🏠 Home 📘 Track 1: Quantum Basics L13 — Creating Entanglement L14 — Measurement Bases L15 — Three Superpowers Together
L14 §4 · Build Something Real ~18 min

Measurement Bases — You Can Measure in Any Direction

Every measurement so far asked one question: is the qubit 0 or 1? That is only one of infinitely many questions you could ask. Change the direction you measure in — and the same qubit gives you completely different, equally valid information. You can only ask one question per qubit. Choose carefully.

✦ One Idea A measurement basis is a choice of question. The same qubit gives completely different information depending on which direction you measure it in — and measuring in one direction permanently destroys your ability to learn what a different direction would have told you.
measurement basis Z basis X basis Y basis mutually unbiased BB84 cryptography
Section 01
① Hook

You've Only Been Asking One Question

🧭
Test your intuition first
Think carefully — this reveals something surprising about quantum measurement.

A qubit is in the state |+⟩ = (|0⟩ + |1⟩)/√2 — a perfect equal superposition. You measure it in the standard Z basis (asking "is it 0 or 1?"). What do you get?

Every measurement in this track so far has done exactly the same thing: take a qubit, look at it, and ask — is it 0 or 1? The qubit collapses to one of those two answers. Done.

But here is something we have not told you yet. That is not the only question you can ask.

You can ask a completely different question — and the qubit gives you a completely different answer. Not a wrong answer. Not a worse answer. A genuinely different piece of information, extracted by a genuinely different kind of measurement.

🤔
Stop and think about what this means
The state of a qubit contains more information than a single 0-or-1 question can reveal. If you measure in the standard way, you learn something — but you permanently destroy the quantum state, and with it, everything you could have learned by asking a different question. There is no way to ask both questions on the same qubit. This is not a limitation of our instruments. It is a fundamental feature of quantum mechanics — built into the geometry of the state space itself.

This idea — that a quantum state can be measured in different ways, each revealing different information, and that you can only choose one — is called measurement bases. It is one of the most practically important ideas in the whole subject, and it is the foundation of quantum cryptography.

⚡ The consequence — stated plainly
A qubit prepared in a definite, known state can give completely random measurement results — not because the qubit is uncertain, but because you asked the wrong question. The randomness is in the mismatch between the question and the state. This is quantum mechanics' most counterintuitive gift.
Section 02
② Intuition

The Compass Analogy

Before any quantum notation, here is the everyday version of choosing a measurement direction.

🧭 Which direction is "forward"?
You are standing in an open field holding a compass needle. The needle is pointing somewhere — let's say northeast.

Someone asks: "Is the needle pointing north or south?" You look along the north-south axis and answer. It is closer to north — you say north. That is your measurement in the north-south basis.

Someone else asks: "Is the needle pointing east or west?" You look along the east-west axis. It is closer to east — you say east. That is your measurement in the east-west basis.

Same needle. Different question. Different answer. Both are correct — they describe the needle from different angles.

The "north-south axis" is the Z basis in quantum computing — the standard 0-or-1 question. The "east-west axis" is the X basis — a completely different question about the same state. You can only pick one axis to measure along. Choosing one permanently destroys all information about the other.

The crucial difference from a real compass: after you measure a qubit in one basis, the state collapses into that basis. You cannot then measure the same qubit in a different basis — it is gone. You had one chance, one question. What you chose determined what you learned, and everything else was permanently lost.

✦
The single most important fact about measurement bases
Choosing a measurement basis is choosing which question to ask. The qubit can only answer one question. And the answer depends entirely on what you asked — not just on the state of the qubit.

The "wrong basis" effect — the most counterintuitive result

Here is the part that feels almost unfair. Suppose someone prepares a qubit in state $|{+}\rangle$ — a perfectly definite, known state in the X basis. If you measure in the X basis, you get the answer "plus" every single time. No randomness. Completely predictable.

But measure the same state in the Z basis — and you get 0 or 1 with 50% probability each. It looks like random noise. The qubit has not changed. The randomness came entirely from asking the wrong question.

