Your rating is climbing. The next points are in your opponent’s reply.
72% of your blitz games are 3+0. Over the last 90 days your rating went from 2364 to 2504, and your accuracy held at 86%, level with your opponents. The engine finds one gap bigger than the rest: short combinations your opponent sees and you don’t.
The short answer
Analysed: AlexandraBotez’s chess.com blitz games: 8,926 games · Stockfish 19 on the latest 2,000 (Nov 18 2022 – Sep 19 2026). Report date: .
Verdict: Rating up, moves steady. Last 90d vs the prior 9 months: accuracy −0.36 (p = 0.70), big mistakes per game +0.08 (p = 0.76), score +0.01 (p = 0.92). None of that is a real change. One number did move: you saved 23% of lost positions, up from 16%. The last 30 days are only 8 games, all unrated, too few to read.
Biggest leaks:
- Two-move tactics against you: ≈1,352 mistakes punished by a short combination · ≈104 rating points, 26 more than your opponents
- Not taking what’s offered: 1,118 missed punishments · 144 free pieces · 20 mates · ≈83 rating points, 15 more than your opponents
- Openings that start you in a hole: White: +0.2 by move 10 · Black: −0.7 · ≈32 rating points, 13 more than your opponents
Free Goat Card for AlexandraBotez · How to review your own games for free · Time pressure in chess, measured
Your game at a glance
- Openings2,330▼ 10
- Tactics2,340▼ 50
- Strategy2,350▼ 30
- Endgames2,290▼ 70
- Brilliant moves112in 2,000 games
Each number is the rating of a player who plays that part of the game the way you do: your opponents’ average rating (2,340), adjusted by how much win-% you lose per move there compared with them, in the same games. Arrows: the change against your opponents compared with Prior 9 mo; grey means it is within the noise. A brilliant move is the engine’s top choice that gives up at least two points of material.
Improving, or lucky?
| Metric | Last 30d | Last 90d | Prior 9 mo | Read |
|---|---|---|---|---|
| Results | ||||
| Games | 8 | 128 | 73 | |
| Rating, start → end | 2504 → 2504 | 2364 → 2504 | 2177 → 2354 | up 140 |
| Score | 68.8% | 55.5% | 54.8% | flat (p = 0.92) |
| Skill: what the engine sees | ||||
| Your accuracy | 75.5 | 86.1 | 86.4 | flat (p = 0.70) |
| Opponents’ accuracy | 75.2 | 86.5 | 86.3 | flat |
| Big mistakes per 40 moves | 3.58 | 1.55 | 1.56 | flat (p = 0.76) |
| Opponent blunders you punish | 50% | 58% | 65% | worse |
| More accurate than opponent | 62% | 47% | 56% | worse |
| Eval at move 10 as White | +1.0 | +0.2 | +0.6 | worse |
| Eval at move 10 as Black | −0.1 | −0.4 | −0.7 | flat |
| Conversion & luck | ||||
| Score after reaching +5 | 75.0% | 78.0% | 78.6% | flat |
| Score after being −5 or worse | 50.0% | 26.5% | 18.4% | saving more |
| Losses that were +5 at some point | 50% | 27% | 32% | better |
| Wins that came from a lost position | 20% | 23% | 16% | more luck |
| Wins on the clock while losing on the board | 20% | 14% | 5% | more luck |
| Losses on time | 0% | 16% | 23% | better |
Rating, rolling 50-game average
Rolling 50-game average: 2318 → 2470. A steady climb through 2026.Accuracy, rolling 50-game average
The two lines move together: you 86.4, opponents 86.7. If you were pulling ahead, the blue line would sit above the orange one.Your clock: wins vs losses
Time left on your clock after your move, in the games you won and the games you lost.
- Move 20
Two seconds apart: 1:46 in wins, 1:44 in losses (medians over your 3+0 games, 686 wins and 801 losses). At move 20 the clock isn’t what separates them.
- Move 30
Seven seconds apart: 0:57 in wins, 0:50 in losses (585 wins, 662 losses). Under a minute either way, and your endgame moves take a median of 0.9 seconds.
The rating has moved; the quality of your moves hasn’t yet. To make 2,500 stick, fix the two leaks where you lose more than similar-rated opponents: two-move tactics and missed punishments.
What you do well
Not everything is a leak. These parts of your game already work, measured the same way as the leaks below: against your opponents, in the same games.
