Game theory is a powerful theory to model strategic decision-making, but also complex biological systems, such as cancer. In the past decades, game-theoretical models have helped to understand cancer, its response to various treatments, and to design better therapies. However, to fully utilize the potential of game-theoretical modelling in designing better anti-cancer therapies, we need more information on cancer population (ecological) and strategy (evolutionary) dynamics in response to treatment. Here we explore how transcriptomics data can be utilized as inputs of game-theoretic models for predicting evolutionary response to irradiation using patient-derived glioblastoma organoids. For that purpose, we utilize both supervised and unsupervised methods to disentangle cell mixtures from publicly available datasets and identify proportions of relevant cancer cell types in patient-derived glioblastoma organoids treated with radiotherapy. We then fit these proportions to the replicator dynamics, the most classic evolutionary game dynamics, to predict both transient evolutionary dynamics and evolutionary stable cell proportions in these organoids. Hereby we demonstrate a methodology that can be utilized to design evolutionary therapies, i.e. therapies that anticipate and prevent evolution of therapy resistance in cancer cells. Our predictions in glioblastoma suggest that hypoxia is the most important factor in identifying short-term response to irradiation, while this seems much less relevant for the long-term response. Further, we conclude that supervised methods are the best way to estimate cancer evolutionary dynamics when therapy resistance is a qualitative evolutionary trait. We provide a pipeline for designing better therapies through testing evolutionary responses in patient-derived organoids, while in parallel the ecological response can be tracked through serum biomarkers and imaging in the corresponding patients.