Added 1 m/s yearly signals have 44.5% to 99.6% recovery across 32 phases in HD 86728 simulations with independent Gaussian noise

We tested added circular velocity signals at public NEID observing times for HD 86728. We used 220 exposures on 162 nights from 2021 to 2025. A header audit removed three unflagged exposures with unsuitable wavelength calibration. We fit the known 31.1503-day signal and instrument offsets. We calibrated a global 1% detection threshold for a 2 to 1000 day search with 10000 Gaussian null draws. At a period of 365.25 days and an amplitude of 1 m/s, correct recovery ranges from 114/256 to 255/256 draws across 32 phases. These fractions are 44.5% to 99.6%. Mean recovery is 80.4% under independent Gaussian noise. A model with independent noise and a fixed 10-day correlated component gives 55.4% mean recovery. Its likelihood gain is small for one extra parameter. We provide a phase-resolved sensitivity map and distinguish exact-period threshold crossing from full search recovery. Neither the independent nor the mixed noise model finds a residual signal above its threshold. A 2.008-day peak crosses the threshold only under a much worse model that forces all extra noise to be correlated. It does not establish a planet candidate. Recovery requires a peak within one inverse time baseline of the added frequency. The results apply to simulated noise, fixed covariance estimates, the stated mean model, and finite tested grids. They do not rule out planets without these assumptions. The package preserves public inputs, header hashes, source code, figures, results, and reproduction instructions.

HD 86728 · NEID · radial velocity · signal recovery · observing window · correlated noise · calibration audit · public data