Jimenez Salas vs Hryzlova

Through move twenty-eight a small structural edge sat on the white side: the isolated c4 pawn, the dark squares without a defender. 29.f4?! was White’s first inaccuracy, and Sofiia Hryzlova took eight minutes and forty-two seconds for the reply. 29...Bb4 balanced the game to zero: the bishop leaves c5 before the capture on e5, and after 30.fxe5 Bxc3 31.exd6 Bf6 White had the passed pawn on d6, but Black had the two bishops off the board and the a1 diagonal in view.

The objectively strongest move was now 32.c3, which keeps the d6 pawn defended and holds the white structure together: 32.c3 g6 33.e5 Bd8 34.R1d4 Kg7 leads to a position where Black stands half a pawn better but White is stable. The game saw 32.d7?, and that is one square too far. The pawn leaves the cover of the e pawn, walks into the diagonal of the bishop on f6 and stays put behind its own lines, without any way to promote for gain.

The answer was 32...Bd8, and it is the only move that holds the position together: 32...c3 33.d8=Q+ Bxd8 34.Rxd8+ Kh7 gives the bishop for the pawn and falls a pawn and a half behind. The bishop on d8 blocks the passed pawn, and Black now stood clearly better: the d7 pawn is an object to blockade, no longer an attacker. Before every advance of a passed pawn late in the middlegame, the question is whether it is still defended after the move, and what the opponent does with it on the new square.

After 32...Bd8 Black stood better, but the white rook now attacks the blockading bishop. The only move that keeps the winning position is 33...Re6, and the reason is concrete: 33...Re6 34.Rxe6 Rxe6 35.Qxc4 Qb6 trades off the most active white attacker, clears the e file for Black’s own rook, and leaves Black with a clean piece for pawns. The white queen can take the c pawn, but the bishop on d8 stays both a blockader and an active force at the same time.

33...Qc7?? was the move Sofiia Hryzlova played after eight minutes and fifty-three seconds of thought, and it is a tactical error. It wants to bring the queen off the edge to a central square and defend not by protecting the e5 rook but by reorganising its own coverage. It misses 34.Re8+, and after 34...Kh7 White has the rook on the eighth, the passed pawn on d7 and the black queen without an active task. Over the next three moves 35.Qb5 g6 36.Qd5 Rf6 37.e5 Rf5 White built up a clearly winning position.

It was humanly understandable, because 33...Re6 offers a rook trade in which the bishop on d8 suddenly looks undefended; the concrete calculation shows that after the double capture on e6 the bishop is defended by the rook and the passed d7 pawn stands without support. An attack on the king is defused by trading the most active attacker off, not by reshuffling the defenders. Nine minutes for a move that leaves that attacker on the board is a calculation that rarely balances.

Between moves thirty-four and thirty-seven White kept the winning position without adding to it; every move melted a shade off the edge, but the advantage was still there. The only supporter of the d7 pawn was the queen on d5, and Black had the c pawn still lying on the queenside. In such a position the move that keeps the attack running is 38.Qd4: the queen to the middle, the h8 diagonal in view, the rook on d1 free for the e file, and White stays a whole pawn ahead.

The game saw 38.Rd4? after two minutes and forty-two seconds. The move wants to double up on the d7 pawn and keep an eye on the c file; it does neither. The d pawn on the seventh is an attacker, not an object to defend, and the rook on d4 does not take the burden of covering it off itself. 38...c3 came after five minutes and thirty-one seconds of thought, and with a single black pawn the balance was on zero again.

After 39.b3 a4 40.b4 came the question of the time control: Black had six minutes and nine seconds for move forty, and 40...Re6 was the only move that put the rook on the half-open file and defended the c pawn; thirty-six seconds were left on the clock when it was played. In a winning position the move that keeps the own activity going matters more than the move that secures a piece of material; a player who defends the passed pawn instead of pushing it hands the initiative away.

The time control had put nineteen minutes and fifty-seven seconds on White’s clock and fifteen minutes and thirty-one on Black’s. 41.Qc5 hit the black queen, 41...Qxe5 took the pawn in the middle, and with 42.Qxe5 Rexe5 the queens were off. In the position that remained Black had two active rooks, the bishop and two passed pawns on c3 and a4; White had the passed pawn on d7 and the rooks on d4 and e8. 43.Rd1 Bg5 put the bishop on the long diagonal and aimed it at the d8 square.

44.g3 would have left the d7 pawn on the board for one more move and allowed the promotion only after a quiet move; 44.d8=Q?? played it at once, onto a square the bishop on g5 was covering. 44...Bxd8 45.Rdxd8 took the new queen for the bishop and ran into 45...Rf1+ 46.Ka2 Rxe8 47.Rxe8: White had two rooks, Black had a rook and the two passed pawns c3 and a4, and every piece of activity sat on the black side.

The rest was bookkeeping in mutual time trouble. After 47...Rf2 48.Kb1 Rxg2 49.Rc8 Rxh2 50.Rxc3 the pawn race began: 50...Rf2? gave half a pawn back, but the black pawns kept running, and after 60.b8=Q g2 61.Qa7 g1=Q the new black queen was on the board before the white checks of 62.Qxf7+ Kh6 63.Qf6+ Kh5 64.Qf5+ Qg5 65.Qf7+ Qg6 66.Qc7 Qg4+ 67.Kf2 Qg1+ led into mate. Before every promotion the question is what the opponent does with the new piece on the promotion square; whoever counts the road to the square but not the attackers on it will give the new piece back.