do one-on-one playtesting sessions to identify problems quickly and avoid wasting production resources.
One problem I noticed early in playtesting was that certain traversal elements, like the poleswing branches pictured above, blended in too well with the environment, causing players to not notice them.
I created a design language to signify poleswing branches by placing these yellow lights on the ends of the branches. This is an effective signifier because its a contrasting color and a lighting attractor. Players found traversal elements easily after the change.
Production Documentation
The level design team was working in this level long before the rest of the team, so I created some documentation to help orient others to the map and so that non-technical team members could understand what we were working on.
​
I also helped coordinate the production of the modular asset kit that was used to build and decorate the manor by contributing to request documentation, providing feedback, and testing results.
Eliminating Production Risks
The original scope of this map was extremely ambitious. The project's funding runway was highly variable so we needed to plan out multiple outcomes based on available funding. I proposed moving the capital to the halfway point of the existing map. The MVP would be based on reaching the capital. Additional areas could be added as funding allowed.
Result
Playtesters loved the level:
​
"It's a level full of personality in terms of environment and design."
"I really liked the verticality, you could fight the enemies down below or sneak up from above. Reminded me of Batman Arkham [Asylum] or the Spiderman games."
"I loved traversing the manor area, it was creepy and had a lot of corners where enemies could hide. Made you stay on the edge of your seat."
"I love the environmental storytelling."
"The secret cave with the fiery cat statue was a pleasant surprise, it was not expected at all and felt rewarding to find and explore."​
"The part where you climb the tower from the outside was awesome. Reminded me a lot of Assassin's Creed."​
"This looks fucking incredible." -CEO, Astral Clocktower Studios
​