Abstract: This study investigates the complex strategic interactions among three core stakeholders—the government, land transferees (new agricultural operators), and smallholder farmers—within the framework of well-facilitated farmland (high-standard farmland) construction in China. It aims to decipher the evolutionary dynamics of farmland transfer and to reveal the underlying mechanisms through which distinct construction models and key economic-agronomic variables influence the process of land consolidation and scaled operations. The analytical foundation is a tripartite evolutionary game model, grounded in the theory of bounded rationality. The model conceptualizes the government as a policy initiator pursuing socio-economic benefits, smallholders as land suppliers weighing transfer rents against farming income, and transferees as scale-seeking entities evaluating costs, subsidies, and risks.The research is empirically anchored in a case study of Xing’an League in Inner Mongolia, a pioneering region for county-wide high-standard farmland construction. The model integrates field-based data, incorporating parameters such as per-mu construction investment, maintenance costs, grain prices, yield increments, transfer rents, and subsidy levels. Two primary construction modes are simulated and contrasted: the traditional direct government investment model, where the state fully funds and executes the project, and the innovative build-first-subsidize-later model, where eligible operators invest upfront and are reimbursed post-construction upon approval. Numerical simulations are conducted using Matlab to visualize the dynamic evolutionary paths of the three parties' strategy choices over time. The simulation results yield several critical insights: 1) Both policy models can steer the system toward an ideal equilibrium (1,1,1), where the government promotes construction, smallholders transfer land, and transferees accept it. However, the build-first-subsidize-later model induces a significantly faster co-evolutionary convergence rate compared to direct investment. Introducing reputation incentives further accelerates this process. A crucial caveat is that if the post-construction subsidy is set too low, it triggers a strategic shift among transferees toward non-acceptance, underscoring the sensitivity of this model to financial adequacy. 2) Key variables exert differentiated, non-linear influences on the evolutionary trajectory. Positive drivers—including longer land transfer lease terms, higher grain yields, and increased market prices—robustly accelerate convergence. Negative inhibitors—such as rising farmland maintenance costs, increased transfer transaction costs, and higher crop production inputs—slow down the system's evolution. Critical regulatory factors like per-mu construction investment, land transfer rent, and scale-operation subsidies exhibit an inverted U-shaped impact; they promote convergence up to an optimal threshold, beyond which marginal benefits decline and may even reverse the trend due to excessive fiscal burden or diminished returns. Furthermore, analysis incorporating a risk preference coefficient reveals that entities with higher risk tolerance and greater expected utility are more predisposed to engage in the build-first-subsidize-later scheme. 3) Significant divergence exists between paddy fields and drylands. In paddy field management, the build-first-subsidize-later model proves most efficient, achieving the fastest tripartite co-evolution, likely due to higher and more stable returns on irrigation infrastructure investment. For dryland systems, transferees face greater self-funding pressure; thus, enhanced subsidies or complementary support mechanisms are essential to sustain their participation incentives and achieve effective scaling. Derived from these findings, the study proposes a multifaceted policy framework. It advocates for the cautious promotion and institutional refinement of the build-first-subsidize-later model, particularly prioritizing its application in high-return paddy field zones. Policy design must involve scientifically calibrating the thresholds of critical variables to avoid diminishing returns. Regionally differentiated strategies are imperative: bolstering financial and technical support for dryland areas and economically lagging regions, while consolidating effective models in core grain-producing zones. Finally, strengthening legal safeguards, formalizing land transfer platforms, and fostering multi-stakeholder coordination are vital to reduce transaction costs, mitigate risks, and ensure the sustainable alignment of well-facilitated farmland construction with the overarching goals of orderly land transfer, agricultural modernization, and national food security.