Crustal deformation patterns are affected by multiscale granulation and healing processes associated with phase transitions between continuum and discrete states of rocks. The ongoing continuum-discrete transitions are accompanied by progressive evolution of disordered fault networks to dominant localized fault zones, development of bimaterial interfaces, and increasing dynamic weakening of fault surfaces. Results on individual fault zones point to three general dynamic regimes. The first is associated with broad range of heterogeneities, little dynamic weakening, power law frequency-size statistics, temporal clustering of intermediate and large events, and accelerated seismic release before large earthquakes. The second is associated with relatively uniform localized structures, significant dynamic weakening, characteristic earthquake statistics, and quasi-periodic temporal occurrence of large events without precursory accelerated release. For a range of conditions, the fault zone response can switch back and forth between the foregoing two dynamic regimes. Higher temperature, fluid content, and thickness of sedimentary cover reduce the seismic coupling in a region and change the properties of local earthquake sequences. Brittle regions with high seismic coupling have few foreshocks and long-duration aftershock sequences with high event productivity, whereas more viscous regions with low seismic coupling have increased foreshocks activity and low-productivity aftershock sequences or swarms. The results provide criteria for organizing data in classes associated with different evolutionary stages and different regional conditions. An ability to recognize the dynamic regime of a given fault zone or a region can increase the information content of the data and lead to improved strategies for seismic hazard assessment.