202)
PC, Web
Made for the REPLAY Master's Thesis - Made in Unity
A mind-reading alien crash-lands on Earth in a village full of quirky people, a politician is campaigning for the ongoing election, and there are rising tensions in the village over outsiders. Welcome to 'The Road Most Travelled'.
For a long time, I have been deeply disturbed by the growth of radicalisation and hatred around me. Both in India and abroad, I have seen how people learn to see the 'other' as an enemy, and how political figures prey on our insecurities to drive this hatred for their own ends.
This inspired me to create 'The Road Most Travelled'. I wanted the game to not only paint a satirical picture of this issue but also let players question what causes this hatred and whether it's ever justified.
This vertical slice is the result of five months of research into psychology, sociology, and game design, and eight prototypes built and playtested under Mary Flanagan's Critical Play methodology. The final build was played and reviewed by 35 participants across 14 countries.
You can read the full design process in my thesis here, or scroll down for the gameplay video and a shorter design diary.
Design Diary
Introduction
The Road Most Travelled is a JRPG where you play as an alien who crash-lands on Earth. You set out on a quest to find your spaceship as you navigate the conflicted country of ‘Desh’. Along the way, you run into desi aunties, angry mobs, corrupt politicians and a host of other memorable characters.
Every decision in this game came from one problem: how do you build a video game where understanding someone can resolve hatred, instead of antagonising them? JRPGs already have a structure built for two people taking turns. I just had to prove to myself, and then to players, that this structure could hold a conversation instead of a violent fight.
Prototype 1 - You can 'Listen' to understand the enemy's insecurity and 'Argue' to pacify them
Two Ways to Fail First
My first instinct was to bolt ‘understanding’ onto a normal turn-based battle, with a bullet hell flavour to it (inspired from Undertale) - dodge some obstacles, reach the enemy's brain, unlock a line of dialogue. It played fine. It also meant nothing. Players had no idea why they should bother de-escalating a hostile stranger instead of just walking away. Mechanics without context are just motion.
So I swung hard the other way; a fully written, branching story with almost no system underneath it. This time the political commentary landed, but the interactivity didn't. Players felt like they were clicking through a script, not making choices.
Between these two failures was the actual lesson: the game needed rules that could argue something on their own, wrapped in a story that gave those rules a reason to matter. Neither half could carry the other.
Prototype 2 - A branching choice based story written in Ink to test the socio-political themes
Cutting Until Only the Essential Remains
I've always admired Fumito Ueda's philosophy of ‘Design by Subtraction’, that a game gets better by removing anything that doesn't serve its core idea, not by piling more systems on top of it. I didn't set out to prove this on purpose. I proved it by accident, the hard way.
By Prototype 4, a single encounter tracked four separate bars at once - a health bar and a hostility bar for the player, and the same pair for the enemy. Each one meant something on paper. In practice, players couldn't hold four numbers in their head while also reading dialogue and playing a minigame. The actual emotional core of the conversation drowned under a spreadsheet's worth of stats.
So I started removing things. The two hostility bars collapsed into one shared Tension bar, tracking the whole encounter's temperature instead of two separate meters. Then the enemy's health bar disappeared too, leaving only the player's health and that single shared Tension bar behind. Once the clutter was gone, did players actually start reading what the bars meant, instead of just managing them.
Prototype 3 - Two health bars and two Hostility bars (hidden, shown as colour of the sprites)
Prototype 4 - Two health bars and two explicit Hostility bars
Prototype 5 - Two health bars and one Tension bar
Making Empathy the Harder Choice
Once the format settled - Listen, Question, and Challenge in place of Attack, plus a Check action to dig for an enemy's hidden insecurity, the real work began - making the mechanics themselves prove the game's point, not just decorate it. Every encounter could end one of two ways: Befriend, resolving things peacefully, or Wipeout, forcibly erasing the enemy's entire memory.
For a while, the mechanics proved the opposite of what I wanted. An early build accidentally made the aggressive minigame easier to execute than the compassionate one, which meant the system was quietly telling players that hatred takes less effort than empathy. I rebuilt the difficulty curve so reaching Befriend demanded real, sustained effort, and Wipeout became the lazy way out. It's a small inversion, but it's the one piece of friction the whole game's argument rests on.