State |+⟩ measured in X basis
🎯
Perfectly predictable
You always get |+⟩. 100% certainty. The basis matches the state — you learn everything with a single shot. No wasted information.
State |+⟩ measured in Z basis
🎲
Completely random
You get 0 or 1 with 50% each. Completely unpredictable — not because the state is uncertain, but because you asked the wrong question.
⚠️
The randomness was not in the state
When you measure $|{+}\rangle$ in the Z basis and get random results, the randomness did not come from the state being uncertain. The state was perfectly definite — it was $|{+}\rangle$, a precise, known quantum state. The randomness came from measuring in the wrong basis. There is nothing you could have done, with any instrument, to extract a definite Z-basis answer from that state. It is not a technology limitation. It is a fact about the geometry of quantum state space.
Section 03
③ Framework

The Three Standard Bases

On the Bloch sphere, every direction through the centre is a valid measurement axis — there are infinitely many. But three come up again and again in quantum computing and cryptography, corresponding to the X, Y, and Z axes of the sphere.

↕
Z Basis
|0⟩ and |1⟩
The standard computational basis. Measures "up or down" on the Bloch sphere. Every lesson so far has used this — asking whether the qubit is 0 or 1. The Bloch sphere's north pole is |0⟩, south pole is |1⟩.
↔
X Basis
|+⟩ and |−⟩
Measures "left or right" across the equator. Asks whether the qubit is in state |+⟩ or |−⟩ — both of which are superpositions of 0 and 1. This is also called the Hadamard basis, since H switches between Z and X.
⊙
Y Basis
|i⟩ and |−i⟩
Measures "into and out of" the Bloch sphere. Involves complex amplitudes. Less common in introductory circuits, but essential in quantum tomography and certain error-correcting codes.

Each basis has exactly two outcomes — because every quantum measurement on a qubit is binary, always collapsing to one of two states. What changes between bases is which two states are being chosen between.

The X basis states written out — you have seen them before

The states $|{+}\rangle$ and $|{-}\rangle$ that form the X basis are not new. They are the states the Hadamard gate creates from $|0\rangle$ and $|1\rangle$:

$$|{+}\rangle = \frac{1}{\sqrt{2}}\bigl(|0\rangle + |1\rangle\bigr) \qquad |{-}\rangle = \frac{1}{\sqrt{2}}\bigl(|0\rangle - |1\rangle\bigr)$$

Both are equal superpositions of $|0\rangle$ and $|1\rangle$ — they have identical probabilities in the Z basis. They look completely identical to a Z-basis measurement. But they are distinct, orthogonal states: the X-basis measurement distinguishes them perfectly. The Z-basis measurement cannot tell them apart at all.

💡
Why |+⟩ and |−⟩ are invisible to Z-basis measurements
Measure $|{+}\rangle$ in Z: you get 0 or 1 with 50% each. Measure $|{-}\rangle$ in Z: also 0 or 1 with 50% each. The two states look completely identical in Z — you learn nothing about which one you had. But measure them in X: $|{+}\rangle$ gives you "plus" with 100% certainty, $|{-}\rangle$ gives you "minus" with 100% certainty. Completely distinguishable in X. Information completely hidden in one basis can be fully exposed in another. This is not a curiosity — it is the entire foundation of quantum cryptography.

Mutually unbiased bases — why Z and X are special together

The Z and X bases are called mutually unbiased: any eigenstate of one basis has equal probability across all outcomes of the other. If you know the state is perfectly defined in Z (say, definitely $|0\rangle$), then measuring in X gives you completely random results. And vice versa.

This mutual randomness is not a coincidence. It is a geometric property of how these two axes are oriented on the Bloch sphere — perpendicular to each other. Two bases that are maximally incompatible in this way are exactly what quantum cryptography needs.

Section 04
④ Theory

The Maths of Basis Change — and Why It Breaks Classical Cryptography

Measuring in a different basis means re-expressing your state as a combination of the basis states you are measuring in — then applying the Born rule to those components. Here are the three most important examples.

