The TrES-1 b entries in a 2026 table give two period-change rates that differ by a factor of ten

Do the printed parameters and public TESS data support the reported period change of TrES-1 b? Table 4 of Kutluay et al. (2026) gives P = 3.0300696522 days and a quadratic coefficient of (-7.14 +/- 2.86) x 10^-11 days. Under the stated definition, these values give -1.487 +/- 0.596 milliseconds per Julian year. The printed rate is -14.9 +/- 0.6. The central rates differ by a factor of 10.019. This check does not show which entry is wrong or measure the physical rate. We also reduce eight public 120 second TESS light curves from 2019-2024. Coverage rules retain 54 transits, with no selection by timing residual. A TESS-only quadratic fit gives -5.18 +/- 28.19 milliseconds per year. Flux, baseline, limb-darkening and sector-error checks leave the constraint weak. The 21 retained 2024 times are 17.16 +/- 10.29 seconds later than a quadratic forecast fitted to the 123 transit times in Hagey et al. (2025). The error rises to 14.73 seconds when the training covariance is scaled for excess scatter. This retrospective test gives no strong forecast failure. The package preserves the exact source data, code, selection audits, times, covariances, figures and checks. No planet detection or physical cause is inferred.

TrES-1 b · transit timing · TESS · orbital period change · quadratic ephemeris · reproducibility · unit conversion