Week 81: three games, two London System Zukertort Variation wins
Week 81 was short: three rapid games between August 13 and 17, 2025, two wins and one loss, and my rating went from 1200 to 1208. Both wins came from the same setup, the one people search for as the London System Zukertort Variation: pawn to d4, bishop to f4, knight to f3, pawn to e3. Chess.com filed one game as "Queen's Pawn Opening: Zukertort Variation" and the other as "Zukertort Chigorin Variation", but to me they were London games. The loss came as Black, in the Modern Defense, in a position the engine says I was winning.
The week in numbers
| Games | 3 |
| Wins / losses / draws | 2 / 1 / 0 |
| Rating | 1200 → 1208 (+8) |
| Average accuracy | 86.1% |
| Blunders | 2 |
| Most played openings | Queen's Pawn Opening Zukertort Variation (1), Queen's Pawn Opening Zukertort Chigorin Variation (1), Modern Defense with 1 e4 (1) |
Three games is a small sample, so I won't pretend this week proves anything. What the numbers do show is a gap I keep seeing in my reviews: my opening accuracy across the three games averaged 90.8%, while my middlegame accuracy dropped to 80.7%. My two blunders of the week came in the same game, and both after move 20.
Two London games, two small lessons
The London System is popular at club level because you can play the same moves against almost anything. That's exactly how I use it. Still, the engine found something to complain about in both wins, and both complaints were about the order of my moves rather than the setup itself.
Against Valtakom (1188) I played my automatic 5.Nbd2. Stockfish rated it a mistake, from +0.55 to -0.55, and preferred 5.c4. The reason is a little square I never look at: b4. Black's knight on c6 could jump there and attack both c2 and a2, and that's exactly what happened. My opponent played 5...Nb4 and, after 6.Rc1, took my a2 pawn. With the pawn on c4 instead, Black's position looks different, and the engine's line simply continues with normal development. I still won the game, which ended after 20 of my moves when my opponent abandoned it, but the pawn on a2 was a free gift I gave away in the opening.
In the other London game, against martinsartabgi (1172), Black had pushed the g- and h-pawns early and I was better (+1.42). I gave a check with 12.Qa4+, which the engine marked as an inaccuracy (+0.67). Its choice was 12.Be2, and its line castles on the next move. The game ended right after my check, so the inaccuracy didn't cost me anything. But it's a habit worth naming: when the position is good and my king is still in the centre, a check feels active, while the bishop move that prepares castling feels boring. The engine prefers the boring one.
The Modern Defense game I should have won
As Black against Mihail139270 (1243), the engine had me at -3.82 (a minus number means Black, me, was better). White had no queen left on the board. My move, 22...Qh6, looked clever: it pinned White's rook on d2 to the king on c1, so the rook couldn't move without exposing the king. But it gave White time. The knight jumped in with 23.Nf6+, then 24.Ng4+, and the evaluation went to +0.67, roughly equal.
The engine's move was one I didn't even consider: 22...Qxb2+, taking the bishop with check. The king can take my queen, and the line shows the follow-up: my d-pawn moves to d5 and attacks the bishop on c4, and the evaluation still says Black is winning. I won't pretend I understand every detail of that line. What I do understand is that I never looked at it, because "my queen can be captured there" ended the thought before it started. Eight moves later I slipped again (30...Nf4) into a lost position, and the game went down as a loss.
The lesson of the week
Look at every check and every capture, even the ones that seem to lose material. My winning move this week was a capture with check that gave the queen back for a bishop, and I filtered it out in a second. In the London games the lesson was the opposite one: when nothing is happening, the quiet developing move (c4 on move 5, Be2 on move 12) beats the active-looking one. Both come down to the same thing. I decide what a move does before I check what it actually does.
All games this week
| # | Date | Colour | Opponent | Result | Opening | Accuracy | |
|---|---|---|---|---|---|---|---|
| 1630 | 2025-08-13 | White | martinsartabgi (1172) | Win | Queen's Pawn Opening Zukertort Variation | 92.1% | game |
| 1631 | 2025-08-17 | White | Valtakom (1188) | Win | Queen's Pawn Opening Zukertort Chigorin Variation | 92.9% | game |
| 1632 | 2025-08-17 | Black | Mihail139270 (1243) | Loss | Modern Defense with 1 e4 | 73.4% | game |
Previous week: week 80. Next week: week 82. The whole story starts on the journey page, and if you want your own games explained move by move, you can review your own games like this.
Frequently asked questions
Is the London System the same as the Zukertort Variation?
Not exactly. Chess.com named both of my London-style games this week (d4, Bf4, Nf3, e3) Queen's Pawn Opening: Zukertort Variation or Zukertort Chigorin Variation. The names follow the move order; the plans I played were London plans.
Did the London System work for me at 1200?
This week, yes: two games with it as White, two wins, at 92.1% and 92.9% accuracy. The only loss came as Black in the Modern Defense.
Why did Stockfish dislike 5.Nbd2 in my London game?
After 5.Nbd2 Black's knight could jump to b4 and attack c2 and a2, and in the game it took the a2 pawn. Stockfish preferred 5.c4, which it rates as about equal.