London System vs e5: what the Englund Gambit taught me in week 32

Week 32 (August 15–18, 2024) brought me back to 1.d4 and straight into the reply every London player meets sooner or later: 1...e5. I played 23 rapid games, won 10, lost 13, and dropped from 985 to 960. Seven of those losses came in a row. Two were Englund Gambits, 1.d4 e5, and I lost both, so the question of London System vs e5 isn't theoretical for me. The answer from the engine is a little embarrassing: the gambit didn't beat me. I beat myself, from winning positions.

The week in numbers

The 1.e4 experiment from week 30 lasted about a week: all ten of my White games here started with 1.d4. With White I went 3–7. With Black, 7–6. My ten wins averaged 91.8 % accuracy and my thirteen losses 68.4 %, which says a lot about how my games are decided. Saturday, August 17 was the bad day: nine games, two wins, seven losses, part of a seven-game losing streak from #789 to #795.

Now the Englund games. In #789 against Jan0ch (1045), after 1.d4 e5 2.dxe5 Bc5 3.Bf4, I was +4.20 at move 18 and played 18.b4 (−3.92); the game swung back, and at move 22 I was +5.72 and played 22.Nxh4, which allowed mate in one. I was checkmated. In #801 against bobobobobobobom (972), after 1.d4 e5 2.dxe5 d6 3.exd6 Bxd6 the queens came off on moves 5 and 6, I was +5.45 at move 11, and the advantage slipped away until 18.Ne1 turned an equal position into −7.69. I abandoned that one. Seven games this week were winning positions I didn't win, and both Englunds are on that list. I also lost twice on time, in #783 with 4 seconds left and in #791.

Three key positions

The knight that would uncover my queen

Game #781, Black against Leons86 (963), in a Colle System. About equal, +0.50, with my queen on a5.

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

The only thing between White's bishop on d2 and my queen on a5 was White's own knight on c3. After 7...Be7, that knight can take on d5: 8.Nxd5 attacks my knight on f6 and my bishop on e7, and by moving it uncovers the bishop on d2 against my queen. I can't take the knight back because my queen is under attack. The evaluation went from +0.50 to +2.77. Stockfish's 7...Qc7 simply steps off that diagonal. Luckily, White played 8.Na4 instead, and I won by checkmate (game on Chess.com).

Leaving pawn tension unresolved

Game #782, Black against Jeyffoo13 (999). I was clearly better, −2.11, and White's pawn on b4 was attacking my pawn on a5.

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

9...axb4 resolves the tension in my favour: the pawn takes, and from b4 it attacks White's bishop on a3 and the pawn on c3, which has no defender. Instead I put my queen on b6, and White could take first: 10.bxa5 attacks my queen and opens the a3 bishop against my bishop on e7. −2.11 became −0.12, about equal. My opponent played 10.e3 instead, and I took on b4 a move later. I won this game too, by checkmate after 57 moves, and then lost the rematch against the same player in the very next game (game on Chess.com).

Retreating instead of trading

Game #783, White against Jeyffoo13 (1007), that rematch. Slightly better for me, +0.48, with Black's bishop on c4 facing my bishop on d3.

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

I retreated with 23.Bc2 and Black answered 23...e5. Stockfish's 23.Bxc4 trades it off: after 23...dxc4 Black is left with a pawn on c4, and in the engine's line my queen goes 24.Qa2 and 25.Qxc4 to collect it. The evaluation went from +0.48 to −1.20 after my retreat. I lost this game on time with 4 seconds on my clock (game on Chess.com).

The lesson of the week

My answer to London System vs e5, from these games, is that taking the pawn was fine: I was winning by the middlegame in both Englund Gambits. The lesson is the same one my whole summer keeps repeating, applied to winning positions: when I'm clearly ahead, every move gets a blunder check against my opponent's checks and captures, and I trade pieces when I can. In #789 one move allowed mate in one at +5.72. Positions #781 and #782 show the same thing from the other side: my opponents let me off, and I won games where I'd given back most of my advantage. At 1000, the side that blunders last loses.

All games this week

Previous: week 31 · Next: week 33 · All weeks: the journey hub · Want to know what happened in your own gambit games? Review your own games like this.