There is a diminishing return to repeated confirmations, since each new instance adds less to the case for a causal theory. In such a situation, experimental failure, unexpected findings, and negative results can be what make for the bigger theoretical breakthroughs. Such results should contribute to theory development and not, as Popper urged, their outright falsification. The failure can show where a theory is to be improved or refined: it is an opportunity for the growth or new knowledge in response to a discrepancy experience. Such a norm is reflected in the non-monotonic reasoning that is useful in thinking about causation.