What superposition actually is
Superposition means a qubit's state is two amplitudes, one for the answer 0 and one for the answer 1. An amplitude is a number that can be positive or negative (or even complex). It does not mean the qubit is "both 0 and 1 at once." When you measure, you square each amplitude to get a chance, so an equal superposition reads 0 on about half of shots.
A qubit's state is two numbers. One is the amplitude for the answer 0. The other is the amplitude for the answer 1. An amplitude is a number that says how strongly the qubit leans toward an answer. We write the state as α|0⟩ + β|1⟩. Here α (alpha) and β (beta) are the two amplitudes.
Amplitudes can in general be complex numbers, but every circuit on this page keeps them as plain real numbers. They follow one rule: |α|² + |β|² = 1. That is because squaring an amplitude's size gives the chance of that answer. All the chances must add up to 1.
That is the whole definition. "Superposition" just means both amplitudes are not zero. After an H gate the state is (|0⟩ + |1⟩)/√2. Each amplitude is about 0.7071. Square it: 0.7071 × 0.7071 ≈ 0.5. So each answer has a chance of 0.5.
So is the qubit 0 and 1 at the same time?
No. That slogan goes wrong in two clear ways.
First, it suggests you could read out both values. You cannot. Measuring the qubit gives you exactly one bit. After that, the state matches the bit you saw. The next lesson is all about this.
Second, "both at once" says nothing about the signs of the amplitudes. In the previous lesson, the signs did all the interesting work. They canceled an answer completely.
Here is a better picture. Think of the state as an arrow pointing in a definite direction, like a compass needle. It is not a dice roll. The direction fixes the odds of every measurement you could make. Gates turn the arrow in a set, predictable way. Chance only comes in at the very end, when you read the qubit. Unlike a compass needle, though, you can never look at the arrow itself. You only see the 0 or 1 it gives you.
What does equal superposition look like at 128 shots?
Does more data make it exactly 50/50?
How close should the counts be?
Think of flipping a fair coin 10 times. Getting 6 heads would not surprise you. Flip it 10,000 times and you will land very close to half. Shots work the same way. The math for this is called binomial statistics.
When you estimate a chance p from n shots, the typical error is √(p(1−p)/n). That is exactly the ± number the panel shows. Let's do the arithmetic for p = 0.5. Then p(1−p) = 0.5 × 0.5 = 0.25.
- 128 shots: √(0.25/128) ≈ 0.044, so about ±4.4 percentage points.
- 1,024 shots: √(0.25/1,024) ≈ 0.016, so about ±1.6 points.
- 16,384 shots: √(0.25/16,384) ≈ 0.004, so about ±0.4 points.
To cut the error in half, you need four times as many shots.
This has two honest results. First, you will never get exactly 50/50, and you should not expect to. Getting exactly 8,192 zeros and 8,192 ones would itself be a little surprising. Second, no number of shots shows you an amplitude directly. You only ever estimate chances from samples, and only to a limited precision. The bar chart is a record of samples. It is not a picture of the state.
What does this look like on real hardware?
On a real device, superposition is a fragile physical setup. It is not just a note in a notebook. Two clocks limit how long it lasts. T1 is how long it takes an excited qubit to relax back toward |0⟩. T2 is how long the qubit keeps its phase, meaning the sign and angle information in its amplitudes. Think of a spinning top that slowly falls over (T1) and wobbles out of step (T2).
Today's superconducting chips usually last tens to hundreds of microseconds. A microsecond is a millionth of a second. Trapped-ion qubits can last seconds or longer. They hold superpositions the longest of any commercial type.
A circuit has to finish well before those clocks run out. That is why gate speed and these times are always given together. Both appear on each device's spec page. The comparison view shows how each type of machine trades speed against lasting time.