quriosity
Sixteen hours of game development where the rules of play are the rules of the universe. Build a game, and let the physics do the teaching.
- When
- 3 to 4 October 2026
- Where
- H204, IIIT Hyderabad
- Teams
- Solo, or up to 4
- Prize pool
- ₹25,000
Kickoff in
Figure 01A qubit, yours to poke
Drag the red dot to set the qubit, then measure it.
Play firstUnderstand later
Nobody falls for quantum mechanics by reading a textbook at two in the morning. People fall for it by poking at it, getting it gloriously wrong, and poking once more. So here is the brief: build a game that teaches one real piece of quantum physics, without a single lecture.
Hello. I am Alice, and this year we are making games where you learn quantum physics completely by accident.
And I am Bob. By accident, yes. Nobody opens a textbook at midnight, but everybody plays one more round.
So the brief is simple. Someone finishes your game, and something in their head quietly clicks.
That little "ohhh, so that is how it works" moment. That is what we are chasing all night.
- 16
hours of building
- 20
hours from kickoff to finals
- 6
options to choose from
- 4
people per team, at most
Seven ground rules
Not legal fine print. Think of these as the physics of the event itself. Bend them and things start to wobble; break them and the whole thing collapses.
- 01
Learn by playing
The whole point, in one line: somebody plays your game and walks away genuinely understanding a quantum idea. No lecture, no quiz at the end. The fun is the lesson.
- 02
Topic first
Please do not lose ten hours to glossy 3D models or clever shaders. A plain square that teaches something true beats a gorgeous game that teaches nothing. Clear and correct wins. Pretty is a bonus.
- 03
Build around the idea
Not an ordinary game with the word quantum printed on the box. The mechanics and the win conditions have to grow out of real phenomena: superposition, collapse, entanglement, gates. Pull the physics out and the game should fall apart.
- 04
Welcome beginners
If a player needs a physics degree, we have already lost them. High school maths and science should be plenty. Keep heavy notation out of sight, unless the gameplay itself explains it the way a patient friend would.
- 05
Learn Struggle Build
Spend the first few hours on a topic your team does not fully understand yet. Notice the exact moment it stops making sense. That moment is gold. Then build the game that clears it up for the next person.
- 06
Original only
No clones, no reskins of quantum games that already exist. Bring your own twist, or better still, a mechanic nobody has dreamt up before.
- 07
Easy to run
Any platform is welcome, as long as the judges can get your game running without a fight. A hosted link or a direct executable is preferred: one click and they are playing. Long setup instructions are where good games go to be forgotten.
Six ways in
Six doors into the same strange house. Pick the one that sparks something and let it shape the entire game. Each is a real piece of physics, and each is stranger than it first appears.
-
Looking at a qubit is never innocent. Ask it the wrong question and it forgets the answer to the right one.
- The principle
- Observing a quantum system means choosing a measurement basis first. A qubit resting in |0⟩ or |1⟩, measured in the computational Z basis, gives the same answer every single time.
- The mechanism
- Measure that very same qubit in the diagonal X basis, the one built from |+⟩ and |−⟩, and the outcome becomes a perfect coin toss. The original information is erased for good, simply because somebody inspected it along the incompatible axis.
- Ideas worth knowing
- Conjugate measurement bases
- Projection operators
- State collapse
- Basis incompatibility
-
A sealed box hides a rule. Classical players have to keep knocking. Quantum players knock exactly once.
- The principle
- A black box function f(x) takes binary strings and answers 0 or 1. It is promised to be either constant, giving the same output everywhere, or balanced, giving 0 for precisely half of the inputs. A classical solver can need 2ⁿ⁻¹ + 1 queries in the worst case to be certain which.
- The mechanism
- Feed the oracle a superposition of every input at once and let phase kickback do its quiet work: U_f|x⟩|−⟩ = (−1)^f(x) |x⟩|−⟩. The many paths interfere, and when the function is balanced they cancel completely. One single query settles the whole question.
- Ideas worth knowing
- Phase kickback
- Walsh–Hadamard transform
- Global parity interference
- Quantum speedup
-
Every move can be undone. That is not a cheat code, it is a law of nature.
- The principle
- Classical gates such as AND and OR throw information away, so there is no recovering what went in. Quantum gates, as long as nobody measures, are unitary: U†U = I. Evolution is strictly reversible, and the state never shrinks or swells on the Bloch sphere.
- The mechanism
- The bit flip X, the phase flip Z and the Hadamard H each rotate the state vector in a precise, predictable way. Any sequence of them can be wound back step by step, or walked through in reverse like a film played backwards.
- Ideas worth knowing
- Bloch sphere coordinates
- Single qubit unitaries
- Reversibility
- State vector trajectories
-
Find the needle in the haystack by making the needle louder, one careful step at a time. Overdo it and the needle goes quiet again.
