Deliver a state, not a particle
Six sealed states need delivery. Each mission starts with a fresh input qubit and one Bell pair already shared between Alice and Bob. Alice runs the Bell step: a controlled-NOT, a Hadamard on the input, then two measurements. The input state and shared entanglement are consumed; the particles do not vanish.
The two random measurement bits travel by an ordinary classical channel. The Z / Sign bit tells Bob whether to apply Z. The X / Flip bit tells Bob whether to apply X. A 0 means leave that operation off; a 1 means turn it on. The two labels matter, even when their arrival order changes.
| Z bit | X bit | Bob’s correction |
|---|
| 0 | 0 | Neither |
| 0 | 1 | Flip X |
| 1 | 0 | Sign Z |
| 1 | 1 | Flip X, then Sign Z |
The radio is not always reliable
Receive both frames, or try an incomplete instruction. Missing bits show “?”. Flagged bits may be wrong. You have two resend stamps for the whole shift: each buys one reliable copy of a classical measurement record. Waiting for the next frame is free. You can return to the radio before committing a delivery.
The route preview names the three troublesome stops: one erased bit, one bit with a 25% flip chance, and one bit with a 50% flip chance. The stamp budget, fault rates, and story are game rules, not experimental benchmarks or laws of teleportation. Resends copy classical data; they do not clone an unknown qubit.
How you score
Each correct pair of correction choices earns 3 points, for a maximum of 18. A lucky guess can earn those points, but is identified as a guess in the replay. “Fully checked” counts deliveries made with both message bits received and no unresolved flags. A wrong correction gets 0 even if that particular input happens to look unchanged. Two explanation questions have a separate score.
Play together
With two people, one is Sender and one is Receiver; the Receiver also handles the radio. With three, add a Courier to receive and resend the classical records. Use the covered handoff screens. Do not tell the Receiver hidden measurement results outside the game’s channel. Rotate roles after a shift. This uses one shared phone and an honor system.
What the phone can—and cannot—show
Before the message, Bob’s local state averaged over the four possible outcomes is maximally mixed. His local measurements cannot reveal the input or which correction he needs. He can guess, but cannot guarantee transfer of an arbitrary unknown state. Entanglement enables no usable faster-than-light message.
The replay compares exact simulator states, including phase. Its fidelity display is not a test a real receiver could perform on one arbitrary unknown qubit. The six teaching states illustrate the protocol; the core calculation also supports general pure qubits. This is a classical, ideal simulation with fictional classical-channel faults, not a real quantum network.
Protocol: Bennett et al. (1993), Teleporting an unknown quantum state. Correction convention: IBM Quantum’s teleportation lesson. Game design and code prepared for SpookyQubits.