Recent focus on the importance of cold, unshocked gas accretion in galaxy formation - not explicitly included in semi-analytic studies - motivates the following detailed comparison between two inherently different modelling techniques: direct hydrodynamical simulation and semi-analytic modelling. By analysing the physical assumptions built into the GASOLINE simulation, formulae for the emergent behaviour are derived which allow immediate and accurate translation of these assumptions to the GALFORM semi-analytic model. The simulated halo merger history is then extracted and evolved using these equivalent equations, predicting a strikingly similar galactic system. This exercise demonstrates that it is the initial conditions and physical assumptions which are responsible for the predicted evolution, not the choice of modelling technique. On this level playing field, a previously published GALFORM model is applied (including additional physics such as chemical enrichment and feedback from active galactic nuclei) which leads to starkly different predictions.