A single test which proves some piece of theory wrong is more valuable than a hundred tests showing that idea might be true.
or simply: