# Game #216: a London System kingside attack against a 1074

> Game #216: I beat angelmariabecerrilandres (1074) with a London System kingside attack. Black's king lost castling early; I nearly let it escape twice.

- URL: https://thechesscoach.netlify.app/journey/upset-game-216/
- Updated: 2026-10-26
- Author: Ignacio Soler

On February 26, 2024, in game #216 of my rapid journey, I beat **angelmariabecerrilandres (1074)** with White while rated around 912. It was a London System kingside attack in the most literal sense: my knight landed on g6 on move 10, Black's king lost the right to castle three moves later, and I spent the rest of the game chasing it. My rating went from 912 to **924**. The gap of 150 points makes it one of the ten biggest upsets in my first 2,116 games.

It was also the cleanest of my early upsets. Stockfish counted **one blunder and one mistake** for me, against three blunders and one mistake for Black, and my accuracy was **90.7** to Black's 85.4. I still came close to throwing it away, and that part is the reason I'm writing it up.

## Where I was that day

This was the first day of [week 8 of the journey](/journey/week-2024-w09/), the week I touched 1000 for the first time. It didn't start well. Games #211 to #214 were four losses in a row, two of them in the same Queen's Pawn Opening line, and I'd dropped to 903. Game #215 was a win against Laid_Akermi (947). Then came this one.

| | |
|---|---|
| Game | #216, February 26, 2024 |
| Opponent | angelmariabecerrilandres (1074) |
| My colour, rating after | White, 924 (from 912) |
| Time control | 20 minutes, no increment |
| Opening | Indian Game (A45) by Chess.com's name; a London setup on my side |
| Result | Win, Black resigned after my 35th move |
| Accuracy (Stockfish) | Me 90.7, Black 85.4 |
| My blunders / mistakes / inaccuracies | 1 / 1 / 2 |

Chess.com files the game as the Indian Game because of the move order, but what I played was my usual London: pawn to d4, bishop to f4, knights to f3 and d2, pawns to e3 and c3. Black played a slow setup with a6 and h6, then traded dark-squared bishops on e5, and my knight took back on e5 on move 8.

## Moment 1: the knight jump that started the attack

On move 9 Black pushed the pawn to f6 to kick my knight from e5. Stockfish called that a blunder. Taking my knight with the knight on d7 kept things about equal; the pawn push gave me a jump that I actually found.

> **Game #216, move 10: Black has just pushed the f-pawn to kick my knight**
> Position (FEN): `rnbqk2r/1ppn2p1/p3pp1p/3pN3/3P4/2P1PN2/PP3PPP/R2QKB1R w KQkq - 0 10`
> Played: Ng6 — Ng6, the knight goes next to Black's king and attacks the rook on h8
> Better: Ng6 — Ng6, the engine's choice too

The knight on g6 attacked the rook in the corner, which had no defender. Black moved the rook to g8. A few moves later I took Black's knight on f8 with my knight, and Black recaptured with the king. That's the detail that shaped the whole game: a king that has moved can never castle. Black's king stayed on f8, in front of a kingside that had already pushed the f- and h-pawns. Black then pushed the e-pawn and the h-pawn (the second push was a mistake, by the engine's count), my pawn captured on e5 and then on f6, Black took back with the g-pawn, and the evaluation stood at **+3.95** for me (evaluations are from White's side).

## Moment 2: the knight that went the wrong way

On move 17 I had about fifteen and a half minutes on my clock against Black's eighteen, and I spent almost two of them on this move. I still didn't find the plan.

> **Game #216, White to move: Black's king is stuck on f8**
> Position (FEN): `r1bq1kr1/1pp5/p1n2p2/3p3p/8/2PBPN2/PP3PPP/R2Q1RK1 w - - 0 17`
> Played: Nd4 — Nd4 attacks the knight on c6 and offers a trade
> Better: Nh4 — Nh4, the knight heads towards Black's king

I offered a knight trade in the middle of the board. Stockfish wanted the knight on h4, closer to Black's king, and its line keeps me winning (+3.84 in a fresh check). After my move the evaluation fell to **+0.91**. When I checked the position for this post, the engine's best answers for Black show why: take the trade on d4, or put the bishop on h3, next to my king, where it attacks my g2 pawn with my rook standing behind it. The tool calls that a double attack (one move that threatens two things at once), right in front of my castled king.

In other words, the attack was running in one direction and I helped my opponent start one in the other.

Black didn't take the trade. They played the bishop to g4, attacking my queen, which the engine also marked as a blunder (+3.46 for me again). I answered with the pawn to f3, which blocks the bishop's line to my queen and attacks it. The bishop went to h3 anyway, I guarded g2 with my rook, and on the next move the knights were traded.

## Moment 3: the right move, in the wrong order

By move 21 I was back to +4.80. Black's queen had come to d6 and the bishop was still on h3.

> **Game #216, White to move: Black's bishop sits on h3, next to my king**
> Position (FEN): `r4kr1/1pp5/p2q1p2/3p3p/3P4/2PB1P1b/PP3RPP/R2Q2K1 w - - 1 21`
> Played: Kh1 — Kh1, the king steps into the corner
> Better: Qd2 — Qd2, the queen comes out first

I tucked my king into the corner. Stockfish called that a mistake and wanted the queen out to d2 first; in its line my king goes to h1 a couple of moves later anyway. Same moves, different order, and the evaluation went from +4.80 to +2.53. I don't fully understand that one yet, and I'd rather say so than invent a reason. What I take from it is that "I'll play that move at some point" isn't the same as "that move is good now".

## Moment 4: checks, and the move I didn't need to find

The next move I played was the queen to d2 anyway, and then the attack ran on its own. My queen checked from h6, my rook checked from e2, and on move 24 Black covered the check with the bishop on e6. Stockfish called that the losing blunder: walking the king to d8 kept Black at +3.14; the bishop block made it +7.70.

> **Game #216, White to move: Black's king on e7, bishop on e6**
> Position (FEN): `4r1r1/1pp1k3/p2qbp1Q/3p3p/3P4/2PB1P2/PP2R1PP/R6K w - - 9 25`
> Played: Rae1 — Rae1, the second rook joins the attack on the e-file
> Better: Bg6 — Bg6, the bishop attacks the pawn on h5 and the rook on e8 at once

The engine preferred the bishop to g6, a double attack on the h5 pawn and the rook on e8. I doubled my rooks instead. That was still completely winning (+6.82), so I don't count it as a mistake; it's just a reminder that when the opponent's king is open, the strongest move is often a bishop or knight landing next to it, not another heavy piece lining up behind.

From there my queen took the f6 pawn with check, my bishop took Black's bishop on e6, one pair of rooks came off on e2, and on move 33 my rook took Black's last rook. On move 35 my rook checked from g7, uncovering a check from my queen on h4 at the same time. Black resigned. In a fresh check, the engine says a rook check from d7 instead would have been checkmate in two; the move I played was a forced mate in three. I had 5:16 left on my clock, Black had 9:46. You can [replay the full game on Chess.com](https://www.chess.com/game/live/102703209931).

## What the engine says about this win

My accuracy was 93.5 in the opening and 89.3 in the middlegame; the game never reached an endgame. Both of my real errors were in the middlegame and both were about direction: a knight that went to the center instead of towards the king, and a king move played before the queen move. Black's three blunders were all about their own king's safety: the pawn push to f6 that let my knight in, the bishop that went to g4 instead of trading, and the bishop block on e6.

For once I don't think the result lied. I made fewer big mistakes, and the ones I made didn't cost the game.

## The lesson I took from game #216

When the opponent's king can't castle, every piece I move should be asking one question: am I getting closer to that king, or further away? The knight to d4 failed that test. It wasn't a bad square in general; it just went to the wrong side of the board, and it handed Black a trade that pointed a bishop at my own king.

The second lesson is smaller: a move that belongs in the plan can still be wrong right now. Stockfish wanted the same two moves I played, queen to d2 and king to h1, just the other way round.

## What came next

Right after this game we played again with colours reversed, and angelmariabecerrilandres (1079) beat me: game #217, the one where I walked my king forward when I was a piece up. That week ended at 982, after my first, brief visit to 1000, which is its own story in [reaching 1000](/journey/reaching-1000/). The whole week is in [week 8 of the diary](/journey/week-2024-w09/), and the full series is on [the journey page](/journey/).

If you want your own games read this way, with the engine pointing at the moment and plain words for why, that's what I built Chess Coach for: here's [how the analysis works](/features/ai-chess-analysis/), and why [Chess.com's review doesn't always explain the moves](/answers/chess-com-game-review-not-explaining-moves/).

## FAQ

### Can you attack the king with the London System?

In this game, yes. My knight jumped to g6 on move 10, took Black's knight on f8 three moves later, and Black had to recapture with the king, so it could never castle. From there my queen, rooks and bishop went after that king until Black resigned on move 35.

### What was the biggest mistake in this London System kingside attack?

Move 17: I offered a knight trade on d4 instead of bringing the knight to h4, towards Black's king. Stockfish's evaluation dropped from +3.95 to +0.91. Black answered with a blunder of their own, and the attack came back.

### How did a 924 beat a 1074 here?

With fewer big mistakes. By Stockfish's count I made one blunder and one mistake; Black made three blunders and one mistake. My accuracy was 90.7 against 85.4.

## Related

- [From 631 to 1385: my chess journey in 2,117 games](https://thechesscoach.netlify.app/journey.md)
- [Week 8 (Feb 2024): touching 1000 Elo for the first time](https://thechesscoach.netlify.app/journey/week-2024-w09.md)
- [Reaching 1000: how many games it took me (game #232)](https://thechesscoach.netlify.app/journey/reaching-1000.md)
- [Free AI chess analysis of your games, explained](https://thechesscoach.netlify.app/features/ai-chess-analysis.md)
- [Chess.com Game Review not explaining moves? Here's why](https://thechesscoach.netlify.app/answers/chess-com-game-review-not-explaining-moves.md)