- Strongest area: Strategy (2,350)Your opponents average 2,340. This is the part of your game closest to, or above, their level.
- EndgamesYour endgames lose fewer points than your opponents’: ≈12, against ≈17.
- Rating up 1402364 → 2504 in the last 90d, against 2177 → 2354 in the prior 9 mo: up 140.
- Score after being −5 or worse26.5% in the last 90d, against 18.4% in the prior 9 mo: saving more.
- Losses that were +5 at some point27% in the last 90d, against 32% in the prior 9 mo: better.
Your best moves
6 leaks, ranked by rating points lost beyond your opponents
Every position below is from one of your own games in the last 120 days. Decide on your move first. Then open the answer and step through what you played and what was best with the ← → buttons (or your arrow keys). Red arrows show what went wrong, green arrows what was best.
1 · Two-move tactics against you
≈1,352 mistakes punished by a short combination · ≈104 rating points, 26 more than your opponentsAfter certain particular dispositions of pieces have proved his undoing, the beginner will develop the perception of threats.
This is your biggest gap. Opponents beat you with short combinations: capture sequences on under-defended pieces (472), attacks on pieces with no retreat (374), checks (313) and forks (173). It costs you ≈104 rating points, 26 more than your similar-rated opponents lose the same way in the same games. No other leak comes close.
- Why it happens
- This is not loose-piece awareness. The combination starts with their reply: a capture sequence, a check, or an attack on a piece that has run out of squares. You calculate your own line and stop one move short of their best answer.
- The habit
- Before any move that captures, pushes a pawn near your king or leaves a piece with two squares or fewer, ask: what is their most forcing reply? Checks first, then captures, then attacks on your least mobile piece.
- The drill
- Three sessions a week of Aagaard’s Grandmaster Preparation: Calculation, full lines on paper, their best defence written at every branch. On the other days, solve the positions from this report from the diagram without moving pieces, then play them out against Stockfish at 3+0.
- What success looks like
- At the next audit, no more than 27% of your big mistakes come from a short combination (now 40%).
Practice this: Forks: one piece, two targets →
2 · Not taking what’s offered
1,118 missed punishments · 144 free pieces · 20 mates · ≈83 rating points, 15 more than your opponentsI have a weakness for any piece in excess of my opponent's numbers - from pawn to queen.
Your opponents hand you material, and 58% of the time you take it. In the prior 9 months it was 65%. You passed on 144 free pieces and 20 mates. In 48% of the misses the winning move was quiet: an attack on a piece with nowhere to go. That costs ≈83 rating points, 15 more than similar-rated opponents in the same games.
- Why it happens
- The win is often quiet: 48% of your misses were an attack on a piece with nowhere to go, not a capture or a check. At 3+0 you continue your own plan and the chance passes. The rate fell from 65% to 58% in the last 90 days.
- The habit
- After every opponent move, one question before your plan: “What did that change?” Which of their pieces lost a defender or a square?
- The drill
- Solve the missed positions in this report and in your own games, then write the opponent’s best defence and your answer. Add Lichess Quiet move and Trapped piece puzzles at Harder difficulty, and Yusupov’s Chess Evolution volumes for calculation with quiet moves.
- What success looks like
- Punish at least 65% of your opponents’ blunders (now 58%), your level from the prior 9 months.
- Lichess Puzzles: Quiet MovepracticePuzzles where the win is neither a check nor a capture. Set difficulty to Harder so they sit near your level.
Practice this: Is it really safe? Counting attackers and defenders →
3 · Openings that start you in a hole
White: +0.2 by move 10 · Black: −0.7 · ≈32 rating points, 13 more than your opponentsYou should keep in mind no names, nor numbers, nor isolated incidents, not even results, but only methods.
By move 10 you are +0.2 as White and −0.7 as Black. In 16% of your Black games you are already 1.5 pawns down. The worst are French-type structures: 31% in the Sicilian French Variation (−1.9 by move 10) and 23% in the MacCutcheon. The gap to similar-rated opponents is 13 rating points: smaller than the tactics leaks, but it repeats every game.
- Why it happens
- The deficit is not about principles. It sits in a few French structures: the Sicilian French Variation (−1.9 by move 10), the Classical Steinitz/Boleslavsky and the MacCutcheon Advance, all scoring 31% or less. You reach the same positions and play the same second-best move.
- The habit
- When you reach a line you have lost in before, stop on the move where it went wrong last time and think. You have the clock for it: 2:38 left at move 10.
