Week 44: how to get better at chess at 1000 Elo — a 13–5 week, looked at honestly
Week 44 was the bounce I needed. From November 4 to 10, 2024, I played 18 rapid games, won 13 and lost 5, and my rating climbed from 994 to 1058, with a high of 1068 along the way. I won ten of my first eleven games. If you're trying to work out how to get better at chess at 1000 Elo, this is what a good week looked like for me — and, as always, the engine found the holes in it.
The week in numbers
| What | Number |
|---|---|
| Games | 18 (#1076–#1093) |
| Wins / losses / draws | 13 / 5 / 0 |
| Average accuracy | 78.2 % |
| Blunders (my moves) | 31 (1.7 per game) |
| Rating | 994 → 1058 (+64) |
| Most played openings | Queens Pawn Opening Zukertort Chigorin Variation (3), Giuoco Piano Game (2), London System (1) |
The most surprising number is the Black score: 6 wins and 3 losses, one week after going 3–12–1 with Black. I hadn't changed my opening yet. After 1.e4 e5 I scored 5–3, with the same move I had won just one game in ten with the week before. With White I went 7–2, including three wins out of three in my usual "Zukertort Chigorin" setup.
Three of the wins were close to perfect: game 1079 against Pythagoras180 (1055), 97.6 % in a short game that my opponent resigned; game 1086 against easy_money_sniper12 (1060), 97.3 % with no inaccuracy at all; and game 1091 against 77SAD (1033), 94.6 %.
How to get better at chess at 1000 Elo: what this good week was made of
I made 31 blunders in 18 games, 1.7 per game, compared with 2.0 the week before. That's not a revolution. What changed more was what happened around the blunders: my opponents' mistakes got punished, and three of my five losses were short games I abandoned (13, 19 and 30 half-moves), not long games thrown away.
The one long game I threw away is the most instructive loss of the week, and it's the third key position below: a completely winning game lost on time.
Three key moments
Game 1077: a knight check that only gave the knight away
With Black against LeoMM189 (987), after 1.e4 e5 2.Nf3 Nc6 3.Nc3 Bc5 4.Nxe5 Nxe5 5.d4, White's pawn on d4 was attacking my bishop on c5 and my knight on e5 at the same time. I gave a check.
A check feels like it buys time, but this one bought nothing: position.py notes that "the knight can now be captured", and White just took it with the queen. The evaluation went from +0.68 to +3.52. The engine's Bd6 saves the bishop and keeps the knight defended, "slightly better for white". I won the game by checkmate anyway, because my opponent later went wrong.
Game 1078: the queen grabbing a pawn on c5
With White against MustafoevM (1015), an Old Benoni, I was a bit better (+0.99) and my queen went after Black's c5 pawn.
Qc4 attacks an undefended pawn and pins f7. But the engine's reply Be6 attacks my queen and, in position.py's words, "opens a line: the queen that has not moved yet now attacks the knight next to its king", my knight on d2. The score dropped to -0.84. The simple e4 kept White slightly better. My opponent didn't find Be6, blundered a piece a move later, and I won the game by checkmate.
Game 1081: the pawn fork I found one move late, and the clock
With Black against MajorScoot (1041), the engine had me at -5.91. White's pawn on h3 was attacking my bishop on g4, and I retreated it.
The bishop was attacked, so I saved it. The engine ignores the threat: d4 is a pawn fork, "attacks the knight on the queen's side; attacks the queen on the king's side; double attack", and it stays "black is completely winning". After Be6 the position was about equal. My opponent then played 17.Nb3, I found 17...d4 one move late, and I was winning by six pawns again. But the game didn't end on the board. On my last move I had about 5 seconds left and the engine had me at -8.18; my clock ran out and I lost.
The lesson of the week
At 1000, getting better doesn't look like brilliant moves. It looks like this week: fewer pieces given away, opponents' mistakes punished, a few clean games. The key moments add two specific habits. First, a check or an attack is only good if it costs the opponent something: Nf3+ and Qc4 both "attacked" and lost value. Second, when my piece is attacked, look for a bigger threat before retreating: d4 forked a queen and a knight while my bishop was still hanging. And one more from game 1081: a winning position is worth nothing at 5 seconds, so when I'm clearly winning in a ten-minute game, I should simplify early instead of calculating long lines.
All games this week
| Game # | Date | Colour | Opponent (rating) | Result | Opening | Accuracy | Link |
|---|---|---|---|---|---|---|---|
| 1076 | Nov 4 | Black | Letxone (1007) | Win | London System | 90.3 % | game |
| 1077 | Nov 5 | Black | LeoMM189 (987) | Win | Three Knights Opening | 78.5 % | game |
| 1078 | Nov 5 | White | MustafoevM (1015) | Win | Old Benoni Defense | 90.6 % | game |
| 1079 | Nov 5 | Black | Pythagoras180 (1055) | Win | Giuoco Piano Game | 97.6 % | game |
| 1080 | Nov 5 | Black | rasrt2222 (1040) | Win | Center Game Accepted Paulsen Attack | 89.6 % | game |
| 1081 | Nov 6 | Black | MajorScoot (1041) | Loss | Giuoco Piano Game | 82.4 % | game |
| 1082 | Nov 6 | White | KHATER525 (998) | Win | Queens Pawn Opening Zukertort Chigorin Variation | 89.1 % | game |
| 1083 | Nov 6 | White | pebrotus (1021) | Win | Queens Pawn Opening Zukertort Chigorin Variation | 83.8 % | game |
| 1084 | Nov 7 | White | yashchikhalkar (1027) | Win | Queens Pawn Opening Zukertort Chigorin Variation | 85.8 % | game |
| 1085 | Nov 7 | Black | nagasateesh (1032) | Win | Giuoco Piano Game Giuoco Pianissimo Variation | 76.3 % | game |
| 1086 | Nov 7 | White | easy_money_sniper12 (1060) | Win | Indian Game | 97.3 % | game |
| 1087 | Nov 7 | Black | ChessTicularTorsion69 (1098) | Loss | Bishops Opening | 46.9 % | game |
| 1088 | Nov 8 | White | vivredanslombre (1047) | Win | Queens Pawn Opening Accelerated London System | 52.7 % | game |
| 1089 | Nov 8 | Black | MCkid92 (1083) | Loss | Scotch Game | 64.6 % | game |
| 1090 | Nov 9 | White | twolf9403 (1047) | Loss | Englund Gambit | 62.8 % | game |
| 1091 | Nov 10 | Black | 77SAD (1033) | Win | Bishops Opening Boi Variation | 94.6 % | game |
| 1092 | Nov 10 | White | pankaj91111 (1046) | Loss | Queens Pawn Opening Zukertort Variation | 60.9 % | game |
| 1093 | Nov 10 | White | xtagir (1030) | Win | Dutch Defense | 64.1 % | game |
Previous: week 43 · Next: week 45 · The journey · Review your own games like this
Frequently asked questions
How do you get better at chess at 1000 Elo?
My week 44 suggests the gains at 1000 come from punishing mistakes and not making new ones: I won 13 of 18 games with 1.7 blunders per game, against 2.0 the week before, and my rating went from 994 to 1058.
Can you lose a completely winning position on time?
Yes. In game 1081 the engine had me at -8.18, completely winning, on my last move, and I lost on time with about 5 seconds on my clock.
Did my Black results improve after a bad week?
Yes, from 3–12–1 the week before to 6–3. After 1.e4 e5 I went 5–3, with the same opening I had scored 1–9 with a week earlier.