Interannual variability and spatial synchrony of seed production, known as masting, have far-reaching ecological impacts including effects on forest regeneration and the population dynamics of seed consumers, predators, parasites, and diseases. Because the relative timing of management and conservation efforts in ecosystems dominated by masting species often determines their success, there is an urgent need to study masting mechanisms and develop anticipatory forecasting tools for seed production. In this study, we aim to establish seed production forecasting ("mast casts") as a new branch of the discipline. To assess the current state of the ability to model intrinsic and extrinsic drivers of masting, we evaluate the predictive capabilities of three models -foreMast, ΔT, and a sequential model based on the developmental steps of seed production - designed to predict seed production in trees using a pan-European dataset of Fagus sylvatica seed production as a case study. Our hindcasting analysis indicates that the models are moderately successful in recreating seed production dynamics, with the best-performing model achieving R2 = 0.66. The availability of high-quality data on prior seed production significantly improved the model's predictive power, with R2 = 0.78 in the sequential model, suggesting that effective seed production monitoring methods are crucial for creating reliable forecasting tools. In terms of extreme events, the models were better at predicting crop failures than bumper crops, likely because the factors preventing seed production are better understood than the processes leading to large reproductive events. To create widely applicable tools, we need models that can make forecasts over large areas with readily available data. We summarize the current challenges and provide a roadmap to help advance the discipline and encourage the further development of mast forecasting.