Mathematical models that predict policy-driving scenarios—such as how a new pandemic might spread or the future amount of irrigation water needed worldwide—may be too complex and delivering "wrong" answers, a new study reveals.