Violation of Bell Inequalities as a Violation of Fair Sampling in Threshold Detectors

TL;DR

This paper demonstrates that the violation of Bell inequalities observed in experiments with threshold detectors can be explained by a sampling bias caused by their thresholding behavior, challenging interpretations of quantum nonlocality.

Contribution

It reveals that the combined effect of detector thresholds can induce a sampling bias, leading to apparent Bell inequality violations without requiring quantum nonlocality.

Findings

Threshold detectors can cause sampling bias in photon detection.

Sampling bias can mimic Bell inequality violations.

Threshold effects are crucial in interpreting quantum optics experiments.

Abstract

Photomultiplier tubes and avalanche photodiodes, which are commonly used in quantum optic experiments, are sometimes referred to as threshold detectors because, in photon counting mode, they cannot discriminate the number of photoelectrons initially extracted from the absorber in the detector. We argue that they can be called threshold detectors on more account than that. We point out that their their functioning principle relies on two thresholds that are usually thought unimportant individually in the context of EPR-Bell discussion. We show how the combined effect of these threshold can lead to a significant sampling selection bias in the detection of pairs of pulses, resulting in an apparent violation of Bell inequalities.