In this paper, I will briefly survey probabilistic logics and investigate logics of probabilistic epistemic change, conjectures, and discovery of chances by resource-bounded agents in the sense of the limitation of the available data and its temporal instability, called dynamic probabilistic logic (DPrL), based on probabilistic dynamic logics and probabilistic Kripke systems, of which equivalence is defined by probabilistic bisimulation. DPrL has actions of form of [pi!] and probabilistic formula (P>=r)p. Furthermore, DPrL will be extended by adding formulas of form of [\exists F]p, a future operator over branching times, and formulas of form of S_ap, of which meaning is `happening of p is significant to agent a', defined by normalization of probability of p by agent a's expectation. The resulting system is called Dynamic Conjecture Logic (DCL), and in DCL we can describe `subjective chance' in terms of the Chance Discovery.