Even at highest levels players make significantlysignificant, but human, mistakes. On the other hand there existexists a great amount of opening theory, with lines studied in deepthdepth, especially the mainlines. Isn't the precision of theory evaluation and human imprecision contradictingin contradiction? Is it generally difficult to punish small opening sins even at the highest (human) level (Evans-Gambit, Cozio defense, 1. ... a6)? If so, is the fact that most advanced players learn and play some kind of mainlines just some kind of gentlemen's agreement? Or do they score significantly better?