You can't meaningfully rebut a statistic with a single counterexample.