- The principle
- Classically, finding one marked item among N unsorted ones takes on the order of N checks. A quantum computer can find it in roughly √N steps, without testing the items one after another.
- The mechanism
- Two moves on repeat. First, an oracle flips the sign of the marked state. Then a diffusion operator, 2|s⟩⟨s| − I, reflects every amplitude about the average. The target grows while everything else fades. Keep going past the sweet spot, though, and the rotation overshoots and accuracy slips away.
- Ideas worth knowing
- Oracle phase marking
- Inversion about the mean
- Amplitude reflection
- Rotation angle cycles
-
Send two bits by posting a single qubit. The secret ingredient is a friend you entangled with earlier.
- The principle
- Holevo’s theorem says a lone qubit can carry at most one classical bit once it is measured. Shared entanglement quietly breaks through that ceiling.
- The mechanism
- Two players share a Bell pair, |Φ⁺⟩ = (|00⟩ + |11⟩) / √2. The sender applies one of four local operations, I, X, Z or XZ, to their own qubit alone and sends it across. The receiver decodes one of 00, 01, 10 or 11. Two whole bits, one physical qubit.
- Ideas worth knowing
- The four Bell states
- Local Pauli operations
- Entangled subsystems
- Bell basis measurement
-
Repair a broken message without ever reading it. Sounds impossible, which is exactly the fun.
- The principle
- Noise from the environment causes bit flips (X) and phase flips (Z) on fragile quantum states. Keeping three copies as a backup is off the table, because the no-cloning theorem forbids |ψ⟩ → |ψψψ⟩.
- The mechanism
- Spread the data across an entangled register of several qubits. Helper ancilla qubits interact with that register through CNOT gates to run indirect parity checks. Measuring only the ancillas reveals where an error struck and what kind it was, without ever measuring, projecting or disturbing the protected superposition itself.
- Ideas worth knowing
- Three qubit flip codes
- Ancilla syndrome extraction
- Non-destructive parity checks
- Stabilizer corrections
One long night
Twenty hours from kickoff to the final pitch, sixteen of them spent building. Pace yourselves, drink water, and save a little wonder for the demo.
-
3 October, 10:30
Kickoff
Doors open in H204. The options are walked through, teams settle in, and everybody picks the one that calls to them.
-
Hours 0 to 3
Learn
Research a topic that confuses you. Write down every "wait, what?" moment, because your README will ask about them later.
-
Hours 3 to 16
Struggle and Build
Turn that confusion into a mechanic. Prototype, break it, playtest it on whoever is still awake, repeat.
-
Hour 16
Submit
The sprint closes. Hand in all five deliverables through the submission form.
-
Before 06:00, 4 October
Finals
Finalists present their games to a panel of domain experts, and the night finally gets to end.
What to hand in
Five things, all through one form. Leave any of them out and the judges get sad, and sad judges are not a winning strategy.
- 01
Your option
Name the option your whole game is built around.
- 02
A playable game
A hosted link or a direct executable, whichever runs most easily.
- 03
A public repository
All of the source code, out in the open.
- 04
A short video
Your team playing the game, with the core mechanic clearly in action.
- 05
A README write-up
What your team learned about the topic in the first three hours, and how that idea became the beating heart of the game.
The submission form lives on its own page, ready when you are.
Open the submission formThe spoils
A prize pool of ₹25,000, split three ways. The glory, as always, is shared by anyone who makes a quantum idea finally click for somebody else.
First place
₹14,000
Second place
₹7,000
Third place
₹4,000
What the guidelines reward
- The physics is accurate and easy to follow
- The core mechanic comes straight from the quantum idea
- A beginner can pick it up and understand it
- The idea is genuinely your own
- It is easy to get running, hosted or as a direct executable
Fair questions
Something we have not covered? Ask us on Discord, find any ISAQC volunteer in H204, or write to us. We are friendlier than a decoherent qubit, we promise.
-
Not at all. The guidelines actually encourage picking a topic your team does not fully understand yet. Confusion is the raw material here.
-
Come alone, or bring up to three friends. Four people is the limit. Mixed teams of programmers, designers and physics enthusiasts tend to do rather well.
-
Whatever you are fastest in. Unity, Godot, Unreal and plain web code are all fair game, and Qiskit or Cirq are welcome for the quantum logic if you want a real simulator underneath.
-
Any of them. Browser, desktop and mobile games are all welcome, as long as they run without a struggle. A hosted link or a direct executable is preferred, so the judges can start playing straight away.
-
No. Pick one option and wrap the entire game around it. Depth beats breadth tonight.
-
Please do not. Clones and direct copies are out. Borrow inspiration if you must, but the central mechanic has to be your own.
-
In two stages. Everything submitted after the sixteen hour sprint is reviewed, and the finalists then present to a panel of domain experts. Every submission must show a meaningful link between its gameplay and its chosen option.