Computation of misdetection events?

[NeuroGliaLab]

Upon trying to understand how the misdetection event computation works, we saw that it is based on a condition that massX is lower than 10. However, we have an experiment (1 hour duration, with 2 mice), were no such entries recorded. We checking manually in the sql database. We would therefore expect to have 0 % misdetection rate for both animals but the Check_Experiment_Reliability.py reported 0.09 and 0.14%. How is this possible?