According to research by UCLouvain and Imperial College London, I have 83% chances of being correctly re-identified in a sampled and anonymized dataset. What about you?