The purpose of this note is two-fold. First, we want to empirically explore how prior distributions influence posterior distributions. Second, we want to explore the trade-off between model complexity and big data. Most techiques taught in Statistics deal with naive models and small data, while the machine learning community tends to deal with simple models and large amounts of data. Bayesian approaches are particularly useful at capturing uncertainty in the parameters, with sparse data and complex models. With implementations like Stan, Bayesian methods are heading in a direction which will allow efficient computation on big data without without losing the transparency of directly specifying a complex model.