In the Dynamical Dark Matter (DDM) framework, the dark sector comprises a large number of constituent dark particles whose individual masses, lifetimes, and cosmological abundances obey specific scaling relations with respect to each other. In particular, the most natural versions of this framework tend to require a spectrum of cosmological abundances which scale inversely with mass, so that dark-sector states with larger masses have smaller abundances. Thus far, DDM model-building has primarily relied on non-thermal mechanisms for abundance generation such as misalignment production, since these mechanisms give rise to abundances that have this property. By contrast, the simplest versions of thermal freeze-out tend to produce abundances that increase, rather than decrease, with the mass of the dark-matter component. In this paper, we demonstrate that there exist relatively simple modifications of the traditional thermal freeze-out mechanism which "flip" the resulting abundance spectrum, producing abundances that scale inversely with mass. Moreover, we demonstrate that a far broader variety of scaling relations between lifetimes, abundances, and masses can emerge through thermal freeze-out than through the non-thermal mechanisms previously considered for DDM ensembles. The results of this paper thus extend the DDM framework into the thermal domain and essentially allow us to "design" our resulting DDM ensembles at will in order to realize a rich array of resulting dark-matter phenomenologies.

In general, non-minimal models of the dark sector such as Dynamical Dark Matter posit the existence of an ensemble of individual dark components with differing masses, cosmological abundances, and couplings to the Standard Model. Perhaps the most critical among these features is the spectrum of masses, as this goes a long way towards determining the cosmological abundances and lifetimes of the corresponding states. Many different underlying theoretical structures can be imagined for the dark sector, each giving rise to its own mass spectrum and corresponding density of states. In this paper, by contrast, we investigate the spectrum of masses that emerges statistically from underlying processes which are essentially random. We find a density of states $n(m)$ which decreases as a function of mass and actually has an upper limit $m_{\rm max}$ beyond which $n(m)=0$. We also demonstrate that this "emergent" density of states is particularly auspicious from the perspective of the Dynamical Dark Matter framework, leading to cosmological abundances and decay widths that are suitably balanced against each other across the dark-matter ensemble. Thus randomness in the dark sector coexists quite naturally with Dynamical Dark Matter, and we examine the prospects for observing the signals of such scenarios in dark-matter indirect-detection experiments.