Number of galaxy cluster candidate members selected photometrically (top) and the fraction of lost spectroscopically confirmed members normalised to N × σ_NMAD = 1 (bottom) as a function of N × σ_NMAD considering the CHANCES low-redshift magnitude-limited regimes: bright end (m_r < 18.5, red) and faint end (m_r ≥ 18.5, blue). Assuming N × σ_NMAD = 4 our target selection process selects the largest number of galaxy cluster candidate members with the lowest fraction loss of spectroscopic members, while assuming N × σ_NMAD = 1 we select the lowest number of galaxy cluster candidate members, with the highest fraction of lost spectroscopically confirmed members. The first panel shows this by considering only our internal LS-DR10-CBPF-z_phot estimations, while the middle and right panels show a direct comparison between LS-DR10-CBPF- and S-PLUS-z_phot, considering 10 clusters with good T80S/S-PLUS coverage (middle panel) and an extended sample considering 16 clusters with T80S/S-PLUS coverage >80%(rightmost panel); see Section 2.3 for details. Assuming N × σ_NMAD = 3.5 and 3 for S-PLUS and LD-DR10-CBPF-z_phot, respectively, successfully reduces the total number of the galaxy cluster candidate members selected with a spectroscopic members loss of ∼2 − 3%.