Quantifying errors without random sampling

Quantifying errors without random sampling