Transforming an Advantage: Giri v Niemann, Grand Swiss 2025

Part 1Samarkand 2025: a must-win game for both players

The FIDE Grand Swiss was an 11-round Swiss tournament in Samarkand, Uzbekistan, with two places in the 2026 Candidates tournament for the top finishers. Before the final round on 15 September five players shared the lead on 7/10: Matthias Bluebaum, Alireza Firouzja, Anish Giri, Hans Niemann and Vincent Keymer. Giri and Niemann were paired against each other. ^[4] ✓ Verified

A draw would probably have given neither player a Candidates place, so both had to play for a win. FIDE's report says both players were under pressure to win; Giri won in just over four hours and fifteen minutes, finished clear first on 8/11, took the $90,000 first prize and qualified for the Candidates for the third time. ^[4][5] ✓ Verified

Chess.com's report describes what Giri got from the opening: a risk-free position with the bishop pair and asymmetrical pawns, plus a half-hour lead on the clock. Giri called it a position with huge potential. The question is how potential becomes a point. ^[5] ✓ Verified

Part 2What does transforming an advantage mean?

Transforming an advantage means changing one kind of advantage into another kind that is easier to win with. A lead in development becomes an attack, an attack becomes an extra pawn, the bishop pair becomes a weak pawn to attack or a better ending. Most advantages are temporary or too small to win on their own; the skill is knowing when to cash one in, and what to buy with it.

Club players tend to hold on to whatever they have. Having the bishop pair, they refuse to trade a bishop; a pawn up, they refuse to give it back. But an advantage is a currency, not a trophy. The right question is never "how do I keep it?" but "what does it buy me if I spend it now, and is that worth more?" ◐ Inferred

In the game below the same choice came up at move 33. Niemann could have gone for opposite-coloured bishops with rooks still on; the move played instead left a pure same-coloured bishop ending, the left-hand board.

Part 3How an English Four Knights produced the bishop pair

Giri opened with the English Opening, and Niemann chose the Four Knights set-up with an early ...Bb4. After 5...Bxc3 6.Qxc3 White already owned the bishop pair, and 7.a3 prepared to gain space on the queenside with b2-b4. The Lichess study notes that Carlsen, among other top players, has played 7.a3 here. ^[3] ✓ Verified

Annotating the game for Lichess, WGM Sabina Foisor judges that 11...Be6 allowed simplification into an ending where the bishop pair gives White a pleasant edge, and suggests 14...Kxe7, keeping the knight centralised, over the game's 14...Nxe7. Stockfish agrees the edge is small: about a third of a pawn. ^[3] ✓ Verified

Then 15...Rc8 16.f3 f6 17.Kf2 Kf7 18.Rd1 Be6 19.h4 b6 20.e4 Nc6 21.Be3. With the queens off, both kings stand in the centre. White's pawns on e4, f3 and h4 take squares from the black knight, and the plan of b2-b4 to open the queenside for the bishops is still in reserve. ◐ Inferred

Part 4Two plans: open the queenside, or lock it shut

The bishop pair is strongest when the position opens, because two bishops cover both colours from a distance and a knight cannot. Giri's kingside pawns took squares from Black's knight; the queenside was where lines could open. Black's task was the reverse: keep the queenside closed, find the knight a strong square, and let the bishops look at pawns they could not reach. ◐ Inferred

Before that, Giri had spent several moves probing: 22.Rd6, then 24.Rd2 and 25.Rd6 again, and 26.Rdd1. Black's knight went back and forth between a5 and b7 in reply. Stockfish shows that 22.Rd6 let the edge shrink to close to level, and Foisor notes that 22...Rhd8 would have equalised; the probing manoeuvres that followed did not change the evaluation much. White's edge was still small when the decisive moment came. ^[3] ✓ Verified

Part 5The decisions that mattered

For the first 26 moves Stockfish's number stayed under a pawn. That is what the bishop pair is worth in a quiet ending: pressure, not a win. Watch where the line on the graph jumps, and notice that the first jump is a king move nobody's report mentioned.