Prototype 6 - Listen minigame
Prototype 6 - Question minigame (changed from previous version)
Prototype 6 - Challenge minigame
Turning Backstory Into Leverage
Hatred is not a natural trait; it often comes from insecurities hidden deep inside of us, which are exploited by powerful people. Understanding these insecurities was the key to understanding another person’s actions. In early versions, discovering a character's fear or shame was just flavour text - nice to read, useless to play. I fixed this by tying every insecurity to a specific point on the encounter's Tension bar: a player could only unlock it by patiently talking someone down first, and once unlocked, it gave a real, tangible advantage in the minigames that followed.
That's the moment the system started arguing something on its own. Empathy wasn't just narratively rewarded - it was mechanically rewarded. Understanding someone's fear made peace measurably easier to reach.
Prototype 7 - How to unlock an insecurity through an Action by collecting the appropriate word
Prototype 7 - 'Check' minigame once an insecurity has been unlocked (Note: The Tension must be below the threshold for 'Check' to work)
Prototype 7 - Once an insecurity is found out, it has in-game effects (appropriate words are highlighted)
A Silent Alien in a Satirical Setting
Mechanics needed a story to mean anything, so I built one around a mind-reading alien who crash-lands in a fictionalised Indian village during the election season, gets mistaken for an outsider, and is swept into communal hatred almost immediately.
Keeping the protagonist silent allowed players to project themselves into the role, rather than watch someone else's story unfold. Wrapping the setting in satire - corrupt politicians, radicalised goons, oppressed minions - helped the game deliver a serious message without ever sounding like a lecture.
The prologue introduces players to the Xandian diplomat
The silent protagonist alien goes on an errand and sees the tension in the village
Tricking the Player, On Purpose
The final fight is the part I'm proudest of. By this point, most players had spent the whole game choosing empathy over force. So I designed the final boss to deliberately block that option (to ‘Befriend’). The climactic confrontation with the game's corrupt politician withholds any hint that peaceful resolution is even possible, pushing players toward violence as the only logical move.
Watching genuinely thoughtful, consistently pacifist players default to Wipeout (the ‘violent’ choice in the game) the moment a peaceful resolution seemed to disappear was the clearest proof I could ask for that the game's argument actually worked - how easy it is to choose hatred. Of course, a minority of people realised the trick and could ‘Befriend’ the boss after multiple tries.
The tutorial encourages Wipeout as the better option
The bossfight tricks players into believing there is no way to decrease Tension
What Wipeout Actually Costs
Befriend and Wipeout carry very different weight in the story. Wiping someone out doesn't just end a fight quickly; it erases them, which is no different in method or morality from the manufactured hatred the whole game spends its runtime critiquing. Choosing it means the player has done exactly what every radicalised character in the game does - refusing to understand someone, and erase them instead.
That mirroring only means something if the story actually follows through on it. So the ending branches into three different outcomes depending on the pattern of choices made across all three encounters. Befriend everyone, and the politician driving the game's hatred is forced to genuinely reform. Wipeout everyone, and he's simply replaced by someone worse, while the community he targeted has already fled, proving that removing one hateful person changes nothing if the system behind him stays intact. Wipeout him specifically, but show real understanding to at least one person earlier, and a new leader still rises except this time, the villagers unite to resist him.
Only the version where empathy wins consistently actually breaks the cycle. Every other path just proves how easily it continues.
Ending 1 - Where everyone has been Wiped out
Ending 2 - Where the boss has been Wipedout but rest aren't
Ending 3 - Where everyone has been Befriended
The Vertical Slice
Eight prototypes later, the final build is a short RPG where every encounter runs on a shared Tension bar instead of a health bar, four minigames stand in for four ways of engaging with a person, and the story resolves into one of three endings depending on how much genuine effort a player was willing to put into understanding someone else. It went out to 35 playtesters across 14 countries and the response validated the mechanics. But the truest answer to "did this work?" wasn't in that data. It was in watching people quietly abandon their own values the moment the story made it easy to.
If you want the exhaustive version - every failed prototype, every playtest quote, every study behind it - you may read the full thesis.