Your chess report
luizspinelli
Rapid · 10 min1 Sept 2026 – 15 Nov 2026games from 1 Sept 2026 to 12 Nov 2026
11 chapters · numbers computed from your games and Stockfish (sf19-lite, depth 12) · text written by AI in English
1. Summary
Rating
Rating after the last rated game of each week.
Results
52.5% score
Accuracy
72.3 average
Chess.com accuracy, 553 of 850 games.
Compared with before
| Previous period | This period | |
|---|---|---|
| Games | 532 | 850 |
| Score | 54% | 52% |
| Rating change | +240 | +290 |
| Accuracy | 70.5 | 72.3 |
1 Jun 2026 – 1 Sept 2026
Diagnosis
You win fast and lose slow. Games of 20 moves or fewer score 63%, and games of 61+ moves score 37%. Games that reach an endgame score 46%, against 61% for games decided in the middlegame. Your losses do not come from time trouble. You finish losses with 4:52 left on average, and 162 of the 168 engine-marked errors came with more than 2 minutes on your clock. The main leak is fast captures and fast moves in won or equal positions: 46 turnarounds against you in the last 100 games, many from 85%+ winning chances (28 Oct · vs 906 · lost, 28 Oct · vs 888 · lost, 29 Oct · vs 872 · lost, 8 Nov · vs 887 · lost, 10 Nov · vs 922 · lost). This is why the rating climbed to about 900 and then stopped. The last 100 games went only 902 → 918, and you score 39% against 900–949 opponents (85 games).
What changed
- Rating: it rose faster. You gained +290 (628 → 918), against +240 (388 → 628) in the previous period.
- Accuracy: it went up a little, from 70.5 to 72.3.
- Score: it was 54% in the previous period and 52% now. The samples are 532 and 850 games, so they are not comparable. Do not read the 2-point drop as decline.
- Opponents: they got stronger every month. The average opponent was 743 in September, 843 in October and 881 in November. Your score stayed between 51% and 57% across those months.
- What did not change: the second half of the period has the plateau. The last 100 games score 50%, and the top rating band scores 39%. Next 30 days, watch your score against 900+ opponents. That band is the real test.
What to keep
- Castling early. When you castle by move 10, you score 55% in 371 games. When you never castle, you score 44% in 177 games. Clean wins like 28 Oct · vs 858 · won (castled move 9, 90 accuracy) and 30 Oct · vs 802 · won (castled move 10, 86 accuracy) follow this habit.
- Bouncing back after losses. After one loss you score 54% (376 games). After two losses you score 55% (162 games). Both are above your 52% average. Losses do not tilt you. Keep that calm.
- The 1.e4 e5 2.Nf3 setup when Black plays something unusual. It scores 77% in 39 games, and 81% in the 32-game sub-line. In the dossier it went 4 out of 4 (28 Oct · vs 879 · won, 11 Nov · vs 871 · won, 11 Nov · vs 922 · won, 11 Nov · vs 893 · won). Keep playing simple developing moves there.
2. Repertoire
All games of the period on the left; the last 100 (with engine) on the right.
