TRGB forward modeling results for NGC 1052–DF2. Top-left panel: the F814W luminosity function of the intrinsic model without contamination (gray) and with contamination added (red). Top-middle panel: the likelihood map from 1000 Monte Carlo runs, marginalized over the TRGB magnitude m_TRGB,F814W and the contamination parameter c, as defined in Equation (1). The parameter values at maximum likelihood are used to construct the best-fit model in the main panel. Top-right panel: the systematic offset (bias, in orange) and the scatter (in green) in the recovered magnitudes using the artificial star test. Middle-right panel: the bias and scatter from the top-right panel, fitted with a fifth-order polynomial. Bottom-right panel: photometric completeness, the fraction of input artificial stars that are recovered at each magnitude. Main panel: the best-fit model (red) of the observed F814W luminosity function (black), with a TRGB magnitude of m_TRGB,F814W = 27.67 from the maximum likelihood. The best-fit model is the intrinsic luminosity function with contamination added (shown in the top-left panel) and photometric errors applied (shown in the right panels). The reduced χ² value for this fit is 1.8. For comparison, the observed luminosity function of NGC 1052–DF4 (Danieli et al. 2000) is scaled and plotted in the dashed yellow line.