GM Nico Georgiadis (SUI, 2535) vs IM Emmanuel Jimenez Garcia (CRC, 2380) · 1-0
Board 2, Switzerland vs Costa Rica · Sicilian Defence, Accelerated Dragon, Maróczy Bind (B38)
By Professor Zugzwang AI Coach . This analysis was written by a Bogner Chess AI coach.
After the defeat against Azerbaijan, Switzerland needed a win on one of the top boards against Costa Rica, and Nico Georgiadis had White against an International Master rated 2380. Against the Accelerated Dragon he chose the Maróczy Bind with : plenty of space, little tactics, and a position in which Black has to ask for every square.
The game is a lesson about one bishop. With White traded off the bishop on g7, the piece Black can least afford to miss in the Dragon, and the black king was left standing on bare dark squares. Nineteen moves later the white queen arrived on e6 and Black resigned.
Trading the Dragon bishop: 12.Bh6
In the Dragon the bishop on g7 is half the concept. It holds h8 to a1 together, it shields the king, and it is the reason Black can live with less space against the Maróczy Bind at all. With and White sent the knight back from g4 and put the bishop on the c1 to h6 diagonal, and forces the trade: the g7 bishop cannot step aside without leaving h6 hanging.
After the queen recapture is the only sensible move, and it is also the best one. The queen is not beautiful on h6, she is useful: she looks at f8 and h7, she keeps Black from castling short, and she comes back into the game later via e3 and e6. That route decided the game.
For your own practice: against the Dragon and the Accelerated Dragon, trading the g7 bishop is nearly always a success, even when it costs you two tempi. And the other way round, with Black: when your opponent sets up Be3 and Qd2, Bh6 is the threat you need a plan against before it arrives.
IM Emmanuel Jimenez Garcia (CRC, 2380)
GM Nico Georgiadis (SUI, 2535)
A retreat that attacks: 16.Bd1
With and Black has chased the knight from b3 and taken space on the queenside. The a4 pawn is the price: it stands far up the board and nothing but the a8 rook defends it. A quarter of players castle here, another quarter put the bishop on d3, and both are respectable.
is better, by roughly a fifth of a pawn, which is not much in this position but is everything in terms of the idea. The bishop goes back to the first rank and in doing so points at the a4 pawn through c2 and b3. White does not win it at once, he pins it down, and seven moves later the bishop really does land there with .
Retreats have a bad name because they look like lost time. You only lose time if the piece does less afterwards than before. On every retreat, ask what the piece aims at from its new square: a bishop that steps back and takes aim at a pawn has not lost a tempo, it has been given a job.
IM Emmanuel Jimenez Garcia (CRC, 2380)
GM Nico Georgiadis (SUI, 2535)
Hunting the king: 24.Bxd7+
Black has invested three moves in a raid on the white pawns with , and , and forgotten two things: his king is still on e8, because the white queen on h6 kept him from castling for nine moves, and that same queen now stands on e3, on the king's file. The queen on b2 and the knight on a3 are at the other end of the board.
is the move that joins the two facts. It removes the last defender and pulls the king onto a square where it is in the way: , and the rook has a clear road. Only a good third of players capture here, the others look for something bigger, and the position only gets slower for it.
The lesson is a counting exercise, not a combination. When your opponent works on your queenside with queen and knight, count how many of his pieces are still near his own king. If there is only one, trading that one piece is usually stronger than any defence of your pawns.
IM Emmanuel Jimenez Garcia (CRC, 2380)
GM Nico Georgiadis (SUI, 2535)
The only rook move: 26.Rb3
The rook on c3 has to move, and exactly one move keeps the game won: . It saves the rook and attacks the queen at the same time, and that single tempo is the whole difference. White is a good four pawns better afterwards.
The obvious alternatives throw all of it away. After the queens come off, the knight sits on c2 in the middle of White's position, and a winning game has become a level one. ends the same way: Black is back in the game and White has nothing left.
As soon as a piece is attacked we look for a safe square for it and stop thinking there. Look instead for the safe square from which the piece attacks something itself. In positions where the opponent is one tempo away from counterplay, that is regularly not one good move among several, it is the only one.
IM Emmanuel Jimenez Garcia (CRC, 2380)
GM Nico Georgiadis (SUI, 2535)
Not the pawn, the king: 27.e5
Ninety-four out of a hundred players take the pawn here with , because it falls with check and because material is countable. That move wins too. It only wins more slowly than , and slowness is the one thing that can still go wrong against a king in the centre.
The reason stands on d7. The black king has not one quiet square, and builds the road he is hunted down: the pawn goes to e6, the e-file opens, and the queen travels via e3 to e6. The game went , and Black resigned: after the queen is gone.
Remember the order, not the moves. When the enemy king is in the centre and you are clearly better, the pawn that happens to be hanging is the slowest winning attempt on the board. Count the squares the king has left first; only when there are enough of them do you count material.
IM Emmanuel Jimenez Garcia (CRC, 2380)
GM Nico Georgiadis (SUI, 2535)
Verdict
The cleanest Swiss win of this olympiad so far, and it needs no combination at all. Nico Georgiadis built the Maróczy Bind, traded off the bishop that holds the black king together with , froze the a4 pawn with and refused to be distracted by the black queen raid on the queenside. When precision was required, it was there: was the only move that kept the win, and was worth more than the pawn almost everyone would have taken. With this point on board two Switzerland won the match against Costa Rica 2.5 to 1.5.
Moves
1-0
Was this analysis useful?