Six of eight selected early quadratic clocks have larger first TESS group mean errors than linear clocks in a 2022 catalog test with fixed errors

We tested whether an early curved transit clock predicts later times better than a linear clock. We tested a fixed 2022 public catalog. We fitted literature event times observed before 25 July 2018. We then predicted each system's first later TESS group mean. We used fixed Gaussian timing errors with a 25.92-second minimum. The score includes the full shared uncertainty of the early fit. Of 41 eligible systems, eight had an early BIC difference greater than 10 in favor of a quadratic clock. Six of these eight had larger absolute mean errors and lower Gaussian log scores. The median quadratic-to-linear error ratio was 3.70. Seven selected curves also failed the early chi-squared fit check at a tail probability of 0.01. Only the known WASP-12 trend passed both early rules. Its later mean error fell from 146.6 to 36.1 seconds. Checks of timing labels, errors, group gaps, and the training cutoff retained more failures than improvements in their selected samples. Source faults and shared reduction errors remain possible. This retrospective test does not establish a new planet or a physical cause. The package includes source tables, complete row decisions, forecasts, figures, code, and an offline repeat. A fresh-directory run gave identical bytes for all 45 derived result and figure files.

transit timing · ephemeris prediction · TESS · quadratic ephemerides · historical forecast · model validation · reproducible data