Week 60: how to play the Modern Defense as Black, and how I didn't
March 3 to 9, 2025: 9 rapid games, 4 wins and 5 losses, rating 1259 → 1251 (−8). The split by colour tells the story. With White I went 3–1, including two very clean wins at 96.2% and 94.7% accuracy. With Black I went 1–4: three games in the Modern Defense and two in the Pirc. So this week's question is the one I should have studied months earlier: how to play the Modern Defense as Black, or at least how my games say not to.
The week in numbers
| Rapid games | 9 |
| Wins / losses / draws | 4 / 5 / 0 |
| Rating | 1259 → 1251 (−8) |
| Record with White / Black | 3–1 / 1–4 |
| Average accuracy | 76.6% |
| Blunders (mine) | 20 |
| Most played openings | Modern Defense with 1 e4 (3), Queen's Pawn Opening Zukertort Variation (1), Pirc Defense (1) |
Two of my Black losses were games I had been winning: game 1365 (peak winning chance 91.9%) and game 1371 (90.0%). None of my 20 blunders came with less than a minute on the clock, so this was not a clock week. It was a bishop week, as the positions show.
My Modern setup is simple: 1...g6, then ...Bg7, ...d6 and castling. The bishop on g7 is the point of the whole thing: it looks down the long diagonal towards White's queenside. White players seemed to know that too. In two of the three Modern games (1365 and 1366), White's dark-squared bishop came to h6 to trade it off.
Game 1365: the capture on b2
With Black against mhmddhshan (1269) (game on Chess.com), White had played an early queen and bishop attack: 2.Qf3, 3.Bc4, 8.Bh6. After 9.Qe3 the engine rated the position −6.60, completely winning for me.
The engine line is 9...Bxb2 10.Bxf8 Qxf8 11.O-O Bxa1: my bishop takes the pawn on b2, White takes my rook on f8, and my bishop then takes the rook on a1. I come out a whole bishop and a pawn ahead. Instead I played Ne5, attacking White's bishop, and after 10.Bxg7 Kxg7 the bishop I needed was gone. −6.60 became −0.20. I was still better after that (−6.07 before move 14), blundered four more times, and 23...Rd8 (−0.75 to +8.37) decided it; the game ended as abandoned on move 29.
Game 1366: taking back with the wrong piece
The next game, with Black against Santiago0020 (1244) (game link), I was clearly better, −1.77. White had castled long, with the queen on d2 and the rook on d1 behind it.
Taking with the queen attacks White's queen, with the rook behind it, and the knight on c3, and offers a queen trade. I took with the knight, and the engine's answer is Bxg7: White's bishop takes mine, attacking my knight on f6 with the knight on d4 behind it, and my rook on f8. −1.77 became +5.57. Here the trade on g7 hurt again. My opponent played 13.Bxg7 Kxg7 14.Ne4, I took a pawn with 14...Nxf3, and later on I won by resignation, but the engine says that was a gift.
Game 1368: the king walks towards the attack
My Pirc game against Ozkilos (1277) (game) shows the other plan White uses against my g6 setups: castle long and push the h-pawn. After 10.O-O-O and 13.h4, the position was +0.92 for White.
Kg7 puts my king one step closer to the open lines. The engine's reply is 14.h5, a pawn move close to my king, and the evaluation goes from +0.92 to +4.00. The engine wanted Nd7, simply developing. The game went 14.h5 h6 15.Qc3+ f6 16.hxg6 hxg5 17.Rh7+, and with 17...Kg8 I allowed mate on the next moves (the engine's evaluation went from −2.67, better for me, to +16.99).
The lesson of the week
In all three games the key piece was the same: my bishop on g7. In 1365 it could have won a rook and I let it be traded instead. In 1366 a careless recapture let White trade it with gain. In 1368 that bishop had left g7 to grab pawns on d4 and f2, and my king stepped onto g7 just as White's h-pawn arrived. If I want to play the Modern as Black, the rule from this week is: when White's bishop or queen lines up against g7 or h6, check first what my bishop on g7 can take, then decide whether to allow the trade. Counting all my Modern games as Black up to here, I was 10–15 in 25 games. Using the bishop before it disappears is the cheapest improvement I can see.
All games this week
| Game # | Date | Colour | Opponent (rating) | Result | Opening | Accuracy | Link |
|---|---|---|---|---|---|---|---|
| 1365 | Mar 3 | Black | mhmddhshan (1269) | Loss (abandoned) | Modern Defense with 1 e4 | 64.1 | game |
| 1366 | Mar 4 | Black | Santiago0020 (1244) | Win (resignation) | Modern Defense with 1 e4 | 66.7 | game |
| 1367 | Mar 5 | White | naveen5513 (1270) | Win (abandoned) | Queen's Pawn Opening Zukertort Variation | 96.2 | game |
| 1368 | Mar 6 | Black | Ozkilos (1277) | Loss (checkmate) | Pirc Defense | 57.4 | game |
| 1369 | Mar 6 | Black | justinjdog101 (1259) | Loss (abandoned) | Pirc Defense Classical Variation | 81.1 | game |
| 1370 | Mar 6 | White | ShockUUuu (1251) | Win (checkmate) | Queen's Pawn Opening Accelerated London System | 94.7 | game |
| 1371 | Mar 6 | Black | ISMAIL-KMS (1287) | Loss (abandoned) | Modern Defense with 1 e4 | 79.1 | game |
| 1372 | Mar 8 | White | Maxim97878 (1253) | Win (resignation) | Queen's Pawn Opening Zukertort Chigorin Variation | 77.2 | game |
| 1373 | Mar 9 | White | cade888 (1270) | Loss (checkmate) | Dutch Defense | 72.6 | game |
Previous week: week 59. Next week: week 61. Every post in the series: journey hub. The engine notes on this page come from the kind of review Chess Coach does for your games.
Frequently asked questions
How do you play the Modern Defense as Black?
In my games the setup was 1...g6, ...Bg7, ...d6 and castling. The engine's best moves often used the bishop on g7: in game 1365 it wanted 9...Bxb2, taking a pawn and attacking the rook on a1, which would have kept a winning position.
What did White do against your Modern Defense?
In two of my three Modern games that week, White brought a bishop to h6 to trade off my bishop on g7: 8.Bh6 in game 1365 and 11.Bh6 in game 1366. In my Pirc loss, White castled long and pushed the h-pawn at my king.
What was your record with the Modern Defense up to then?
Counting every Modern Defense game I had played as Black up to the end of this week: 25 games, 10 wins and 15 losses.