The number of muons in extensive air showers predicted using LHC-tuned hadronic interaction models, such as EPOS-LHC and QGSJetII-04, is smaller than observed in showers recorded by leading cosmic ray experiments. In this paper, we present a new method to derive muon rescaling factors by analyzing reconstructions of simulated showers. The z-variable used (difference of initially simulated and reconstructed total signal in detectors) is connected to the muon signal and is roughly independent of the zenith angle but depends on the mass of primary cosmic ray. The performance of the method is tested using Monte Carlo shower simulations for the hybrid detector of the Pierre Auger Observatory. Having an individual z-value from each simulated hybrid event, the corresponding signal at 1000 m from the shower axis, and using a parametrization of the muon fraction in simulated showers, we can calculate the multiplicative rescaling parameters of the muon signals in the ground detector even for an individual event. We can also study its dependence as a function of zenith angle and the mass of primary cosmic ray. This gives a possibility not only to test/calibrate the hadronic interaction models, but also to derive the $\beta$-exponent, describing an increase of the number of muons as a function of primary energy and mass of the cosmic ray. Detailed simulations show dependence of the $\beta$-exponent on hadronic interaction properties, thus the determination of this parameter is important for understanding the muon deficit problem.