As White
| Opening | Games | W · D · L | Score | Last 100: games | Accuracy | Errors / game |
|---|---|---|---|---|---|---|
| Scotch Game | 122 | 57% | 17 | 76.2 (8) | 1.8 | |
| Scandinavian Defense | 40 | 55% | 2 | — | — | |
| Kings Pawn Opening | 39 | 77% | 4 | 71.8 (4) | 1.0 | |
| Philidor Defense | 31 | 63% | 5 | 76.8 (3) | 1.4 | |
| Caro Kann Defense | 24 | 56% | 3 | — | 0.3 | |
| French Defense | 22 | 39% | 2 | — | — | |
| Four Knights Game | 20 | 55% | 2 | — | — | |
| Petrovs Defense | 20 | 63% | 4 | 81.3 (4) | 0.3 |
As Black
| Opening | Games | W · D · L | Score | Last 100: games | Accuracy | Errors / game |
|---|---|---|---|---|---|---|
| Petrovs Defense | 84 | 50% | — | — | — | |
| Kings Pawn Opening | 42 | 57% | 1 | — | — | |
| Four Knights Game | 39 | 44% | 4 | — | 0.8 | |
| Indian Game | 37 | 47% | 9 | — | 2.9 | |
| Bishops Opening | 25 | 58% | 4 | 76.0 (3) | 1.8 | |
| Giuoco Piano Game | 20 | 23% | 2 | — | — | |
| Queens Pawn Opening | 16 | 44% | 2 | — | — | |
| Vienna Game | 16 | 41% | 3 | — | 3.3 |
Best lines
- C50Italian Game · 3 games · 100%
- C20Kings Pawn Opening Leonardis Variation · 11 games · 82%
- C40Kings Pawn Opening Kings Knight Variation · 32 games · 81%
- B07Pirc Defense · 11 games · 77%
- A00Undefined · 4 games · 75%
- B12Caro Kann Defense Advance Variation · 4 games · 75%
- C41Philidor Defense · 15 games · 73%
- C42Petrovs Defense Classical Damiano Variation Kholmov Gambit · 11 games · 73%
Worst lines
- C22Center Game Accepted Paulsen Attack · 3 games · 0%
- C58Italian Game Knight Attack Polerio Bishop Check Line · 3 games · 0%
- C50Giuoco Piano Game Giuoco Pianissimo Italian Four Knights Variation · 6 games · 8%
- A45Trompowsky Attack · 5 games · 20%
- D04Colle System · 4 games · 25%
- C48Four Knights Game Spanish Variation · 4 games · 25%
- C28Vienna Game Falkbeer Vienna Gambit · 4 games · 25%
- C45Scotch · 8 games · 25%
Keep
As White, keep 1.e4 e5 2.Nf3 against Black's odd second moves (77% in 39 games). Also keep it against 1.e4 e5 2.Nf3 d6 (63% in 31 games). Calm development plus early castling is working there.
Change
As Black, the Italian Game (1.e4 e5 2.Nf3 Nc6 3.Bc4) is your worst line: 23% in 20 games (4 wins, 15 losses). In 28 Oct · vs 893 · lost the engine flagged 7...Nd4 and 8...Nxf3+, so the trouble starts early. Prepare one simple plan: develop, castle, and avoid early knight jumps. Also note the Scotch as White (3.d4). It scored 46% in the dossier's 14 games, with 1.82 errors per game.
3. Critical positions
The 10 moments where your win chance fell the most in the last 100 games, plus every missed mate. Board from your side, before your move.
- Clock after the move
- 2:50
- Time spent on it
- 6 s
- Grade
- Blunder · decisive
Game 7 Nov · vs 874 · lost, endgame, you were down a rook (−5). Rh5 was mate in one (Mate). You played Rf1 after 6 seconds. The engine reply runs f2 Rh5+ Kg8 Rg5+: the mate is gone and you stay −5. Your chance fell from 100% to 7%. Look for checks first, even when losing.
- Clock after the move
- 4:23
- Time spent on it
- 7 s
- Grade
- Blunder · decisive
Game 28 Oct · vs 906 · lost, you were up +7. You took a pawn with Rxc4, and the engine reply was Ra8#, mate (Mate). Bc6 kept everything (b4 Re8 b5, still +7). The chance fell from 93% to 0%. You spent 7 seconds. Before grabbing, check every enemy check against your king.
- Clock after the move
- 3:37
- Time spent on it
- 3 s
- Grade
- Blunder · decisive
Game 10 Nov · vs 922 · lost, you were up +2. You took a bishop with Rxe7, and the reply was Qxg2#, mate (Mate). The quiet b3 kept +2 (Rd8 Qf3 Rg5). The chance fell from 90% to 0% after 3 seconds of thought. Same lesson as 28 Oct · vs 906 · lost: a free piece means nothing if you get mated next move.
