Introduction

In 1935, Einstein, Podolsky and Rosen pointed at a strange feature of quantum mechanics. Measure one of two entangled particles and the other one instantly seems to "know" the result, however far apart they are. Einstein found this absurd — "spooky action at a distance" — and argued the particles must have carried a hidden script all along, fixed at the source. This is the idea of local hidden variables: the world is local (no influence travels faster than light) and real (outcomes are decided in advance).

For thirty years this looked like pure philosophy, impossible to test. Then in 1964 the physicist John Bell did something extraordinary: he turned the question into arithmetic. He showed that if the world were locally real, a certain sum of measurement correlations could never exceed a fixed number. Quantum mechanics predicted it could.

One side had to be wrong, and now an experiment could decide which.

Try It: Break the Bound

This demo runs a simplified CHSH experiment. Two detectors, far apart, each randomly choose one of two angles and measure an entangled pair. We tally how often the two outcomes agree, combine four such correlations into a single score S, and watch where it lands.

<p class="hint">Two far-apart detectors each pick one of two angles and measure an entangled pair. We combine four correlations into one score <b>S</b>. Any <b>local, pre-decided</b> universe obeys <b>S &le; 2</b>. Quantum mechanics reaches <b>2&radic;2 &asymp; 2.83</b>. Pick a model and run the trials.</p>
<div class="controls">
  <label><input type="radio" name="model" value="quantum" checked> Quantum (entangled)</label>
  <label><input type="radio" name="model" value="classical"> Classical (hidden script)</label>
</div>
<div class="meter">
  <div class="scale">
    <div class="zone-ok" style="width:70.7%"></div>
    <div class="bound" style="left:70.7%"><span>S = 2</span></div>
    <div class="tsirelson" style="left:100%"><span>2&radic;2</span></div>
    <div id="bar" class="bar" style="width:0%"></div>
  </div>
</div>
<div class="status" id="status">Choose a model and press Run.</div>
<div class="btns">
  <button id="run" type="button">Run 20000 pairs</button>
  <button id="reset" type="button" class="ghost">Clear</button>
</div>

* { box-sizing: border-box; }
body { font-family: system-ui, sans-serif; color: #222; margin: 0; }
.hint { font-size: .9rem; color: #444; margin: 0 0 .8rem; line-height: 1.45; }
.controls { display: flex; gap: 1.2rem; flex-wrap: wrap; margin: .3rem 0 1rem; font-size: .92rem; }
.controls label { cursor: pointer; }
.meter { margin: 1rem 0 .6rem; }
.scale { position: relative; height: 34px; background: #eef2f6; border: 1px solid #cdd9e3;
         border-radius: 8px; overflow: visible; }
.zone-ok { position: absolute; top: 0; left: 0; height: 100%; background: #dceede;
           border-right: 2px dashed #0a7d33; border-radius: 8px 0 0 8px; }
.bar { position: absolute; top: 0; left: 0; height: 100%; background: #1d3557;
       opacity: .82; border-radius: 8px; transition: width .5s ease; }
.bound, .tsirelson { position: absolute; top: -20px; transform: translateX(-50%);
                     font-size: .72rem; font-weight: 700; color: #0a7d33; white-space: nowrap; }
.tsirelson { color: #c92f3c; }
.status { font-size: 1rem; font-weight: 600; margin: 1.1rem 0 .6rem; min-height: 1.4em; line-height: 1.4; }
.status.ok { color: #0a7d33; }
.status.bad { color: #c92f3c; }
.btns { display: flex; gap: .5rem; flex-wrap: wrap; }
button { font: 600 14px system-ui, sans-serif; padding: .45rem .9rem; border: 1px solid #1d3557;
         background: #1d3557; color: #fff; border-radius: 8px; cursor: pointer; }
button.ghost { background: #fff; color: #1d3557; }

const fmt = (s, v) => s.replace(/%(\w+)%/g, (_, k) => String(v[k] ?? ''));
// CHSH Bell test. Two detectors choose a setting (a in {0, 45 deg}, b in
// {22.5, 67.5 deg}). The "score" S = E(a0,b0) - E(a0,b1) + E(a1,b0) + E(a1,b1),
//  * Local hidden-variable theories are PROVEN bounded by S <= 2 (Bell 1964).
//  * Quantum entangled pairs reach S = 2*sqrt(2) ~ 2.828 (Tsirelson bound).
// We simulate both and watch the bar cross 2 only for the quantum model.
const DEG = Math.PI / 180;
const A = [0 * DEG, 45 * DEG];          // Alice's two settings
const B = [22.5 * DEG, 67.5 * DEG];     // Bob's two settings
const SCALE = 100 / (2 * Math.sqrt(2)); // map S onto the 0..2sqrt2 bar (%)

// --- Quantum model: outcomes +/-1 with correlation E = cos(2*(a-b)).
// For a singlet-like pair, P(agree) = cos^2(a-b). Sample a shared random "phase" lambda and apply the quantum rule honestly per trial.
function quantumOutcome(a, b) {
  // Use the singlet correlation directly: agree with prob cos^2(a-b).
  const pAgree = Math.cos(a - b) ** 2;
  const agree = Math.random() < pAgree;
  const alice = Math.random() < 0.5 ? 1 : -1;
  const bob = agree ? alice : -alice;
  return [alice, bob];
}

