Pinching-antenna systems (PASS) represent a promising customizable wireless access mechanism in high-frequency bands, enabled by dielectric waveguides and movable dielectric particles, called pinching antennas (PAs). In this work, we study optimal position allocation of PAs in PASS for multicasting in the downlink when a line-of-sight (LoS) link does not necessarily exist between all users and the PAs. The multicasting problem is solved by leveraging minorization-maximization (MM) principle to yield a provably convergent algorithm. In each run of the MM based procedure, we solve a convex surrogate problem using two methods called the candidate search method (CSM) and the bisection search method (BSM). With both BSM and CSM, we not only report superior performance of the multicasting PASS in non-LoS environments compared to conventional antenna systems (CAS), but also determine that BSM yields better overall computational complexity when the number of users and PAs increases. For example, we report that when we have 8 PAs and 25 users, the execution time with the CSM is approximately 2.5 times that with the BSM.