Bell's Inequality
When a coin lands heads, the result was always going to happen; we just couldn't see all the details. Is quantum chance like that, or is it truly random? Remarkably, you can settle this with pure logic and a little counting. Four short steps: try to copy a pair of twins with a plan, see why no plan can work, meet the real twins, then see the one way out and why physicists refuse it.
Try to copy the twins
Two twins sit in separate rooms with no way to talk. A judge asks each one a question, do you like cats, dogs or birds?, and they answer yes or no. Tested over and over, real twins always show two patterns:
The obvious explanation: before being split up, they agreed a plan. Make a plan and see if it copies them.
Rule 1 is easy: both twins follow the same plan, so the same question always gets the same answer. Now Rule 2. The judge can ask different questions three ways. Tap one to see it:
Different answers. No match this time.
Why no plan can ever work
You tried one plan. Here is every plan there is. The red zone above is below 1 in 3, and not one of them gets there.
| Cats | Dogs | Birds | C&D | C&B | D&B | Agree |
|---|---|---|---|---|---|---|
| Y | N | N | ✗ | ✗ | ✓ | 1 in 3 |
| Y | Y | N | ✓ | ✗ | ✗ | 1 in 3 |
| Y | N | Y | ✗ | ✓ | ✗ | 1 in 3 |
| N | Y | Y | ✗ | ✗ | ✓ | 1 in 3 |
| N | Y | N | ✗ | ✓ | ✗ | 1 in 3 |
| N | N | Y | ✓ | ✗ | ✗ | 1 in 3 |
| Y | Y | Y | ✓ | ✓ | ✓ | 3 in 3 |
| N | N | N | ✓ | ✓ | ✓ | 3 in 3 |
Your plan is highlighted. C&D means one twin is asked about cats and the other about dogs.
- No making it up on the spot. If they did, the same question would sometimes get different answers, and Rule 1 breaks. So every answer has to be decided in advance, and that leaves just these eight plans.
- Something always repeats. Three questions but only two answers, so every plan gives two questions the same answer. That is the ✓ in every row.
- The judge catches it 1 time in 3. So every plan agrees at least 1 time in 3, never 1 in 4. Swapping plans between rounds doesn't help: it only averages rows that are all 1 in 3 or more.
Particles really do this
The twins are a stand-in for a real experiment:
What this kills, and the one way out
The table rested on one assumption we never questioned: that the judge picks the questions freely, without the twins knowing in advance.
Local determinism
The common-sense picture: each particle carries its answers from the start, and nothing travels faster than light. That is exactly the twins with a plan, and step 2 shows no plan can reach 1 in 4. With free questions, this picture is finished.
Superdeterminism
If the twins knew the questions in advance, faking 1 in 4 would be easy. Superdeterminism says they effectively do: the particles, the random number generator, the starlight and the physicist's own choice were all set up together at the Big Bang, in exactly the way that makes 1 in 4. The questions were never free.
Why physicists won't take the deal
It's a conspiracy
Every coin flip, computer random number, human choice and billion-year-old ray of starlight would have to be secretly coordinated with particles it never met, for no purpose except making quantum physics look random. Nothing rules that out. It is just an explanation that could explain anything.
It would break all of science
Every experiment assumes the tester's choice is independent of what is being tested. A drug trial trusts that the coin deciding who gets the pill isn't secretly linked to who would recover anyway. Allow that link here and you must allow it everywhere, and then no experiment can prove anything.
A few serious physicists do defend superdeterminism, but it is a small minority view, and no superdeterministic theory yet predicts the 1 in 4 in advance.
So quantum is random. Truly random.
A coin toss only looks random: its result is fixed by how it was flipped, and we just can't follow the details. Particles are different: no hidden plan can explain them.
The answer does not exist until the particle is measured, and at that moment nature decides it by pure chance. The only alternatives are influences faster than light, which clash with relativity, or a universe-wide conspiracy that would undermine every experiment ever run. Most physicists take the randomness instead.
This way of telling it follows David Mermin's 1981 version of Bell's argument. The proof is pure counting; the 1 in 4 comes from quantum theory and the experiments.
Try another puzzle →