Our demographic study provides a detailed picture of the short-term, weekly, mortality fluctuations charactering females and males over more than a decade. We follow an evidence-based approach, since we use data from the Human Mortality Database to detect stylized empirical evidence about the behaviour of mortality, during normal times and in the most recent time span, highly affected by the COVID-19 pandemic disease. Our study relies on time-series mortality data collected at a finer scale than traditionally done in the actuarial literature and encompasses different age groups and gender. The empirical evidence represents the starting point for exploring the future mortality patterns. Our quantitative analysis, namely based on stochastic mortality modelling, indeed, answers the question whether accounting for higher frequency mortality data delivers more reliable mortality projections, namely projections that approximate more closely the realized mortality phenomenon.