- The drill
- Run your Black games in those lines through Stockfish and find the first move that drops the eval. Check that position in the master database, pick one line and write its move-10 plan in a sentence. Start with 3.Bd3 in the French, where you played 3…Nf6 six times and the engine prefers 3…dxe4. Then play 10 games from each repaired position against an engine or a sparring partner.
- What success looks like
- Average eval at move 10 as Black at or above −0.3 (now −0.7).
- Lichess Analysis board + Opening ExplorerpracticeEngine and master database on one board. Check your move-10 positions against strong games.
Practice this: Opening rules that win games →
4 · One-move oversights
≈502 mistakes · 15% of all · ≈57 rating points, 8 more than your opponentsIt never even passed through my mind that this was threatened.
267 times a piece went to a square where it could be taken; 208 times an attacked piece stayed where it was. Your big-mistake rate is 2.3 per 100 moves with more than 90 seconds left, and more than doubles below that (4.9 at 45–90s, 5.3 at 20–45s). The check is there when you have time. It drops out when you don’t. Against similar-rated opponents the gap is 8 rating points.
- Why it happens
- You know these. With more than 90 seconds you make 2.3 big mistakes per 100 moves; at 45–90 seconds it is 4.9. The check you run with time drops out under a minute and a half.
- The habit
- Below 90 seconds, keep one check: what does their most active piece see after my move? The positions above went to a long bishop diagonal and a pawn attacking two pieces.
- The drill
- Set up equal middlegames from your own games and play them against Stockfish with 60 seconds each. It trains the check at the clock where you drop it.
- What success looks like
- Big mistakes at or below 1.3 per 40 moves (now 1.55), measured on your next 50 games.
Practice this: The blunder check →
5 · Giving back won games
Reached +5 in 976 games · lost 186 · drew 61 · ≈29 rating points, 6 more than your opponentsFirst restrain, next blockade, lastly destroy.
You reached +5 in 976 games and lost 186 of them. 32% of your big mistakes come from winning positions, at a median of 0:36 on the clock. On its own it looks big, but your opponents lose almost as much: you lose only 6 rating points more. It ranks fifth for that reason.
- Why it happens
- Your opponents give back almost as much, so the gap is small. Of the games you lost from +5, 66 ended by resignation, 63 on time and 57 by checkmate. The median clock when you blunder a winning position is 0:36: the opponent has nothing to lose, and you have little time to check.
- The habit
- At +5 with under a minute, switch modes: find their one counterplay idea and stop it, then trade. Skip the fastest win if it needs calculation you have no time for.
- The drill
- Pick 20 of your games lost or drawn from +5. Play each from the +5 position against Stockfish at 1+0 until you win it twice in a row.
- What success looks like
- Score at least 86% after reaching +5 (now 78%).
- Aron Nimzowitsch: My System (Wikipedia)bookRestrain, blockade, destroy: the prophylactic method for won positions.
Practice this: Converting a won game: trade pieces, not pawns →
6 · Endgames: the last point
999 endgame mistakes · 53% in rook+minor endings · ≈12 rating points, 5 fewer than your opponentsThe rooks belong behind passed pawns, behind their own in order to support their advance, behind the enemy's in order to impede their advance.
This one is a strength: you lose 5 fewer rating points here than similar-rated opponents. 59% of your games reach an endgame, and from a winning one you score 90%. The rest of that 10% is wins you had already earned. 53% of your endgame mistakes come in rook+minor endings, and 11 winning endgames were lost on time. The technique is there; the clock isn’t.
- Why it happens
- Your technique is ahead of your opponents’. What costs points is speed: a median of 0.9 seconds per endgame move, and 11 winning endgames lost on time. Most mistakes come in rook+minor endings, where one pawn race decides.
- The habit
- Before any capture or king move in an ending, count the pawn race. The positions above were a missed trade into a won pawn ending, a pawn grabbed too early and a king going the wrong way.
- The drill
- Once a week, play rook+minor endings from your own games against Stockfish at 1+0, from both sides. For theory, the rook-ending chapters of Dvoretsky’s Endgame Manual.
- What success looks like
- Convert at least 95% of your winning endgames (now 90%).
- Jeremy Silman: Silman’s Complete Endgame CoursebookSplit by rating. Read only the master-level section.