- Clock after the move
- 8:46
- Time spent on it
- 4 s
- Grade
- Blunder · decisive
Game 8 Nov · vs 887 · lost, opening, material even. You played Bf4+. The reply Qxf4+ gxf4 Rxa8 Kb1 leaves you −3. Qxa7+ (Kc6 Qa4+ Kc7) wins a pawn instead (+1). The chance fell from 93% to 11% after 4 seconds. Your queen sat on a8, and the reply line captures it there.
- Clock after the move
- 9:08
- Time spent on it
- 2 s
- Grade
- Blunder · decisive
Game 29 Oct · vs 808 · lost, you were up +1. Nxe4+ was a Fork: Kg1 Nxg5 h4 ends at +5. You played Ng4+, and the reply Qxg4 left a Hanging piece (−2). The chance fell from 92% to 11%. You spent 2 seconds. Two knight checks were possible, and you picked one without comparing them.
- Clock after the move
- 9:18
- Time spent on it
- 6 s
- Grade
- Blunder · decisive
Game 29 Oct · vs 872 · lost, you were up +1. You played Qxc5, and the reply dxc5 Bxd7+ Qxd7 Bd2 ends at −7: a Hanging piece. Your queen was taken by a pawn. Bf4 kept +1. The chance fell from 89% to 11% after 6 seconds. Before any queen capture, ask: "Is that square defended?"
- Clock after the move
- 9:14
- Time spent on it
- 11 s
- Grade
- Blunder · decisive
Game 30 Oct · vs 826 · lost, you were up +1. You played Ne4, and the reply Qxd5 ends at −8: a Hanging piece (your queen). Qe4+ traded queens and kept +1 (Qe2 Qxe2+ Kxe2). The chance fell from 86% to 9%. Before moving, check what is attacked right now, starting with your queen.
- Clock after the move
- 7:01
- Time spent on it
- 0 s
- Grade
- Blunder · decisive
Game 28 Oct · vs 895 · lost, endgame, you were up +3 (a bishop). You played Kc4 in 0 seconds. In the reply line, Black's g-pawn reaches g2 while your b-pawn is only on b5. Bg8 (Kf5 Bd5 Ke5) kept control. Material never changed, yet your chance fell from 88% to 12%. Count pawn races before moving.
- Clock after the move
- 2:40
- Time spent on it
- 23 s
- Grade
- Blunder
Game 12 Nov · vs 927 · lost, endgame, material even. h6 was mate (Mate). You played hxg6, and the reply Rxg6+ Kh8 Rxa6 Rg8+ leaves you −1. Your chance fell from 100% to 31%. You spent 23 seconds, so time was not the issue. You simply did not look at the checking and mating moves first.
- Clock after the move
- 5:18
- Time spent on it
- 2 s
- Grade
- Blunder
Game 30 Oct · vs 811 · lost, endgame, material even. Re1 was mate (Mate). You took a rook with Nxh6 in 2 seconds instead. The chance fell from 100% to 50%. Together with 7 Nov · vs 874 · lost and 12 Nov · vs 927 · lost, this makes three missed mates in one. The habit is the same each time: you took or moved before checking all checks.
4. Where you go wrong
168 errors and blunders flagged by Stockfish in 100 games (inaccuracies left out).
By phase
Errors per 100 of your moves.
By move number
Count of errors by the move they happened on.
By time left on your clock
Errors per 100 moves, by the clock after the move, as a share of the starting time.
Winning positions not converted
31 games not won after reaching 75%+
Swings in the last 100 games: 46 against you, 39 in your favour.
