Rad Chmiel made two astonishing blunders at the end of his game:

- the first was 63...Rxg5??, which actually deserves 3 question marks because it is wrong on that many level:

(1) Black's pieces were in optimal positions and he was the verge of placing white into zugwang - all he had to play was 63...a4, and white would have had to either lose either one of his pawns on the very next move (or the bishop if white played 64.Bh4?).

(2) Black gave up the most powerful piece on the board, one capable of cutting off white's king from the queenside, for one of lesser quality and value.

(3) Not only did black lose out materially, but he also allowed white to move one step closer to the queenside by capturing on g5.

- the second blunder was 65...Kxf6??, when again 65...a4 wins easily as white cannot get inside the pawn's "square" (i.e. the square bound by a4-a1-d1-d4) and therefore cannot capture the a-pawn which must therefore promote. Instead, the move played was simply a needless material grab that did nothing but allow white the vital tempo to get to a well-known theoretically drawn ending. I'm surprised a 1700-rated player would allow this.