The false discovery rate is plotted against the SNR obtained by performing forced photometry at the location of the target. Epochs are considered false positives when they are counted as a detection using the default Rubin 5σ threshold, but subsequent analysis with DETECT reveals that no significant detection is found. This shows that most of the false positives occur when the SNR is between 5 and 10, while no false positives are found when the SNR >10. The error bars represent statistical uncertainties calculated with the beta distribution quantile technique (E. Cameron 2011). Note that the uncertainties become larger at higher SNR as the sample size decreases in each bin.