IM Fabian Bänziger (SUI, 2426) vs IM Mukhammadali Abdurakhmonov (UZB, 2390) · ½-½
Board 2, Switzerland vs Uzbekistan 3 · Queen’s Gambit Declined, Exchange Variation, Positional Variation (D35) · 90 minutes for 40 moves, then 30 minutes, plus 30 seconds per move
By Professor Zugzwang AI Coach . This analysis was written by a Bogner Chess AI coach.
Round eight, table 23, Switzerland against Uzbekistan 3. On board two sat IM Fabian Bänziger with White against IM Mukhammadali Abdurakhmonov, 2390, thirty-six points below him and playing at home on Uzbek soil. In round eight a team asks this board for one thing: a half point that keeps the match together, and if the game allows it, the full point. With White you want to try for both.
Bänziger tried, and traded queens on move ten so that Black kept doubled f-pawns and the open g-file into the bargain. Then came instead of tucking the king away, and over the next six moves the assessment slid from a small White edge into an equally small White deficit. From move twenty-six on he played only moves that force the draw. In the closing stretch the clock stood closer to him than to the opponent.
The double queen trade: 10.Qxf6 gxf6
White had placed the queen on f3 with , from where she looks twice at d5 and at the bishop on f5, and sidestepped. Then came the sequence that fetches the endgame in this line: . The queen trade is the price for making Black recapture on f6, and neither recapture is free. Taking with the bishop costs castling; taking with the g-pawn costs the pawn structure.
The bishop on h7 sits afterwards on a diagonal that leads to nothing, the black king has to find a way along the first rank, and the pawns on f6 and f7 defend each other but nothing else besides. The compensation is the open g-file, which the rook takes almost by itself, and the bishop pair, which pays for the reckoning that anyone who plays the Exchange has to keep in mind: two bishops are worth half a pawn while they remain two.
For White the balance after move ten was a third of a pawn to the good, no great edge, but a real one. The only condition the position asks is that his king gets somewhere. Ordinarily it goes to c1 or to d2, unhurriedly, and White then sets about pinning the bishop on h7 to the wall. That is how one plays these endings. That is not how this one was played.
IM Mukhammadali Abdurakhmonov (UZB, 2390)
IM Fabian Bänziger (SUI, 2426)
The attack without king safety: 11.h4
saves the castling move for later and wants instead to pick at the black king’s pawns, which opened themselves with . The idea is not a bad one: the pawn wants h5, then g4 and h6, the h7 bishop’s diagonal would be blocked, and with the rook on h1 the battery on the open file is already there. Eleven minutes forty-nine seconds Bänziger spent on it.
The order in which such a plan comes together is a different one. The king first has to go somewhere that is no longer the middle. The main line runs , and after it White stands not quite half a pawn better: the king on d3 is an active endgame piece, the bishop on h7 has traded with its own bishop, and then the rook goes onto the c-file or the g-file in its own time. does almost the same job, a hair worse, but still with the king on land.
sells part of that edge. After the natural reply the position stood almost in balance, and the white king’s home was what White had promised himself for the opponent: a target, only on the wrong side. What began as a small advantage in the opening became a good start to a dangerous plan, and the move the plan needed from the outset never came: the king did not, in the whole game, ever castle.
The rule for your own games: in an endgame with the bishop pair against doubled pawns you win with patience, not with tempo. The attack you do not have costs more than the attack you begin one move later. Before any h-pawn advance the question worth asking is where your own king sleeps if the plan does not work.
IM Mukhammadali Abdurakhmonov (UZB, 2390)
IM Fabian Bänziger (SUI, 2426)
The wrong piece forward: 17.Rg1 instead of 17.Nh5
After six more moves the balance stood at zero, and White had a choice between two continuations of the original idea. One was , the knight going from f4 to h5, controlling g7 from there and clearing the way for the h3 knight to come to f4. The main line is quiet: , and the position keeps its balance, with two knights on the kingside and a rook running along the third rank.
The other was , the rook taking the g-file before anything has happened there. The move prepares , the pawn advance the plan has been asking for since move eleven. But the rook on g1 is a passive move in a position where passivity no longer works: it changes nothing about the cover of Black’s d5 pawn, about the bishop on h7, or about what the black king does next. asks those three questions, asks none.
After Black had walked his king to safety, moved the rook to a5, and kept a good half pawn to himself over the next three moves. followed, thought about for thirteen minutes five seconds, and then , eleven minutes fifty-one, to do afterwards what would have been at the start. The clock at that point already showed twenty-two minutes for White against forty-nine for Black.
Objectively was not a coarse mistake, and the reasoning behind it was sound: occupy the file before it is opened. Practically it was the move that pushed the wrong piece forward. The rule: when an attacking piece and an attacking preparation are both on offer, the piece almost always wins, because it counts the opponent’s replies. A pawn and a rook count nothing; they only open the file.
IM Mukhammadali Abdurakhmonov (UZB, 2390)
IM Fabian Bänziger (SUI, 2426)
The path to the draw: 26.Ng5
The road here was a sequence of moves that were all the best on the board. After Black had opened the structure without having to, and White could trade back: cleared the queenside rooks and the c2 weakness. Then , the pawn push that suddenly has the right answers, , and the black king stands in the middle of the board.
From here everything runs by force. aims at f7, chases the king, takes the bishop that a moment before was ruling the long diagonal, and returns the pawn it costs. The black rooks, bishop and knight have all found a move in between, but none of them changes the count: after the exchange a white rook with a bishop and a pawn less stands against a black rook with bishop and knight and a pawn more, the same material balance as before.
The last task was the rook-and-bishop against rook-and-knight ending. Here White found the move that points at the thin edge of the position: , the knight put next to the black bishop. The move offers the piece, because after follows and the remaining material cancels itself out. and a rook move later the clocks stood at four minutes fifty-two for White, but nothing threatened him beyond losing on time, and ended the game at the handshake.
The rule for your own games: when you are supposed to stand better and yet find yourself defending, the move that forces material off is usually the best one. The exact calculation is one of the things that a world-class player needs and that a club player can already practise. Count first, trade second, never the other way round.
IM Mukhammadali Abdurakhmonov (UZB, 2390)
IM Fabian Bänziger (SUI, 2426)
Verdict
Half a point with White on board two, from a position that started at move ten as a small edge and had to be managed as a small deficit by move nineteen. The plan of attacking with before tucking the king away was the hard choice of the game, and the four long thinks totalling nearly fifty-three minutes show that it did not come easily. What came after was what the craft asks for in such cases: no more mistakes, seventeen best moves in a row, a liquidation that has to be calculated all the way to the ending, and a knight on g3 that clears the path for the bishop. Half a point against a titled player, without the clock ever falling under two minutes: for the Swiss team on a day like this, that is the point that carries a match.
Moves
½-½
Was this analysis useful?