Z basis measurement on |+⟩ — certain state, random result
$$|{+}\rangle = \frac{1}{\sqrt{2}}|0\rangle + \frac{1}{\sqrt{2}}|1\rangle$$ $$P(\text{get }0) = \left|\frac{1}{\sqrt{2}}\right|^2 = \frac{1}{2} \qquad P(\text{get }1) = \left|\frac{1}{\sqrt{2}}\right|^2 = \frac{1}{2}$$
The state $|{+}\rangle$ expressed in the Z basis has equal amplitudes on both $|0\rangle$ and $|1\rangle$. Born rule: square the amplitudes. Both outcomes are 50%. Complete randomness — from a perfectly definite, known quantum state. The randomness is in the basis mismatch, not the state.
X basis measurement on |+⟩ — aligned basis, certain result
$$|{+}\rangle = 1 \cdot |{+}\rangle + 0 \cdot |{-}\rangle$$ $$P(\text{get }|{+}\rangle) = |1|^2 = 1 \qquad P(\text{get }|{-}\rangle) = |0|^2 = 0$$
$|{+}\rangle$ expressed in the X basis has all its amplitude on $|{+}\rangle$ and none on $|{-}\rangle$. Born rule gives 100% certainty. The state is an eigenstate of the X-basis measurement — it "belongs" to this basis. The answer is always "plus." No randomness at all.
X basis measurement on |0⟩ — roles reversed
$$|0\rangle = \frac{1}{\sqrt{2}}|{+}\rangle + \frac{1}{\sqrt{2}}|{-}\rangle$$ $$P(\text{get }|{+}\rangle) = \frac{1}{2} \qquad P(\text{get }|{-}\rangle) = \frac{1}{2}$$
Now the roles reverse. $|0\rangle$ is perfectly definite in Z — it is an eigenstate, certain result. But in X, it is an equal superposition of $|{+}\rangle$ and $|{-}\rangle$. Measuring in X gives completely random results. Certainty in one basis, complete randomness in the other. This is the meaning of mutually unbiased.

BB84 — how measurement bases make eavesdropping physically impossible

Measurement bases are not just a theoretical curiosity. They are the foundation of the most important quantum cryptography protocol ever invented. The key insight is brutal and elegant.

🔐 BB84 — the core mechanism
In 1984, Charles Bennett and Gilles Brassard published a protocol that uses measurement bases to distribute cryptographic keys with physical security guarantees — not computational ones.

Alice sends Bob qubits, each randomly chosen from {|0⟩, |1⟩, |+⟩, |−⟩}. Bob measures each one — but doesn't know Alice's basis, so he randomly picks Z or X. Half the time he guesses right, half wrong. After all qubits are sent, they publicly compare which bases they used (not the results). They keep only the matching-basis measurements. Those bits form the secret key.

Now here is why it is unbreakable. If Eve intercepts a qubit and measures it, she doesn't know Alice's basis either. If she guesses wrong, she gets a random result — and must resend something to Bob. But she doesn't know the original state, so she resends the wrong thing. When Alice and Bob compare a sample of their key bits, they will find errors introduced by Eve. No errors = no eavesdropper. Errors = someone was listening.

Eavesdropping leaves a physical trace — guaranteed by the laws of quantum measurement. Classical encryption is hard to break. Quantum key distribution is physically impossible to break without detection.
1
Alice prepares qubits in random states from two bases
Each qubit is one of {|0⟩, |1⟩} (Z basis) or {|+⟩, |−⟩} (X basis), chosen randomly. She keeps her choices secret.
2
Bob measures each qubit in a randomly chosen basis
Bob doesn't know Alice's basis. He picks Z or X randomly for each qubit. When he guesses right, his result matches Alice's. When wrong, he gets a random result unrelated to Alice's bit.
3
Basis reconciliation — publicly announce which bases were used
Alice and Bob publicly compare their basis choices (not results). They discard all measurements where they used different bases — roughly 50%. The remaining measurements are perfectly correlated and form the raw key.
4
Error checking — detect any eavesdropping
They compare a small sample of their key bits. If Eve intercepted qubits, she introduced errors (by measuring in the wrong basis and resending). Errors above the noise threshold mean the channel is compromised — abort. No detectable errors = provably secure key.
🔮
Measurement bases appear everywhere in quantum computing
Quantum teleportation: choosing which Bell basis to measure in determines what correction Bob must apply. Quantum error correction: measuring in different bases reveals different types of errors without collapsing the logical qubit. The Hadamard gate: switching between Z and X bases is literally what H does. Grover's and Shor's algorithms both exploit the fact that certain states look simple in one basis but encode structure that interference can extract in another.
Section 05
⑤ Interactive

Measurement Basis Explorer

Drag the state arrow on the 3D Bloch sphere to any direction — theta (up/down) and phi (around) — then choose a measurement basis and measure. See exactly how the same state gives certain or random results depending on which axis you ask about.

