> or just silently accept wrong results

Silently wrong results are very fashionable these days, you know. Deterministic results are very 2010s.