- 12 Nov · vs 927 · lost peak 100%
- 7 Nov · vs 874 · lost peak 100%
- 30 Oct · vs 811 · lost peak 100%
- 11 Nov · vs 897 · lost peak 99%
- 7 Nov · vs 873 · lost peak 95%
- 8 Nov · vs 887 · lost peak 93%
- 28 Oct · vs 888 · lost peak 93%
- 28 Oct · vs 906 · lost peak 93%
What to stop doing
- Playing critical moves in seconds. Your losses end with 4:52 on the clock. 162 of 168 engine errors came with more than 2 minutes left. Six of the ten critical positions above took 7 seconds or less (28 Oct · vs 906 · lost, 29 Oct · vs 872 · lost, 29 Oct · vs 808 · lost, 30 Oct · vs 811 · lost, 8 Nov · vs 887 · lost, 10 Nov · vs 922 · lost). This costs you the most, because it turns wins into losses. Instead, spend 15–20 seconds on any move when you are ahead, when the opponent just gave a check, or when you are about to capture.
- Grabbing material without checking the reply. Look at 28 Oct · vs 906 · lost Rxc4, 10 Nov · vs 922 · lost Rxe7, 29 Oct · vs 872 · lost Qxc5, 29 Oct · vs 846 · lost Qxd7 and 1 Nov · vs 841 · lost Qxg7. Each was a capture marked as a blunder (a "blunder" is a move that throws away a big part of your winning chance). Hanging piece appears 12 times in what you allowed. Instead, before every capture, list the opponent's checks and captures after it.
- Resigning quickly. Resignation is 29 of your 48 dossier losses and 164 of 377 in the period. In 29 Oct · vs 849 · lost you resigned after 3 moves, down one pawn. In 28 Oct · vs 867 · lost, 29 Oct · vs 877 · lost and 30 Oct · vs 826 · lost you resigned around moves 12–15. Instead, play on at this level. Opponents at 900 also blunder: they lost 196 games to you by resigning, and 19 more on time.
5. Tactics
Errors and blunders in the last 100 games, by the tactical theme the algorithm found in the engine line.
What you missed
15 of 168 errors had a theme in the best move, the one you did not play.
You missed 15 tactics. The biggest groups are Fork (7), Hanging piece (4) and Mate (3). Train 10 mate-in-one and fork puzzles daily. Before you solve, say out loud every check you see.
What you allowed
27 of 168 errors gave your opponent a theme in the reply.
You allowed 27 tactics, mostly Hanging piece (12), Fork (8) and Mate (5). Practice "what does the opponent threaten?" puzzles. After each of your candidate moves, check every piece the opponent can capture next.
127 errors with no theme found. Themes looked for: Mate, Hanging piece, Fork, Pin, Skewer, Discovered attack. A theme only counts when the algorithm is sure of it in the line.
6. Strategy indicators
From all games of the period, and the engine ones from the last 100 games. They point to habits; they do not evaluate the positions.
King safety
Castled by move 10, later or never; king still on the d or e file at move 15 without castling.
You score 55% when you castle by move 10 (371 games), 51% when you castle later (277 games) and 44% when you never castle (177 games). Test castling by move 10 in every game for the next 50.
Pawn structure
Your pawns at move 20. Shorter games are left out.
An isolated pawn is a pawn with no friendly pawns on the files next to it. With one, you score 45% (316 games); without isolated or doubled pawns, you score 53% (299 games). Note it in reviews and see whether your errors cluster there.
Bishop pair
At move 20. Shorter games are left out.
When only you keep both bishops, you score 61% (69 games). When only the opponent has them, you score 50% (59 games). Try not to give up a bishop for a knight without a reason.
Simplifying when ahead
Last 100 games, by the engine: the first time you were 2 pawns up or more (in two positions in a row), whether each side lost a piece in the next 8 moves. Games that ended sooner are left out.
How losses slip away
Engine-checked losses of the last 100: to one decisive move, or gradually (your chance fell 30 points or more over up to 10 moves, none of them a mistake). Losses with neither are left out.
7. Endgames
By the pieces on the board when the endgame began (kings and pawns don't count).