// --- Classical model: a local hidden variable. Each particle carries a
// pre-set answer fixed at the source, independent of the far detector. Best deterministic strategy still cannot beat S = 2.
function classicalOutcome(a, b) {
  // Shared hidden variable: a random source angle. Each side answers sign(cos(2*(setting - lambda))) -- a fixed local rule, no communication.
  const lambda = classicalOutcome._lambda;
  const alice = Math.cos(2 * (a - lambda)) >= 0 ? 1 : -1;
  const bob = Math.cos(2 * (b - lambda)) >= 0 ? 1 : -1;
  return [alice, bob];
}

function correlation(model, a, b, n) {
  let sum = 0;
  for (let i = 0; i < n; i++) {
    if (model === 'classical') classicalOutcome._lambda = Math.random() * Math.PI;
    const [x, y] = model === 'quantum' ? quantumOutcome(a, b) : classicalOutcome(a, b);
    sum += x * y;
  }
  return sum / n;
}

function computeS(model, n) {
  const e00 = correlation(model, A[0], B[0], n);
  const e01 = correlation(model, A[0], B[1], n);
  const e10 = correlation(model, A[1], B[0], n);
  const e11 = correlation(model, A[1], B[1], n);
  return Math.abs(e00 - e01 + e10 + e11);
}

function selectedModel() {
  return document.querySelector('input[name="model"]:checked').value;
}
function setStatus(t, cls) {
  const s = document.getElementById('status'); s.textContent = t; s.className = 'status ' + cls;
}
function setBar(S) {
  document.getElementById('bar').style.width = Math.min(100, S * SCALE) + '%';
}

function run() {
  const model = selectedModel();
  const N = 5000; // per correlation -> 20000 pairs total
  const S = computeS(model, N);
  setBar(S);
  if (S > 2.01) {
    setStatus(fmt('S = %S%  ->  beats 2! No local hidden-variable theory can do this.', { S: S.toFixed(3) }), 'ok');
  } else {
    setStatus(fmt('S = %S%  ->  stuck at or below 2, exactly as local realism predicts.', { S: S.toFixed(3) }), 'bad');
  }
  console.log(model, 'CHSH S =', S.toFixed(4));
}

document.getElementById('run').addEventListener('click', run);
document.getElementById('reset').addEventListener('click', () => {
  setBar(0); setStatus('Choose a model and press Run.', '');
});
document.querySelectorAll('input[name="model"]').forEach(r =>
  r.addEventListener('change', () => { setBar(0); setStatus('Press Run to measure.', ''); }));
Result
Console

Any local hidden-variable universe is bound by S ≤ 2 — no clever pre-arranged script can do better. Run the simulator with the quantum correlations and S climbs to about 2.83 (the value 2√2, Tsirelson's bound). That extra 0.83 is the gap Einstein's worldview cannot cover. Switch to the classical model and you will see S stay pinned at or below 2, no matter how you tune it.

The Real Result

So what exactly did Bell establish, and how firm is it?

  • It is a proven theorem. In 1964 John Bell showed mathematically that any theory that is both local and built on pre-determined values must satisfy the inequality. The CHSH version (Clauser, Horne, Shimony, Holt, 1969) writes it cleanly as S ≤ 2.
  • Quantum mechanics violates it. The same theory predicts S = 2√2 ≈ 2.83 for ideal entangled pairs — a clean, parameter-free contradiction. The gap is not a rounding error; it is a wall between two pictures of reality.
  • Experiment took the quantum side. Starting with John Clauser (1972), sharpened by Alain Aspect (1982) closing timing loopholes, and pushed to airtight "loophole-free" tests by Anton Zeilinger and others (2015), measured values land near 2.4–2.8, far above 2.
  • It won the 2022 Nobel Prize in Physics, awarded to Aspect, Clauser and Zeilinger for these experiments.

The verdict is blunt: the universe is not both local and pre-determined. At least one of those comforting assumptions is false. Entanglement is a real, irreducible correlation — not a hidden message agreed on in advance. This is the bedrock that powers ideas like Shor's algorithm and the strange speedups of quantum computing.

Where It Matters

A Bell violation is not just a curiosity — it is a certificate. If your numbers beat 2, no local, pre-arranged process could have faked them, and that guarantee is bankable:

  • Device-independent cryptography: two parties can share a secret key whose security rests only on a measured Bell violation, even if the hardware was built by an adversary.
  • Certified randomness: a violation proves the outcomes were not pre-determined, so they are genuinely random — the gold standard for cryptographic seeds and lotteries.
  • Verifying quantum hardware: a Bell test is a sanity check that a device is really exploiting entanglement and not just classical noise.
  • Foundations of computing: the same nonlocal correlations underpin the resources behind quantum simulation and quantum advantage.

What began as Einstein's complaint about "spooky action" is now a tool engineers use to build things classical physics simply cannot.

Conclusion

Bell's genius was to take the grandest question — is reality local and pre-determined? — and shrink it to a number you can measure in a lab. Common sense demands S ≤ 2. Entangled particles cheerfully return 2.83. There is no loophole left to hide in.

So the spookiness Einstein hoped to explain away is real, and the 2022 Nobel Prize made it official. The next time you hear that quantum computers exploit something "deeper" than classical machines, remember: that depth has a name, a number, and an experiment. It is the same strangeness that gives quantum simulation its power.

Share this article

Pick a channel — or use your device's native share sheet.

Comments

Loading comments...

https://www.kipuhub.com/en/article/bell-inequalities/Content licensed under CC BY-NC 4.0.