How would collider bias predictions differ quantitatively in actual Bell test data compared to quantum mechanics?

Collider Bias and Bell Test Predictions: The Quantitative Gap

The key insight is that collider bias—a selection effect arising from how particles are reconstructed from decay products—creates a fundamental asymmetry between quantum predictions and classical (local realist) models at colliders.

### The Core Problem

At traditional Bell tests, experimenters measure particle properties directly. At colliders, they infer properties (like spin) indirectly from decay kinematics. This indirect measurement introduces collider bias: the act of selecting events based on decay-product angles *depends on* the unmeasured parent-particle properties in a way that mimics Bell-test violations without requiring genuine quantum nonlocality [AG-2024.04-2060].

More starkly: a classical local hidden variable model (LHVM) can be constructed that reproduces *exactly the same scattering cross-sections* and spin-correlation measurements as quantum mechanics, making them experimentally indistinguishable *in principle* at colliders [AG-2024.01-521]. This means the quantitative predictions converge—not because the physics is the same, but because collider bias creates an unfalsifiable shadow of quantum correlations.

### When Differences Do Appear

However, the picture is more nuanced. LHVMs employing *classical* spin correlations (rather than arbitrary hidden variables) *can* be excluded by collider data—they produce measurably different predictions [AG-2024.01-521]. The quantitative gap emerges when you move beyond reconstructing spin correlations from cross-sections alone.

For systems with sufficient structure—like Higgs decays to four leptons—the full spin density matrix can be reconstructed [AG-2024.08-1023], and quantum predictions can violate generalized Bell inequalities (e.g., CGLMP) *with high significance* while classical models respect them [AG-2024.08-1023]. In one concrete example, dihadron correlations from light-quark pairs show potential Bell-inequality violation at >5σ under realistic conditions [AG-2025.01-1106].

### The Fundamental Loophole

Critically: Bell locality tests are *not possible* under the standard collider-measurement scenario, because the detector inefficiency (particles don't report correct outcomes with 100% fidelity) closes no loopholes—it creates one [AG-2025.08-1228]. Quantum correlations remain measurable and informative, but they cannot definitively rule out local realism using standard Bell inequalities.

Bottom line: Quantitatively, quantum and certain classical models can give *identical* collider predictions for cross-sections and naive spin correlations. Differences emerge only when (1) you assume classical correlations have specific structure, or (2) you reconstruct the full density matrix and test higher-order Bell operators. The collider-bias loophole is not a minor experimental caveat—it's a fundamental feature of indirect measurement.