A fixed 40.54-day wave has gain ranges that include zero in five GJ 3323 velocity products

We test whether the reported 40.54-day signal of GJ 3323 predicts a held observing interval. We fix both published periods. We fit circular waves on the first intensive interval and keep their coefficients fixed for the next interval. A constant offset is allowed in the held interval. The discovery table has 56 early and 67 held velocities. The 40.54-day wave reduces held weighted squared error by 10.1 percent. Its nominal 95 percent range from 30-day block resampling is -45.7 to 39.9 percent. A newer published reduction gives 8.2 percent, a joint-template SERVAL product gives 19.6 percent, and two corrected RVBank products give -0.4 and -2.7 percent. All five 30-day ranges include zero. The 5.3636-day positive control gives a 39.3 percent gain in the discovery table, with a range of 13.1 to 55.4 percent. A 1.49 m/s injected wave gives positive gains at all 24 tested phases in the discovery, newer, and joint-template products, but none in the split-template SERVAL product. These phase trials are conditional tests, not completeness estimates. The data products share spectra. No new observation or new planet is found. The result does not reject the reported planet candidate. It shows that this fixed-wave test has limited precision and depends on velocity extraction. Exact inputs, code, source hashes, figures, and offline steps are included.

astronomy · GJ 3323 · radial velocity · HARPS · planet candidate · stellar activity · time validation · velocity extraction · null result