Week 86: how to beat the Modern Defense, as shown five times against me
Week 86 (September 15 to 21, 2025) cost me 30 points: 16 rapid games, 6 wins and 10 losses, from 1252 to 1222. As White I scored 5–4. As Black I won once in seven games, and the Modern Defense, my main answer to 1.e4, went 0–5. So this is a post about how to beat the Modern Defense, written from the losing side. The honest answer from these five games is less exciting than an opening trap: my opponents mostly developed normally and waited for me to go wrong in the middlegame, and I did.
The week in numbers
| Games | W / L / D | Rating | Accuracy | Blunders |
|---|---|---|---|---|
| 16 | 6 / 10 / 0 | 1252 → 1222 (-30) | 74.2% | 33 |
Most played openings: Modern Defense with 1 e4 (4), Queen's Pawn Opening Horwitz Defense (2), Queen's Pawn Opening Zukertort Chigorin Variation (2).
Ten losses is a lot, but the more useful number is this one: in 8 of the 10 losses, Stockfish says my best winning chance during the game was above 80%. I wasn't being crushed out of the opening. I was getting good positions and giving them back. My average opening accuracy this week was 87.1%; my middlegame accuracy was 71.5%.
How to beat the Modern Defense: five games against me
Here is what each opponent did against 1...g6, and where the game actually turned:
- Sliklas (1273), game 1678: early Bc4, Nc3, Nf3 and castling. I was winning on move 26 (-4.23; minus means Black, me, was better) and blundered twice in three moves. More on that position below.
- Tanoo111 (1243), game 1682: d4 and d5, grabbing space. My 12...b6 was an opening blunder (-2.42 to +2.51), and later, when I was winning again at -4.25, 32...h6 allowed a mate in three.
- ikomah (1277), game 1684: the Geller System with c3, Nd2 and Qb3. I was at -5.22 when 22...Qa5 threw it back to +0.30. I lost on time with 11 seconds left after my last move, at 88.1% accuracy.
- ajaycs135 (1236), game 1688: Nc3, Bc4, Nf3, d4 and then the push e4-e5. After the pawns came off, 7...Nxd4 took a pawn that wasn't free (-0.31 to +4.20), and the game was over after my 8th move.
- shamloo1995 (1212), game 1691: d3 and an early h4. Two middlegame blunders, 18...Bh5 and 20...f6 (+7.00 after it), and I was checkmated.
If you're the White player looking for a plan against the Modern at around 1200, my games say: develop, castle, keep pieces protected, and push e5 only when the tactics work, like in game 1688. If you're the Black player, like me, the lesson is harsher. The opening wasn't the problem in four of the five games.
The key positions
Against Sliklas (1273) I had a winning middlegame (-4.23). White's queen sat on g5, next to my knight on h5. Stockfish wanted 26...f6, a pawn attacking that queen, which had no defender, so White has to deal with it first. I played 26...Re7, a quiet rook move. The engine's reply for White shows what that allowed: 27.e5, a pawn attacking my queen on d6, and then 28.g4, a pawn attacking my knight on h5. My rook move had already dropped the evaluation to about equal (-0.09). In the game White played 27.g4 straight away, and two moves later my 28...g5 made it +4.20 for White.
My worst move of the week came on move 3, as White, and I won the game anyway. Against 215Raj (1259) the game went 1.d4 e6 2.Bf4 Qf6. Black's queen was now looking down the f-file at my bishop on f4. I played 3.Nf3 as if nothing had happened, and Stockfish's best answer for Black was simply 3...Qxf4, taking the bishop (the evaluation after my move was -6.73). The engine's move was 3.Bxc7: the attacked bishop escapes by capturing a pawn, and from c7 it attacks the knight on b8. My opponent replied 3...d6, missed the bishop, and I won by resignation 45 moves later, with 9 blunders of my own along the way.
The last one is quieter. Against bangjay88 (1273) the position was level (+0.36). I played 20.Rac1. Stockfish's answer for Black was 20...Ng3: the knight attacks my rook on f1 and, by leaving e4, opens the line of the bishop on f5 towards my queen on c2. That's a discovered attack, two threats at once. My opponent found it; the evaluation after my rook move was already -1.30. The engine's 20.Nh4 attacks that bishop first. I lost this one by checkmate on move 60.
The lesson of the week
A losing streak with one opening makes you want to change openings. The data says not yet. In four of my five Modern losses I was better at some point, and the games turned on middlegame blunders, not on anything my opponents prepared. What links all three positions above is simpler: a piece of mine was on a line with an enemy piece (queen and knight on g5 and h5, bishop on f4 and queen on f6, queen on c2 and bishop on f5), and I didn't look at that line. Before each move, I now try to check which of my pieces sit on the same file, rank or diagonal as an enemy piece.
All games this week
| # | Date | Colour | Opponent | Result | Opening | Accuracy | |
|---|---|---|---|---|---|---|---|
| 1678 | 2025-09-15 | Black | Sliklas (1273) | Loss | Modern Defense with 1 e4 | 59.0% | game |
| 1679 | 2025-09-16 | White | 215Raj (1259) | Win | Queen's Pawn Opening Horwitz Defense | 64.8% | game |
| 1680 | 2025-09-16 | White | bangjay88 (1273) | Loss | Queen's Pawn Opening Zukertort Chigorin Variation | 88.7% | game |
| 1681 | 2025-09-17 | White | iwan-sopian (1253) | Win | Indian Game | 95.6% | game |
| 1682 | 2025-09-17 | Black | Tanoo111 (1243) | Loss | Modern Defense with 1 e4 | 55.7% | game |
| 1683 | 2025-09-18 | White | yasermoein2 (1258) | Win | Englund Gambit | 85.9% | game |
| 1684 | 2025-09-18 | Black | ikomah (1277) | Loss | Modern Defense Geller System | 88.1% | game |
| 1685 | 2025-09-19 | White | camelo1455 (1232) | Loss | Queen's Pawn Opening Horwitz Defense | 60.0% | game |
| 1686 | 2025-09-19 | Black | Lzetoo (1253) | Loss | Nimzowitsch Defense Declined Williams Variation | 68.1% | game |
| 1687 | 2025-09-19 | White | gogetsomesleepbro (1202) | Win | Queen's Pawn Opening Zukertort Chigorin Variation | 95.3% | game |
| 1688 | 2025-09-19 | Black | ajaycs135 (1236) | Loss | Modern Defense with 1 e4 | 70.2% | game |
| 1689 | 2025-09-20 | White | Ghostking226 (1219) | Loss | Indian Game | 56.5% | game |
| 1690 | 2025-09-20 | White | razornjaja (1245) | Loss | Queen's Pawn Opening Zukertort Variation | 56.0% | game |
| 1691 | 2025-09-21 | Black | shamloo1995 (1212) | Loss | Modern Defense with 1 e4 | 69.8% | game |
| 1692 | 2025-09-21 | Black | Soni1253 (1188) | Win | Nimzowitsch Defense Declined Williams Variation | 78.3% | game |
| 1693 | 2025-09-21 | White | GeloidasPapas (1233) | Win | Indian Game East Indian London System | 95.5% | game |
Previous week: week 85. Next week: week 87. The whole diary is on the journey page. For how these reviews are produced, see can Claude analyze chess games?.
Frequently asked questions
How do you beat the Modern Defense at club level?
Going by my five losses with it this week, not with a trick. Four of my opponents simply developed and waited; in each of those games Stockfish says I was better at some point and then blundered in the middlegame. Only one game was decided early, after White pushed e4-e5 and I grabbed a pawn on d4 that wasn't free.
Which setups did my opponents use against 1...g6?
Two played an early Bc4 with Nc3 and Nf3, one gained space with d4 and d5, one used the Geller System with c3 and Nd2, and one played d3 with an early h4.
What was the free bishop I missed as White?
In game 1679, after 1.d4 e6 2.Bf4 Qf6, Black's queen attacked my bishop on f4. Stockfish wanted 3.Bxc7, saving it by capturing a pawn. I played 3.Nf3, which allowed 3...Qxf4, but my opponent missed it and I won.