This math thing: you're doing it wrong. Theories are not "arbitrary", and you cannot outbest theory "in practice".

They only way to do this is if you have an ad-hoc faulty theory.