Week 13: how to beat the London System as Black, then lose with it as White

Week 13 (April 1–5, 2024) ended at my best rating so far, tying the 1051 I'd touched in week 9: 26 games, 11 wins, 14 losses, 1 draw, 1036 → 1051 (+15). The most instructive pair of games came on Friday morning against the same opponent, eleven minutes apart. First I showed how to beat the London System as Black, in 15 moves. Then I played the London myself and lost.

The week in numbers

The blunder rate was the worst so far: 72 in 26 games, close to three per game, 51 of them in the middlegame. Two games show it at its most chaotic: game #330, a win on time against micard1948 with 7 blunders of mine, and game #335, a loss to TANKPINOU with 8. With White the London went 1 win and 3 losses. With Black I faced it once and won.

Beating the London as Black: game #349

Against hosseinkameli76 (1008) I had Black, and the game went 1.d4 d5 2.Bf4 Bf5 3.e3 Nf6 4.Nf3 e6. The order matters: my light-squared bishop came out to f5 before I played ...e6, so it wasn't locked behind my own pawns. White checked with the bishop on b5, I blocked with the knight, played ...c6, and after the knights came off on d7 my queen was already developed. A few moves later the dark-squared bishops were traded on d6, my queen recaptured, and it was now pointing at h2. I castled, put my knight on g4, and when White played h3 my queen took on h2: checkmate on move 15. My accuracy was 98.4%, with no blunders and no inaccuracies at all.

Is that how to beat the London System as Black in general? One game is not a repertoire, and my opponent's h3 was a blunder (the engine marks it as allowing mate in one). But the setup did what I needed: normal development, the bishop outside the pawn chain, and nothing loose. See the game.

Eleven minutes later the colors were reversed, I played the London, and my opponent used the very same ideas: ...c6, ...d5, ...Bf5, ...e6. I got through the opening fine (94.6% accuracy) and then collapsed in the middlegame (57.8%). I lost that one.

Three positions the engine flagged

Evaluations are from White's side.

Game #327: two ways to attack the queen

With Black against micard1948 (1100), White's queen had come out to g4 early, and I was already clearly better (-3.68) on move 7. I played ...d6, which opened my bishop's line to the queen on g4. The engine preferred ...f5, attacking the queen directly with a pawn, and its line keeps Black winning. After ...d6 I was still better, but by much less (-1.50).

r . b q . r k .
p p p p . p p p
. . n . . . n .
. . b . p . . .
. . . . P . Q .
. . P P . . . P
P P . B . P P .
R N . . K B N R

The rest of the game was a mess on both sides, and I resigned on move 12 after losing a rook.

Game #328: the London move I didn't play

With White against petrsvetrr (1100), in a London System, I was slightly better (+0.95) on move 30 with eleven minutes left. I moved my queen to e2 to attack Black's rook on g4. The engine wanted the pawn break c3–c4: the pawn can be taken, but the line shows why it still works. After my queen move the engine's reply is the bishop taking my pawn on h4, slightly better for Black (-1.29). My opponent chose a queen move instead, and my next move, f2–f3, was a much bigger blunder (+0.54 to -5.02).

. . r . . . q .
p p . k . p . .
. . p . p b . .
P . . p . p . p
. P . P . . r P
. . P . P Q P .
. . . . . P N .
R . . . . R K .

I lost, with 94.6% accuracy in the opening and 75.4% in the middlegame. The same London pattern as in week 3.

Game #329: kick the queen first

With Black against pickep (995) the position was equal (+0.12). The queens faced each other on b6 and b3. I moved my queen to c6. That let White play a2–a3 and attack my bishop on b4, and the engine called it clearly better for White (+2.62). Its move was ...a4 first, attacking White's queen and keeping things level.

r . . . . r k .
. p . n . p p p
. q . . p n . .
p . . p . . . .
. b . P . B . .
. Q . B P P N .
P P . . . P . P
R . . . . R K .

I won this one in the end.

The lesson of the week

The London teaches both colors the same thing: the opening is the easy part. As Black, getting my light-squared bishop out before ...e6 and developing normally was enough against a 1000-level London. As White, the moves I kept missing were the middlegame pawn breaks, like c3–c4 in game #328, which are what give the London's quiet setup a purpose. Games #349 and #350, eleven minutes apart, show it perfectly: the same structure, the same opponent, opposite results, decided after move 12 both times.

All games this week

← Week 12: a missed mate in 3 · Week 14: eleven losses in a row → · My first London week is week 3 · The journey page · Get the same phase-by-phase review for your games