Comment by JohnKemeny

10% is ridiculously high compared with published papers in math journals where the number of critical mistakes is near zero.