SrTi0.5Mn0.5O3 (STM55) perovskite has been proven experimentally to be a promising material for hydrogen production using thermochemical water splitting . However, optimizing the performance of STM55 in this application requires synergistic study of its thermochemical properties using both experiments and computations. In particular, there are several questions with regard to its crystal structure, magnetic ordering, electronic structure, and more importantly its off-stoichiometry variation as a function of temperature and oxygen chemical potential. Herein, we tackle these questions employing density functional theory calculations equipped with onsite Hubbard (U) terms applied to Mn and Ti 3d states. Since prior studies did not convincingly determine best values for U_Ti and U_Mn, we decided to systematically vary these Hubbard terms between 0 and 5 eV in steps of 1 eV to understand their effect on the properties of STM55. Our analysis revealed that U_Mn has a more dominant impact on the properties of STM55 in comparison to U_Ti. Higher U_Mn  decreases  the oxygen vacancy formation energy and hence the oxide becomes more  reducible. More importantly, we found out that U_Mn = 1 eV is the best in computing oxygen off-stoichiometry of this oxide in comparison to prior experiments. However, the temperature dependence of the off-stoichiometry is not well-reproduced by our DFT calculations. Further analysis showed that mere reliance on temperature-independent formation energies of oxygen vacancies is not sufficient to correctly reproduce the temperature dependence of the off-stoichiometry. This hints to the importance of including thermal effects in the DFT study of this compound. In addition to this, our study showed that STM55 behaves as a metal (regardless of the ingredients of the DFT calculations), and is best modelled as a ferromagnet in the pm3m crystal structure.  We also predicted the enthalpy and entropy of reduction of this oxide based on an Arrhenius-type analysis. While our model is capable of predicting the temperature dependence of these quantities, the U_Mn that leads to the best agreement with experiments is different than the one that best describes the off-stoichiometry. This led to the conclusion that there is no unique set of Hubbard parameters that can completely describe the behavior of STM55.  Our work furnishes the grounds for further studies on this oxide, especially kinetic studies. Additionally, our results have general implications for the class of metallic oxides.