How does a statistician ensure that an analysis that comprises of outputting N results is correct? Can this be done without manually checking each of the results? Some statistical approaches for this task of proof-calculation are described -- e.g. capture-recapture estimation and sequential decision making.