I will present a new model for the formation and evolution of galaxies. I will first introduce the ingredients included in standard Semi-Analytical Models (SAMs), and will summarize the successes and limitations of such models. A different complementary approach will be presented which is conceptually simpler and closer to the halo model. In our approach all the processes which govern galaxy formation are set by the host halo mass and redshift, but the galaxies are evolved within the complex structure of dark-matter merger-trees. Although the physical 'recipes' are simple and smooth, the properties of galaxies vary significantly due to the merger-histories of their host halos. I will compare our approach to a standard SAM, and will also explore a set of very
different models which all fit the observational data well. This will be used to identify the observational constraint which could distinguish between different models, and to explore the level of uniqueness inherent in galaxy formation models.