🧭 3D Bloch Sphere — Measurement Basis Explorer
Left-drag: rotate view · Right-drag / Shift+drag: move state vector · choose basis · measure
INTERACTIVE
LEFT-DRAG: ROTATE  ·  RIGHT-DRAG: MOVE STATE
Choose measurement basis
Outcome probabilities (live)
Outcome |0⟩
50%
P = 0.500
Outcome |1⟩
50%
P = 0.500
Collapse the state
Random state challenge
What just happened
Right-drag (or Shift+drag) to point the state vector anywhere on the sphere. Left-drag to rotate the view. Then choose a basis and measure.
Measurement history
🔬 Four experiments to run in 3D
1. Click |0⟩ ↑ preset. Switch to Z basis — probability cards show 100%/0%. Measure in Z repeatedly: always get |0⟩. Now click X basis — cards flip to 50%/50%. Measure in X: completely random. Same state vector, different axis, completely different answer.

2. Click |+⟩ → preset. The state vector points along the X axis. Switch to X basis — probability shows 100%/0%. Now switch to Z — shows 50%/50%. Rotate the sphere by left-dragging to see why: the state vector is perpendicular to the Z axis, so Z measurement is maximally uncertain.

3. Click |i⟩ ⊙ preset. This is the Y-axis state — impossible to see in a 2D cross-section. Switch to Y basis — probability shows 100%/0%. Switch to X or Z — completely random. Left-drag to orbit the sphere and observe the state vector pointing directly out of what used to be the screen plane.

4. BB84 in 3D. Right-drag (or Shift+drag) to set the state vector to a random direction. Measure in Z — note the result (Alice's bit). Reset. Set the same direction again. Measure in X instead (Eve measuring in the wrong basis): random result. Reset and use Y: random again. Eve cannot know Alice's basis, so every wrong-basis measurement introduces detectable errors.

5. Random State Challenge. Hit 🎲 Random State — Measure All Three Bases. A random qubit lands on the sphere; three identically-prepared copies are measured in Z, X and Y simultaneously. Watch the probability column: one basis will show a high-confidence result, the others will be close to 50/50. That high-confidence basis is the one most "aligned" with the state. Hit it repeatedly — every random state tells a different story about which basis sees it most clearly.
Quick Check
Lesson Summary

What You Now Know About Measurement Bases

  • 🧭
    A measurement basis is a choice of question
    The Z basis asks "0 or 1?" The X basis asks "|+⟩ or |−⟩?" Any direction through the Bloch sphere is a valid basis. Different questions reveal different information — and you can only ask one per qubit.
  • 🎯
    Certainty in one basis means randomness in a complementary basis
    A state definite in Z gives completely random results in X, and vice versa. This mutual unpredictability is not a flaw — it is a deep consequence of the geometry of quantum states, and it is what makes quantum cryptography possible.
  • 📐
    Measuring in a new basis means re-expressing the state — then Born rule
    To find probabilities in a given basis, write the state as a superposition of the basis states and square the amplitudes. The probabilities depend on the overlap between your state and the basis — not on some absolute "randomness" of the qubit.
  • 🔐
    BB84 uses measurement bases to make eavesdropping physically detectable
    An eavesdropper who intercepts qubits must guess the basis — wrong guesses introduce errors that Alice and Bob can detect. The security comes not from computational difficulty but from the fundamental laws of quantum measurement. Classical encryption can theoretically be broken. BB84 cannot be broken without leaving evidence.
  • ⚛️
    Measurement bases appear throughout all of quantum computing
    Teleportation, error correction, Grover's, Shor's — the ability to choose which question to ask, and the consequences of that choice, run through all of quantum information science. The Hadamard gate is literally a basis-change operation. You will use this concept again and again.
How clearly do measurement bases click?

Superposition. Interference. Entanglement.
You've built a Bell pair. You understand measurement bases.
Now — how do these three superpowers work together
in a real quantum computation?

→ Three Superpowers Together — L15
Sources & Further Reading
← Previous
Creating Entanglement
L13 — H + CNOT = Bell pair
← → navigate lessons · Space = mark complete