The machine‐side oscillation (MSO) of the full converter‐based wind generation (FCWG) systems is a critical threat to the reliable wind power supply. The introduction of the auxiliary resonance controller (ARC) to the machine‐side converter (MSC) controls of FCWG can effectively improve the converter‐driven stability of the power grid, but it would also complicate MSOs of FCWG. Thus, it is of great significance to study the damping feature of MSOs of FCWG. In this paper, the energy flow analysis (EFA) is applied to quantitatively investigate MSOs of FCWG. Firstly, the configuration and machine‐side control loops of FCWG are briefly introduced. Then, a mode screening‐based EFA is proposed for evaluating the damping feature of MSOs of FCWG in the time domain by applying the principle of the Laplace transform. After that, the eigenvalue analysis is conducted for MSOs of FCWG in the frequency domain to lay a foundation for revealing the essence of EFA. On this basis, the consistency of EFA with the eigenvalue analysis is strictly proved based on the Parseval's Theorem, which is applicable for arbitrary control schemes of FCWG. Finally, the proposed EFA is applied in numerically investigating multiple types of MSOs of FCWG in case studies.