By endgame type
| Endgames | Games | W · D · L | Score | Converted | Last 100: games | Errors / 100 moves |
|---|---|---|---|---|---|---|
| Pawn endgamessmall sample | 1 | 50% | — | 1 | 9.1 | |
| Minor-piece endgamessmall sample | 6 | 33% | 2/2 | 2 | 0.0 | |
| Rook endgames | 20 | 38% | 3/4 | 2 | 1.9 | |
| Rook and minor-piece endgames | 325 | 45% | 62/90 | 41 | 4.8 | |
| Queen endgamessmall sample | 5 | 40% | — | — | — | |
| Queen with other pieces | 151 | 51% | 57/65 | 17 | 2.8 |
Converted: of the endgames you entered 2 or more points of material up, how many you won.
Errors per 100 of your moves, from the start of the endgame.
Rook endgames
You score 38% in rook endings (20 games) and won 3 of the 4 you entered ahead. The dossier has only 2 of them, too few to measure errors.
Rook and minor-piece endgames
This is your most common endgame, at 45% in 325 games. You converted only 62 of 90 advantages, and you make 4.8 errors per 100 moves here. That is the most expensive phase of your play.
Queen with other pieces
With queens still on, you score 51% (151 games). You converted 57 of 65 advantages, with only 2.8 errors per 100 moves. That is much cleaner than rook-and-piece endgames.
8. Patterns
Sessions and tilt
97 sessions, 8.8 games each on average, longest 32.
After one loss you score 54%, and after two losses 55%. Both are above your 52% average, so you do not need a forced break after a single loss.
Game number in the session
Your first game of a session scores 50% and games 6 and later score 50% (426 games). Games 3–5 peak at 56% (236 games), so warm up and stop around game 5.
Time of day
Afternoon is your best time: 59% in 182 games, with 74.5 accuracy. Night scores 50% in 449 games and late night 51% in 201 games, so move your sessions earlier when you can.
Day of the week
Saturday is best at 56% (187 games), and Monday and Wednesday are lowest at 49%. The gap is small, so do not plan around weekdays.
Stronger and weaker opponents (last 100 games)
Opponent's rating minus yours, before the game.
In the last 100 games, the scores in each rating gap range from 43% to 55%. Each gap has only 16–23 games, so there is no clear pattern. Opponent strength is not what decides your results.
9. Games to review
The games where you had the biggest advantage and still did not win (not already shown in section 3). Key moment drawn.
- Best win chance
- 99%
- Key moment
- 18. Rde8 (best Rxe3+)
- Clock then
- 6:17
You reached 99% winning chances, then played 18...Rde8 with 6:17 left, and the chance fell to 41%. The engine wanted Rxe3+, a check that you skipped. Open the Chess.com analysis at move 18 and see why that check was better.
- Best win chance
- 95%
- Key moment
- 39. Rxc5+ (best Rh8)
- Clock then
- 5:58
This is the Scotch. You were at 95% before 12.Be2, and later at 88% before 39.Rxc5+, which fell to 31% with 5:58 on the clock. The engine preferred the quiet Rh8 over the check. Compare both moves in the analysis.
- Best win chance
- 93%
- Key moment
- 41. h1=Q (best Rxd5)
- Clock then
- 3:35
At move 41 you promoted with h1=Q, and the chance fell from 93% to 34%. The engine wanted Rxd5 instead. Check in the analysis why promoting was not the best move there. Also note 58...Kb6, played with under 2 minutes left.
- Best win chance
- 93%
- Key moment
- 22. Bxd4 (best Rxe4)
- Clock then
- 8:03
You were at 89%, then 20...cxd4 and 22...Bxd4 dropped it to 50%. The engine preferred exd4 and then Rxe4. Look at those two captures, which were made with over 8 minutes on the clock.
- Best win chance
- 92%
- Key moment
- 13. Qxd7 (best O-O)
- Clock then
- 8:25
At move 13 you took with Qxd7 instead of castling, and the chance fell from 62% to 10%. You had 8:25 left. This game is the clearest example of "castle first, grab later". Review it with the king-safety numbers in mind.
10. Two-week plan
Two-week plan
Week 1
- Do 15 minutes of mate-in-1/2 and hanging-piece puzzles every day. Track your solve rate.
