Comment by philipwhiuk

Underneath this means that you have say 50 tests and you grade each of them out of 10, then there was no test it did worse on.

The data doesn't support it being better on every test (sometimes the score will be the same imperfect one, sometimes both will have gotten a perfect score).