Week 124: stuck at 1200 Elo in chess, with 83% accuracy
Five games from June 30 to July 4, 2026, one win and four losses, and my rating landed on a round number: 1223 → 1200. The odd part is the accuracy: 83.3% on average, one of my better weeks by that measure. If you're stuck at 1200 Elo in chess and your accuracy looks fine, this week may sound familiar. Mine shows where the points actually went.
The week in numbers
| Games | 5 |
| Wins / losses / draws | 1 / 4 / 0 |
| Rating | 1223 → 1200 (−23) |
| Average accuracy | 83.3% |
| My blunders | 9 |
| With White / with Black | 1–2 / 0–2 |
| Openings | Englund Gambit, Zukertort Chigorin, Pirc Classical, London System, Modern Averbakh |
What "stuck at 1200" looks like in my data
From game 2034 on April 6 to game 2114 on July 4, I played 81 rapid games, and my rating never went above 1244 or below 1197. That's three months inside a 47-point band. My peak in this series, 1385, was in November 2025, so this isn't a beginner's ceiling; it's a long slide followed by a long flat line.
This week's four losses fit that pattern in two different ways. In two of them, my opponents simply played better: FeddoM (1229) had 92.7% accuracy and no blunders, and La-Gruyererie (1229) had 97.0% and no blunders, checkmating me on move 23. I made one blunder in each game. In the other two, I had been winning: +7.52 against aatifahmed60 (1250) and −5.71 as Black against x-9186116211 (1245), and I gave both back. My one win was a 13-move London System against MK_Ultra_84 (1184), at 96.9%.
Three moves that cost the week
Game 2110: a pawn pushed into a capture
White against aatifahmed60, Englund Gambit (1.d4 e5). On move 10 the position was about equal (+0.23).
Pawn to e4 puts a pawn where nothing defends it, and Black's knight on f6 can take it while attacking my queen on d2. The engine's evaluation went to −1.48. My opponent took on f3 instead, and later I was winning by a lot (+7.52 after Black's 18th move). Then Black put a bishop on g4, and taking it was a blunder (−1.55); my next move made it −5.01. I lost on move 40 (game 2110).
Game 2111: a pawn break that lost the queen's side
White against FeddoM, Zukertort Chigorin. On move 23 Black was slightly better (−0.51), and Black's pawn on a4 was attacking my pawn on b3.
Pawn to c4 attacks the b5 pawn, which sounds active. But Black takes on b3 first, and then on c4, and my queen's side pawns fall apart: clearly better for Black, −2.54 after my move. Taking on a4 first keeps the damage small. In the game Black took on b3, I took back with a knight instead of the queen, and three moves later I was lost (game 2111).
Game 2112: a capture that opened my king
Black against x-9186116211, Pirc Defense Classical Variation. On move 21 White was slightly better (+0.64). White's bishop had just taken my knight on f5, and White's queen sat on e5.
Taking the bishop back with the g-pawn restores the material, but it opens the g-file in front of my king. White's queen checks from g3, and that same move uncovers White's bishop on c3 against my knight on f6. The engine had it at +5.97. Rook takes on d5 instead attacks White's queen on e5 and, behind it, the bishop on f5, and keeps it close (+0.79 in the engine's line). White missed the queen check, and I was later winning at −5.71 before giving it back twice. The game ended on move 31 (game 2112).
The lesson of the week
Check captures twice. In the two games I had been winning, six of my seven blunders were captures: three in game 2110 and three in game 2112, including the pawn capture above that opened my own king. A capture feels forcing, so I play it fast. Before I take anything now, I ask what the capture opens or leaves behind: a file toward my king, an undefended piece, a square for their queen. That won't help in games where my opponent plays 97% without a blunder, but the games I'd been winning are exactly the ones that would move me off 1200.
All games this week
| Game # | Date | Colour | Opponent (rating) | Result | Opening | Accuracy | Link |
|---|---|---|---|---|---|---|---|
| 2110 | Jun 30 | White | aatifahmed60 (1250) | Loss | Englund Gambit | 75.6% | game |
| 2111 | Jul 1 | White | FeddoM (1229) | Loss | Queen's Pawn Opening, Zukertort Chigorin Variation | 85.5% | game |
| 2112 | Jul 2 | Black | x-9186116211 (1245) | Loss | Pirc Defense, Classical Variation | 70.5% | game |
| 2113 | Jul 3 | White | MK_Ultra_84 (1184) | Win | London System | 96.9% | game |
| 2114 | Jul 4 | Black | La-Gruyererie (1229) | Loss | Modern Defense, Averbakh Variation | 88.0% | game |
The previous diary entry is week 123 (there were no games in between), and the next and last one is week 125. The journey hub has the full series. If your own rating is stuck, review your own games like this and look at where the evaluation swings, not just at the average accuracy.
Frequently asked questions
Why am I stuck at 1200 Elo in chess?
In my case it wasn't the openings. Between April 6 and July 4, 2026, across 81 games, my rating stayed between 1197 and 1244. The points were lost on single moves in the middlegame: in week 124, four losses came down to one or two moves each.
Can you lose with high accuracy at 1200?
Yes. In game 2111 I played 85.5% and in game 2114 88.0%, and lost both, because my opponents made no blunders at all (92.7% and 97.0%). One blunder of mine was enough each time.
How long can a 1200 plateau last?
Mine had lasted three months by this week: from game 2034 to game 2114 I never went above 1244 or below 1197. My peak, 1385, had come in November 2025.