POLARIZED
A First-person Puzzler
Summary
Polarized is a first-person puzzle level inspired by Portal 2 about solving physics-based puzzles using magnetic forces. Stranded on an abandoned oil rig-like laboratory on a foreign water covered planet, the player must find their way to the escape shuttle.
Goals
-
Develop and prototype puzzle mechanics
-
Design good puzzles and make the player feel clever
-
Teach the player new mechanics through out the level
-
Follow a balanced difficulty curve
Project Breakdown
-
Solo 8 weeks halftime (4hours/day) (4 weeks scripting + 4 weeks designing the level)
-
Created using Unreal Engine 5 and Blender
-
All scripts created by me
-
All 3D models by me
-
Marketplace assets/plugins used
Overview
1
2
3
4
5
6
1. Onboarding
2. The Reveal
3. Polarity Fields
4. Claw Upgrade
5. The Other Rig
6. The Drones
Commentary Walkthrough
Developing Fun Mechanics
The best puzzles are built on simple mechanics rather than complex ones. The challenge should come from the clever ways these mechanics are utilized, not from their inherent complexity. By designing different mechanics that all interact with one another, the potential for unique and engaging puzzles increases exponentially.
If it wasn’t obvious, this level is heavily inspired by Valve’s Portal. It’s a puzzle game about positioning and movement. What makes Portal’s puzzles so great is that every challenge revolves around a single central mechanic: Portals. It’s a unique yet simple physics-based concept, abstract but easy to grasp.
Being a huge physics enthusiast, I’ve always been fascinated by magnetic fields and forces. Magnets are an everyday item, meaning players will have an intrinsic understanding of how they behave. The core properties of attraction and repulsion serve as the foundation for every puzzle. The gameplay revolves around positioning and moving magnetic objects as well as the player themselves.
Magnetic Cube
The Magnetic Cube is the core puzzle element. It is physics-based, can be picked up, placed down, and used to activate weighted buttons. The cube can be negatively (blue) or positively (red) charged. Unlike real magnets, these cubes are monopoles, meaning they have a single polarity. Two cubes with the same polarity repel each other, while two cubes with opposite polarities attract.
Magnitic Plate
A static object with a magnetic charge that can be placed on floors, walls, or ceilings. The Magnetic Plate can be connected to an electric object, becoming a Generator Plate. Generator Plates are activated when an oppositely charged object is connected to them.
Polarity Field
Switches the polarity of any object passing through it.
EradicationField
Destroys any magnetically charged object passing through it.
Polarity Claw
The Polarity Claw is a handheld electromagnetic tool that, when charged with a polarity, can:
-
Attract and grab movable opposite poles from afar
-
Repel and launch grabbed identical poles at high speeds
-
Attract and pull the player towards static opposite poles from afar
-
Repel and launch the player away from static identical poles
-
Additionally, when passing through a Polarity Field, the Polarity Claw's charge switches poles.
Drone
A floating enemy AI that keeps a safe distance while attacking the player from afar.
Drones are dipole objects, meaning they have both a positive and negative pole. While airborne, a drone’s primary pole faces the player. However, when grabbed, the drone flips, revealing its secondary pole on the backside.This means that drones can be attracted from a distance and launched from the player's hands. Otherwise, they behave like Magnetic Cubes, including the ability to activate Magnetic Plates. When passing through Polarity Fields, their poles are reversed.
Designing Good Puzzles
RULES
These are rules I have developed and followed during this project.
1. Solvability Over Difficulty
It's better to have a puzzle that's too easy than one that's too hard. There should always be tension before solving the puzzle, but never frustration. A puzzle that is too difficult to solve is a bad puzzle.
3. Understanding the puzzle
The player must understand the problem before solving it. A puzzle solved by accident leaves the player confused and unsatisfied, having learned nothing new. Always design for "aha!" moments, not confusion.
5. Mechanically easy to execute
The player should never have to execute a pixel-perfect jump or spend 10 minutes trying to solve the puzzle the wrong way. Buffer zones must be used to clearly communicate which actions are possible and which are not. Once the player has figured out the solution, it should be easy to execute.
7 Every puzzle must be unique
If the core solution to a puzzle is identical to a previous puzzle without any variation, it should not exist. Each puzzle should present a new, unique challenge, and the player should leave each one having learned something new.
2. Strive for simplicity
To avoid unintended red herrings, all components of a puzzle must play an essential role. Remove all unnecessary elements, refining the puzzle through subtraction until it reaches its purest form. If a puzzle component can be used in multiple ways, it should be.
4. All information on the table
Every piece of the puzzle must be clearly visible to the player. A player who discovers a hidden off screen component after prolonged struggle will feel cheated, not triumphant.
6. Fail-safe design
The player should never be able to get themselves or an essential object stuck in a position where they cannot escape. The player should never have to kill themself to reset a puzzle or destroy a puzzle component as part of the solution.
