In this paper, we establish that, for statistically convex-cocompact actions, contracting elements are exponentially generic in counting measure. Among others, the following exponential genericity results are obtained as corollaries for the set of hyperbolic elements in relatively hyperbolic groups, the set of rank-1 elements in CAT(0) groups, and the set of pseudo-Anosov elements in mapping class groups.Regarding a proper action, the set of non-contracting elements is proven to be growthnegligible. In particular, for mapping class groups, the set of pseudo-Anosov elements is generic in a sufficiently large subgroup, provided that the subgroup has purely exponential growth. By Roblin's work, we obtain that the set of hyperbolic elements is generic in any discrete group action on CAT(-1) space with finite BMS measure.Applications to the number of conjugacy classes of non-contracting elements are given for non-rank-1 geodesics in CAT(0) groups with rank-1 elements.Theorem 1.2 (Mod). In mapping class groups, the set of non-pseudo-Anosov elements in a sufficiently large subgroup Γ has growth-negligible. In particular, if Γ has purely exponential growth, then the set of pseudo-Anosov elements is generic.An obvious instance with purely exponential growth is a subclass of sufficiently large subgroups, whose action on Y is itself SCC. This was studied in [40] as SCC subgroups, a dynamical generalization of convex-cocompact subgroups. We emphasize that there exists indeed non-convex-cocompact non-free and free subgroups admiting SCC actions on Teichmüller space. Hence, for these groups, the set of contracting elements is (exponentially) generic.In terms of Theorem 1.2, it would be desirable to investigate which sufficiently large subgroups have purely exponential growth. The similar problem has been completely answered in the setting of Riemannian manifolds with variable negative curvature by work of T. Robin [36]. In fact, his results were obtained in a more general setting. For any proper action on a CAT(-1) space, Roblin showed that the growth of the action is of purely exponential if and only if the corresponding Bowen-Margulis-Sullivan measure is finite on the geodesic flow. From this, we obtain the following application to discrete groups on a CAT(-1) space.Theorem 1.3 (CAT(−1)). Suppose that G acts properly on a CAT(-1) space such that the BMS measure is finite. Then the set of hyperbolic elements is generic.As suggested in Riemannian case, it seems interesting to develop an analogue of Roblin's result for sufficiently large subgroups in mapping class groups.We are now turning to the class of SCC actions, which admits stronger consequence by Theorem A. This class of groups encompasses many interesting clases of groups, listed as in [40]. Again, we first consider the instance of mapping class groups. They are known to act on Teichmüller space by SCC actions a result of by Eskin, Mirzakhani and Rafi [18, Theorem 1.7], as observed in [1, Section 10]. This is the starting point of our approach to mapping class groups,...