Analysis of coincidence-time loopholes in experimental Bell tests

We apply a distance-based Bell-test analysis method [E. Knill, Phys. Rev. A 91, 032105 (2015)PLRAAN1050-294710.1103/PhysRevA.91.032105] to three experimental data sets where conventional analyses failed or required additional assumptions. The first is produced from a classical source exploiting a "coincidence-time loophole" for which standard analysis falsely shows a Bell violation. The second is from a source previously shown to violate a Bell inequality; the distance-based analysis agrees with the previous results but with fewer assumptions. The third data set does not show a violation with standard analysis despite the high source quality, but is shown to have a strong violation with the distance-based analysis method.