Isn't the opposite then a negative selection bias? If we have N children killed per year doing an activity A, and N anecdotes of children being perfectly okay,

Isn't the opposite then a negative selection bias? If we have N children killed per year doing an activity A, and N anecdotes of children being perfectly okay, neither matter without knowing the denominator of the number of child-activities per year.