Do you notice any biases when adding IDs to unknowns?

For me, I’ve noticed that most of the unknowns I try to identify come from new users, folks having 200 or fewer observations, who haven’t yet learned that they need to add an initial ID.