Practice this: Converting a won game: trade pieces, not pawns →
The plan
Four rules, starting today
- Their forcing reply, every move. Before you release a move, name their checks, captures and attacks on your least mobile piece. Capture sequences (472), trapped pieces (374) and checks (313) did most of the damage.
- Their move first, your plan second. You cash in 58% of their blunders, down from 65%. Half the misses were quiet: look for their piece with no squares before you continue your plan.
- Repair the Black move-10 positions. As Black you are −0.7 by move 10, and 16% of your Black games start 1.5 pawns down. The French lines in Week 3 are where it happens.
- Spend the clock where it pays. Your big-mistake rate more than doubles below 90 seconds (2.3 to 4.9 per 100 moves). You play the opening at 1.4s a move and still have 2:38 at move 10. Spend the surplus on the moves that capture, trade or open your king.
Before every move
- Their most forcing reply: checks, captures, the piece with no squares.
- Their last move: what lost a defender or a square?
- Capture or pawn move near my king: what opens?
- Winning: stop their one idea, then trade.
Week by week
Their forcing reply
Your game, move 27. You were +6.9. What was the threat?
- Daily 30 min: Aagaard’s Grandmaster Preparation: Calculation. One exercise at a time, full lines written before you check. At every branch, write their best defence first. That is the move you skip.
- Daily 10 min: solve the positions above from the diagram, without moving pieces. Name their forcing reply out loud, then play each one out against Stockfish at 3+0 from your side.
- Games: 3+0 rated, as usual. After each loss, find the first eval drop and label the reply you missed: capture sequence, check, trapped piece or fork.
Repair the Black move-10 positions
Move 8 as Black. White’s d5-knight is eyeing c7. Which move takes a pawn and guards c7 at once?
- Export your Black games in the Sicilian French Variation, the Classical Steinitz/Boleslavsky and the MacCutcheon Advance. Run them through Stockfish and mark the first move that drops the eval.
- Check each of those positions in the Lichess master database. Pick one line per structure and write the move-10 plan in one sentence.
- Start with 1.e4 e6 2.d4 d5 3.Bd3. You played 3…Nf6 there six times; the engine prefers 3…dxe4.
- Play 10 games from each repaired position against Stockfish or a sparring partner at 3+0. Then 20 rated games as Black.
Cash in the quiet wins
White’s b5-pawn can be taken two ways. Which one?
- Daily 20 min: Lichess puzzle themes Quiet move and Trapped piece, difficulty set to Harder or Hardest. 48% of your misses were quiet wins.
- Two sessions a week: Yusupov’s Chess Evolution volumes, the last three books of his course. Solve every exercise before you look.
- After each game, open every opponent blunder the engine flags. Did you see the winning move? If you saw it and rejected it, find the reason.
Winning positions under a minute
You are +6.8. Black’s queen just went to b5. What attacks it?
- Pick 20 of your games lost or drawn from +5. Play each from the +5 position against Stockfish at 1+0 until you win it twice in a row.
- At +5 with under a minute: name their one counterplay idea before you move, and stop it. Trade pieces when it costs nothing.
- Keep 20 seconds for the finish. 63 of the games you lost from +5 ended on time.
Rook + minor endings at speed
Endgame, you are +4.2. The knight is checking you. What pawn ending do you get if you take it?
- Twice this week: rook+minor endings from your own games, played out against Stockfish at 1+0 from both sides. 53% of your endgame mistakes come from them.
- Before any capture or king move in an ending, count the pawn race. Your median endgame move takes 0.9 seconds; this one deserves three.
- For theory, Dvoretsky’s Endgame Manual, the rook-ending chapters. Your technique is already ahead of your opponents’, so keep it short.
Audit again
- Play 60 rated 3+0 games, half of them as Black. Stop a session at 15 games: from game 16 on you score 43%, against 48% in games 1–3.
- Re-run this analysis on them. You pass if big mistakes are ≤ 1.3 per 40 moves (now 1.55) and you score 86% from +5.
- If you don’t pass, repeat weeks 1–2. Two-move tactics are the leak where you trail similar-rated opponents most.
Games: the latest 2,000 blitz games from the chess.com public API (Nov 18 2022 – Sep 19 2026). Engine: Stockfish 19, 50,000 nodes per position; puzzle lines at 3,000,000. Win% model and accuracy formula: lichess. A big mistake loses ≥ 20 win-% in one move. Rating-point costs: 695 Elo per point of expected score, counting only mistakes the opponent cashed in. Trends: Welch z-test against Prior 9 mo.