GM Alexandra Kosteniuk (SUI, 2495) vs IM Mai Narva (EST, 2411) · ½-½
Board 1, Switzerland vs Estonia, women’s event · Sicilian Defence, Najdorf Variation, Opocensky System (B92)
By Professor Zugzwang AI Coach . This analysis was written by a Bogner Chess AI coach.
Round three, table 16, Switzerland against Estonia: GM Alexandra Kosteniuk had White on board one against IM Mai Narva, and the Estonian put the Najdorf on the board. Kosteniuk chose the Opocensky System with , the quietest of the main Najdorf roads, and after nine moves she still had exactly what that setup promises: space on the kingside, a closed centre and a clear plan.
The plan was never carried out. put the king precisely where the pawns were supposed to go, and a real advantage shrank to almost nothing in one move. From there the game turned: on move 24 White traded the wrong piece, on move 32 Black stood on the brink of winning and handed it back on that very move with a single bishop move. A draw in 48 moves, and the game teaches in both directions.
The pawn that was not taken: 9...exf4
asks the e5 pawn a question, and the answer is . After that the position stays in balance. The game saw , a perfectly natural developing move, and it costs about two thirds of a pawn, because White now gets to push that pawn herself.
is the move Black had to prevent. It attacks the bishop on e6, and it shuts the centre for the time being: no pawn can capture anything any more, so only the wings decide. The bishop went to , the best of the three tries; and both lead to slightly worse positions.
The lesson holds for every f4 in the Sicilian: work out first what the position looks like once White is allowed to play f5. If your own bishop gets chased and the pawns freeze, capturing on f4 is not a matter of taste, it is the move.
IM Mai Narva (EST, 2411)
GM Alexandra Kosteniuk (SUI, 2495)
The storm that never started: 11.O-O
With the centre closed, the board belongs to whoever advances on a wing first, and White was ready: , and the storm is rolling while the black king is still in the middle. That is worth about three quarters of a pawn and more than any alternative, even if and are not far behind.
The game went . The king thereby walks to the side where the pawns are supposed to run, and from then on every storming move needs thinking about twice. One move was enough for the advantage to be all but gone.
The game demonstrates it itself: , the same moves as in the variation, only four moves later and with the king on g1. In the meantime Black had castled, and cost another tempo to step off the g-file that White’s own storm is meant to open. A rule for closed centres: the wing first, the king second. Castle on the side you intend to attack and every pawn tempo costs you twice.
IM Mai Narva (EST, 2411)
GM Alexandra Kosteniuk (SUI, 2495)
The wrong piece traded: 24.Nxc5
The game tipped over in two steps. recaptured the pawn at once; , leaving it alone for a moment, was almost a whole pawn better. Two moves later offered a trade, and White accepted: , and Black is a good pawn better without a single pawn having fallen. would have kept the balance, and so would or .
The reason stands on c5. As long as a knight sits there, it is a knight like any other. A bishop on c5 looks down the whole diagonal to g1, and the white rook is standing on it. That is why the rook could not stay on g1 after : the bishop would simply have taken it, and Black takes a rook for a bishop any day. was not an attacking move, it was a move out of the way.
Out of that comes a rule of thumb for every exchange: do not ask which of your pieces is badly placed, ask which enemy piece is left standing on the square afterwards. On c5 a knight was harmless and a bishop was a permanent threat. Count that before you trade and you will trade your own advantage away less often.
IM Mai Narva (EST, 2411)
GM Alexandra Kosteniuk (SUI, 2495)
The bishop that had to stay: 32...Ba7
From move 26 on Black played the better game. drove the bishop off g5, asked it the next question on h6, then the rooks came to the d-file and the bishop came through f2 into the white position: . White was not defending in the most stubborn way; was about a pawn better than .
The winning road starts with a check: , a pair of rooks leaves the board, White’s active rook is gone, and the f5 pawn drops next move. The whole thing hangs on one small point: after the white king may not capture, because the bishop on f2 covers g1.
The game went . The bishop really was attacked, would have collected it, and retreating is a reasonable thought. It only takes the bishop off the square where it was needed: from a7 it no longer covers g1, and with the check the entire winning road disappears. came anyway, and the lead shrank from two and a half pawns to less than one. Before you pull a piece out of a threat, ask what it is currently covering. A bishop guarding one single square behind enemy lines is worth more than a bishop standing safely.
IM Mai Narva (EST, 2411)
GM Alexandra Kosteniuk (SUI, 2495)
Passed pawn for passed pawn: 42.c5
After the time control both players had time to calculate again, and White found the most practical plan of the game. , and what looks like a sacrificed pawn is a trade: White’s passed pawn disappears and so does Black’s. Three pawns against three from there on.
was compulsory rather than a preference. The bishop moves beside it would have left Black a good two pawns better, and Black could not recapture without handing the g7 pawn to the two rooks on the g-file. followed, and every exchange made the draw firmer.
The practical value of this plan lies in the order of things. A passed pawn is not an end in itself, it is leverage: it forces the opponent to leave something else standing. gives nothing away, it trades your own trump for the only one your opponent has. When you are worse, look for exactly that trade: not the one that wins material, but the one that takes the opponent’s threat off the board.
IM Mai Narva (EST, 2411)
GM Alexandra Kosteniuk (SUI, 2495)
Verdict
In the match it was not quite enough: Ghazal Hakimifard won on board three, Mariia Manko lost on board two, Camille De Seroux drew on board four, 2 to 2 against Estonia. For Alexandra Kosteniuk it is half a point made of two halves. The first was lost with , the second was saved against an opponent who stood on the brink of the win on move 32 and gave it back with ; would have offered one last try, but instead came and everything was off the board. If you take one thing from this game, take this: a plan in chess has an expiry date. was strong, and four moves later the same move was merely normal.
Moves
½-½
Was this analysis useful?