How do university master's students in mathematics reason about a simple problem involving coin tosses? We answer to this question by analyzing the participants' written reflections in terms of recency and equiprobability effects understood non-normatively. Overall, the participants display a strong tendency to deem the problem's coin to be fair despite indication that this should be the case and contrary to explicit evidence presented in the problem. In view of this, we further elaborate on the connection between the participants' answers and their education in mathematical probability.