The 2025 Kepler-451 TESS table has two used rows for 888 eclipse events

We checked the public machine-readable Table 1 of Er et al. (2025). Its 2,382 TESS timing rows describe 1,489 primary eclipse events. Of 2,376 rows marked for use, 888 events have two times and 600 have one. Three published linear clocks and a test based only on time gaps give the same event groups. The five published TESS means agree with averages over all retained rows to within 0.001 s. Giving each event equal weight changes a mean by at most 0.216 s. Selecting the smaller or larger stated error in each event changes a mean by at most 0.551 s. Exact bounds over every choice of one recorded time per event have a maximum width of 8.147 s. The published errors are 21.74 to 24.79 s. The stated error rule does not reproduce these errors from the retained rows. A rule that uses the scatter of individual residuals gives values within 0.009 s. The table has no cadence labels, so it cannot identify which row the authors intended to retain. This audit supplies an event record, source data, code, figures, and independent checks. It does not establish or reject a planet. The method wording was checked in the June 2026 author version.

Kepler-451 · TESS · eclipse timing · data audit · circumbinary planets · reproducible research