if the test differences are significant I would agree to not pollote production but often they are not so the harm is minimal.
if the test differences are significant I would agree to not pollote production but often they are not so the harm is minimal.