Week 10: how to get better at chess at 1000 Elo, from a quiet 9-game week
Week 10 (March 11–17, 2024) was the quietest so far: just 9 games, 4 wins and 5 losses, rating 1023 → 1022 (-1). My accuracy was high (82.1%) and, for once, I didn't lose a single game from a winning position. The whole week splits cleanly in two: 4 wins and 1 loss with White, 0 wins and 4 losses with Black. That made it a good small sample for a question I kept asking myself: how do you get better at chess at 1000 Elo when you're not obviously blundering everything?
The week in numbers
| Games | 9 (#292–#300) |
| Rating | 1023 → 1022 (-1) |
| Wins / losses / draws | 4 / 5 / 0 |
| As White | 4 W, 1 L |
| As Black | 0 W, 4 L |
| Average accuracy (Stockfish) | 82.1% |
| Blunders | 15 |
| Blunders per game | 1.7 |
| Time control | 20 min (7), 10 min (2) |
| Most played openings | Queen's Pawn Opening Horwitz Defense (2); Old Benoni Defense (1); Indian Game (1); Scotch Game (1); Giuoco Piano Game (1) |
With so few games, the individual results say more than the averages. The White games were mostly clean: three wins above 87% accuracy. The four Black losses were each different:
- Game #294 (Scotch Game, vs Jag_97470, 1095): lost in 17 half-moves after a blunder on move 5.
- Game #296 (Giuoco Piano, vs bradg1082, 1100): my knight went to h5 on move 7; after a short exchange White's queen took it and I was a pawn down (+1.72).
- Game #297 (Ruy Lopez, a 10-minute game vs Kingofsobradinhocity, 1009): 92.5% accuracy in the opening, 45.3% in the middlegame.
- Game #300 (King's Fianchetto, vs 7osny_Salah67, 1129): no blunders at all, and I abandoned it after 31 half-moves.
So the answer to "how do I get better at 1000?" was not "blunder less" in general. It was: my Black openings were costing me games before the middlegame even started.
Three moments from nine games
Evaluations are from White's side.
Game #292: a queen move instead of a direct attack
With White against ahmedellethy1577 (1089), in an Old Benoni, I was winning (+5.36). I moved my queen to a4, attacking the knight on c6. The engine preferred bishop to h4, attacking the black queen on d8, which keeps the full advantage. After my queen move I was still clearly better (+2.58).
What happened next is a little embarrassing. Black's reply allowed checkmate in one. I didn't see it and played the bishop move instead. Black's next move allowed mate in one again, and this time I found it: knight to d6, checkmate on move 16.
Game #293: taking the wrong pawn
Against Dankey202 (1025), White again, the position was slightly in my favor (+1.02). Two black pawns had advanced on my queen's side, on a4 and c4, and both could be captured. I took the a4 pawn with my rook. The engine wanted the other capture, b3 takes c4, which keeps my edge. After my rook capture the engine's reply is the knight jump to e4, and Black is slightly better (-1.29).
This was my only White loss of the week; I resigned after 50 half-moves.
Game #294: the f-pawn and the h5 check, again
With Black against Jag_97470 (1095) I had played my f-pawn to f6 on move 3. On move 5 the position was still close to level (+0.73), but my king's diagonal from h5 to e8 was open, because the pawn on f7 had left. I retreated my knight to c6. White's queen jumped to h5 with check, and the engine's line wins my rook in the corner: +5.30.
That's exactly what happened in the game: queen check, the knight took on g6, and White's queen took my rook on h8. I resigned on move 9. The engine's move, f5, attacks the pawn on e4 and opens my queen's diagonal to White's knight on h4, and keeps the game close to level.
The lesson of the week
At 1000, getting better can mean something narrow and boring: fix the one part of your game that keeps losing. For me that week it was the first eight moves with Black. Game #294 was the same mistake as game #68 in week 3: an early f-pawn move that opens the diagonal from h5 to my king. So my rule with Black became simple: don't move the f-pawn in the opening unless I've checked Qh5+ first. A rule like that won't make me a better player overall, but it removes a whole category of 17-move losses.
All games this week
| # | Date | Colour | Opponent (rating) | Result | Opening | Accuracy | Link |
|---|---|---|---|---|---|---|---|
| 292 | Mar 11 | White | ahmedellethy1577 (1089) | Win | Old Benoni Defense | 83.5 | game |
| 293 | Mar 11 | White | Dankey202 (1025) | Loss | Indian Game | 81.5 | game |
| 294 | Mar 11 | Black | Jag_97470 (1095) | Loss | Scotch Game | 69.9 | game |
| 295 | Mar 12 | White | caxam (1017) | Win | Queen's Pawn Opening Horwitz Defense | 92.8 | game |
| 296 | Mar 12 | Black | bradg1082 (1100) | Loss | Giuoco Piano Game | 80.4 | game |
| 297 | Mar 13 | Black | Kingofsobradinhocity (1009) | Loss | Ruy Lopez Opening Classical Central Variation | 67.7 | game |
| 298 | Mar 14 | White | VisirChess (987) | Win | Queen's Pawn Opening Zukertort Variation | 90.3 | game |
| 299 | Mar 15 | White | diane864 (1013) | Win | Queen's Pawn Opening Horwitz Defense | 87.2 | game |
| 300 | Mar 17 | Black | 7osny_Salah67 (1129) | Loss | King's Fianchetto Opening | 85.5 | game |
Previous week: week 9, the best openings at 1000. Following week: week 11, back under 1000. The series index is on the journey page. If you'd like a coach that spots your repeated patterns, like my f-pawn habit, here's what Chess Coach does.
Frequently asked questions
How do you get better at chess at 1000 Elo?
In my 9 games at 1000 the problem was specific: 4–1 with White, 0–4 with Black, and two of those Black losses were decided in the first eight moves. Fixing the openings I play as Black mattered more than general study.
Why do I lose so quickly with Black?
In game #294 I had pushed my f-pawn early, which opened the diagonal from h5 to my king. One knight move later White's queen checked on h5 and won my rook. It was the same pattern as game #68 in week 3.