- In every game, say "checks, captures, threats" before each capture. Afterwards, write down how many captures you made in under 5 seconds.
- Replay 28 Oct · vs 906 · lost, 29 Oct · vs 872 · lost, 29 Oct · vs 808 · lost, 30 Oct · vs 826 · lost and 10 Nov · vs 922 · lost. For each one, write one sentence: "what I missed".
Week 2
- Spend 30 minutes on your Scotch plan (develop, castle, queen moves once). Play 10 games with it and count the queen moves in the first 10.
- Do two endgame sessions of 20 minutes each, converting won positions from 28 Oct · vs 895 · lost and 7 Nov · vs 873 · lost against the computer.
- Measure your blunders per game in your next 20 analyzed games. The target is under 0.7.
What to study
- A blunder check before every move (top priority). The data behind it: 12 Hanging piece and 5 Mate allowed, plus three missed mates in one (30 Oct · vs 811 · lost, 7 Nov · vs 874 · lost, 12 Nov · vs 927 · lost). The routine is short: look at checks, then captures, then threats, for both sides. Do 15 minutes of puzzles daily, mostly mate in 1–2 and hanging pieces. Then replay 28 Oct · vs 906 · lost, 29 Oct · vs 872 · lost, 30 Oct · vs 826 · lost and 10 Nov · vs 922 · lost and find the opponent's reply yourself before looking.
- Your Scotch as White (1.e4 e5 2.Nf3 Nc6 3.d4). It has the most games and 1.82 errors per game. In 29 Oct · vs 850 · drew, 7 Nov · vs 837 · lost, 8 Nov · vs 887 · lost and 8 Nov · vs 888 · lost your queen moved three times in the first 10 moves. Learn one plan: after 4...Nxd4 5.Qxd4, develop both bishops and knight, then castle, before the queen moves again. Replay 29 Oct · vs 872 · lost, 29 Oct · vs 846 · lost, 7 Nov · vs 873 · lost and 8 Nov · vs 887 · lost for 20 minutes each.
- Rook and minor-piece endgames. These are 325 games at 45%, with 4.8 errors per 100 moves and only 62 of 90 advantages won. Twice a week, take a won position from your own games and play it out against the computer until you win. Start with 28 Oct · vs 895 · lost from move 45.
- Black against the Italian (1.e4 e5 2.Nf3 Nc6 3.Bc4). You score 23% in 20 games. Write down one simple setup: knight to f6, bishop to c5 or e7, castle, and no early knight jumps. Then replay 28 Oct · vs 893 · lost to see where it went wrong.
Routine rules
- When to play: in the afternoon whenever possible (59%, against 50% at night).
- How long: stop after 5 games. Games 3–5 score 56%, and game 6 onward drops to 50%. Your average session is 8.8 games, and the longest was 32.
- Before the first game: solve 5 puzzles. First games score only 50%.
- After a loss: you do not need to stop. Your post-loss scores are 54–55%. Instead, take 1 minute to name the move that lost it.
- During each game: you have time to spare, so use it. Spend at least 15 seconds on any move when you are winning, when you are about to capture, or when the opponent gives a check.
11. How to track
Today's value comes from your games; the target comes from the AI.
Clock left in losses
4:52Target: 3:00 or lessToday you finish losses with 4:52 left. Using that extra time on key moves should lower it. Your error rate does not rise with less time on the clock, so spending it costs you nothing.
Serious errors per game (engine)
0.92Target: 0.6 or lessToday you make 0.92 blunders per game. Most of them are fast captures and missed checks. The blunder-check routine targets exactly these.
Errors per game (engine)
1.68Target: 1.3 or lessToday the engine marks 1.68 errors of all kinds per game. Since 162 of 168 came with plenty of time, slowing down should cut them directly.
Losses by resignation (last 100 games)
29 of 48Target: under 20 of 48Today 29 of your 48 losses are resignations. Playing on gives opponents at this level the chance to blunder back.