The game went 28...Rcd8 29.Rxd8 Rxd8 30.b4 Nb3 31.Rc3 Rd6. ChessBase India calls 31...Rd6 the beginning of the end, and Stockfish preferred 31...Kd6, keeping the rook on the back rank. Giri then played 32.bxc5, though Stockfish's choice was 32.g5 first; both keep a clear advantage. ^[6] ✓ Verified

I had a difficult choice – I could just move the bishop or take [with the rook] on b3. My logic was that, if moving the bishop wins, it's OK. But if taking on b3 wins and I don't do it and don't qualify for the Candidates, I will never forgive myself.

Black had a different choice at move 33: 33...Rxa6 instead of 33...Bxb3, when 34.Bxc5+ leaves White a pawn up with rooks on and bishops of opposite colours. Foisor judges that with the rooks on and a passive black king Black has little hope there, and Stockfish agrees it is lost too, about three pawns for White, though less clearly than the game. ^[3] ✓ Verified

Part 6The turning point: 36.g5 and the second weakness

A pawn up with same-coloured bishops is usually winning, but not automatically. If Black can set the kingside pawns on dark squares and keep the king in front of them, the white king may have no way in. The ending still had to be transformed once more: from an extra pawn into an open road for the king. ◐ Inferred

After 36...Be6 37.Ke3 Kc5 38.Kf4 Kd4 39.gxf6 gxf6 Black's kingside is down to two pawns, f6 and h7, and the white king is in front of them. Now the last transformation: 40.a4! and 41.a5, sending the a-pawn up the board to drag Black's king away, and only then e4-e5 to make a passed pawn on the other side. ^[3] ✓ Verified

Part 7What Niemann could have done

Black's defence had two good chances, one strategic and one concrete. The strategic one was 27...Nb3, putting the knight on its best square before White's break and locking the queenside with ...c4. The concrete one was 32...Nxc5, the recapture that kept every piece safe. ^[3][4] ✓ Verified

The key point is what Black avoided: after 32...Nxc5 there is no exchange to return, because no black piece is overloaded. A transformation needs something to transform. When the defender leaves nothing loose, the attacker has to keep building, and the defender keeps hoping. ◐ Inferred

Part 8A different angle: Giri won by spending the bishop pair

The usual lesson from a game like this is "the bishop pair wins endings". Stockfish tells a more useful story. For twenty moves the bishop pair was worth less than a pawn. Giri won not by keeping it but by spending it three times: the bishops' pressure bought an open queenside, the open queenside bought an extra pawn, and the extra pawn bought an entry for the king. By move 35 the bishop pair, the advantage everyone wrote about, no longer existed. ◐ Inferred

Each transformation traded a larger but vaguer advantage for a smaller but clearer one. The bishop pair is a matter of degree; an extra pawn in a same-coloured bishop ending is a fact. That is also why the choice at move 33 mattered less than it felt to Giri: both moves were winning, and Giri took the one that left less to calculate, with the Candidates at stake. ◐ Inferred

Part 9How to transform an advantage in your own games

• What exactly is my advantage, and is it still growing?
• Which enemy piece is doing two jobs?
• What kind of ending does the trade leave: same-coloured or opposite-coloured bishops, pawns on one wing or both?
• After the trade, where is my second front?

The bishop pair itself is the subject of Kramnik v Tiviakov, and the finish of this game, a passed pawn on one wing to stretch the king while the other wing breaks, is the principle of two weaknesses. For why the opposite-coloured version so often draws, see Topalov v Shirov.

Your turnPractise it

Set each position up, decide your plan before you look, then play it out against Stockfish to see whether it holds.

Part 10What to take away

• An advantage is a currency. Giri's bishop pair was worth under a pawn until it was spent.
• Transform toward clarity. An extra pawn in a same-coloured bishop ending is easier to win than a vague edge.
• Overloaded pieces are where transformations start. The knight on b3 guarded itself and c5, and could not do both.
• Check the ending, not just the material. The same extra pawn wins with same-coloured bishops and draws with opposite colours.
• One advantage is rarely enough. 36.g5 and the a-pawn turned an extra pawn into two fronts.
• The decisive error is often quiet. 27...Ke7 lost more than the famous 32...bxc5.