8 Iterative playtesting
The gap between designer intent and player perception is where puzzles break. Every player approaches puzzles in different ways. By observing real behaviors and adapting the design, the puzzle can be tailored to work for as many different players as possible.
Teaching Complex Mechanics
Onboarding
Learning by doing is the best way to teach complex mechanics. No first-time player wants to face a wall of text when launching a game. Learning should be diegetic, fun and engaging, and it is the designer’s job to make players excited to learn more.
The first puzzles let players test mechanics like picking up cubes, placing them on buttons, and experiencing magnetic interactions and eradication fields. Each mechanic is introduced separately in a safe environment. These rooms are not meant to be challenging and require minimal problem-solving.
Ki-Sho-Ten-Ketsu
I use the Ki-Sho-Ten-Ketsu structure for teaching mechanics, which essentially means Introduction, Exploration, Twist, and Resolution.
By introducing one mechanic at a time and allowing players to explore their interactions, they develop a strong understanding of the game’s rules. Once the fundamentals are established, I introduce twists, such as the ability to switch a magnet’s polarity, allowing the player to explore once more before introducing yet another twist. Suddenly, the player can use the Polarity Claw to attract and repel magnetic objects. Although the previously explored mechanics function the same way, the player can now interact with them in completely new ways, shifting their perception of puzzles. The resolution comes when the player demonstrates mastery of every puzzle mechanic in the final challenge
The fourth room is the first real test of problem-solving. Here, players must use repulsion to solve a puzzle. The solution relies on intrinsic knowledge of magnetism, but players can also brute force it by wandering until repulsion occurs. Either way, they learn how repulsion works.
Design Problems and Solutions
The Fan Room
The first version of the fan room is a perfect example of breaking my own fail-safe design rule. The intended solution was to realize that polarity fields also change the polarity of the Polarity Claw and use this to launch the cube over the field from inside the chamber, then retrieve it from the outside. However, playtesters immediately brought the cube out through the polarity field and tried to throw it in from above. This sometimes got the cube stuck on the balcony with no way to retrieve it.
I suspect this behavior was influenced by the previous polarity field puzzle, where players had to carry the cube out, launch it over, and retrieve through the polarity field. I had not foreseen this, and it needed to be fixed.
Removing the Balcony?
I considered removing the balcony entirely, but this eliminated a key part of the puzzle, launching yourself to grab the cube mid-air. While not the core of the puzzle, it made it unique. Removing the balcony would have broken my rule of keeping puzzles distinct and made the puzzle too easy.
Adding an Eradication Field!
The best solution was to prevent players from throwing the cube in from outside. Placing an eradication field on the balcony that only deactivated once the cube was on it solved the issue perfectly.
The Final Puzzle
The final puzzle was the last one designed in the level. It needed to meet two key requirements: Utilize every major mechanic and be outdoors with the rocket as a backdrop
The first version broke nearly every design rule I had. It was overly complex, had four movable magnetic objects, was mechanically clunky even when the solution was known, and had multiple ways to soft-lock the player or force a reset. It was an absolute mess even before any playtests. All it had going for it was that it was incredibly unique.
In the second version, I made a key discovery: drones cannot activate weighted buttons. This mechanic was incorporated into the puzzle, but all the other issues remained. With the deadline approaching, I had to move on. Technically, it worked, but I knew it was bad.
Eventually, I decided to fix it. I identified the puzzle’s core elements: using drones to release cubes, using cubes to unlock the polarity field and switch drone polarity, placing drones in cages, and using cubes to activate buttons. With two cubes and two drones, this process had to be repeated twice, making it unnecessarily long.I reduced the steps by removing one drone and one cube, getting closer to the core of the puzzle. This required rearranging many elements, but the result was a design that was less complicated, mechanically smoother, less punishing, and encouraged more creative use of the cube and drone.
Reflection
Designing puzzles is hard. Designing them under time pressure with a deadline approaching is waaaaay harder. During the development I’ve spent 20% of my time building puzzles, 20% experimenting with ideas, and the last 60% just staring at the screen, feeling time slip away.
Despite this, puzzle design might be my absolute favorite part of level design. I love challenging myself, and the payoff of creating a great puzzle is a satisfaction like no other. That feeling when inspiration suddenly strikes and you come up with an amazing idea, then seeing the player think, behave, and solve the puzzle exactly as intended.
If I had to redo the project, I might consider cutting the drones. They are incredibly cool and fitting, which is why I wanted to showcase them, and I don’t want to cut them from the current level. But even without them, I already have a set of solid mechanics that could be explored much deeper. Even before getting the polarity claw charge, the potential for unique puzzles is enormous. I could have sacrificed some of the level’s twists to allow for more exploration, but that would have made it much more boring. I am glad with the decisions I made and very proud of the end product.
Thank you for reading!!!