Game #91: beating a 1054 after a Philidor Defense queen trade
On January 24, 2024, in game #91 of my rapid journey, I beat Juds0n777 (1054) with Black while rated around 900. It came out of a Philidor Defense queen trade on move 7, and the whole game is a lesson about where my king ended up after that trade. My rating went from 895 to 907, a gap of 147 points to my opponent, one of the ten biggest upsets of my first 2,116 games.
It was not a clean win. Stockfish counted four blunders for me and three for White. I'm writing it up anyway, because the mistakes have a pattern I can actually use.
Where I was that week
This game sits in week 3 of the journey, my busiest and one of my worst weeks so far: 43 games, 15 wins, 27 losses, rating 899 → 872. The four games before this one were a loss, a win, a loss and another loss, the last one to danigno (1064) with the London System. Right after game #91 came my longest losing run of that week, six in a row, games #92 to #97. So this win was a single good result in the middle of a slide, not a turning point.
| Game | #91, January 24, 2024 |
| Opponent | Juds0n777 (1054) |
| My colour, rating after | Black, 907 (from 895) |
| Time control | 10 minutes + 5 seconds |
| Opening | Philidor Defense (C41) |
| Result | Win, checkmate on move 40 |
| Accuracy (Stockfish) | Me 83.0, White 76.4 |
| My blunders / mistakes / inaccuracies | 4 / 0 / 1 |
The queen trade, and my king on d8
The opening went 1.e4 e5 2.Nf3 d6, the Philidor. White played a slow d3 and pushed d4 two moves later; my knight took the pawn on e4, pawns were traded in the center, and on move 7 White took my queen with check on d8. I had two ways to take back: with the king or with the bishop on e7.
I took with the king. The game analysis flagged it as an inaccuracy; when I checked the position again for this post, the engine rated both recaptures "about equal". So the recapture itself wasn't a mistake worth worrying about. What mattered was the consequence: with no queens on the board, my king stayed on d8, in the middle of the board, on an open file. Everything that went wrong for me in the next ten moves happened on that d-file.
Moment 1: White's free bishop, and my bishop to the wrong file
On move 11 White put a bishop on f5, attacking my bishop on e6. Stockfish called that a blunder: I could simply take it. Its line shows White can then check my king with the knight on f7 and grab my rook in the corner, and Black is still winning after all of it. Instead, I moved my bishop away to d5.
The tool describes my move as a double attack (a move that threatens two things at once): my bishop hits the pawn on g2 and the rook on h1 behind it. That's probably what I saw. What I didn't see is where the bishop was standing: on d5, on the same file as my king on d8. The engine's best answer for White was knight to c3, attacking the bishop; my rook to e8; then White castles queenside, and the rook arrives on d1 behind my bishop. That's a pin (a piece that can't move without exposing a more valuable piece behind it, here my king), and the engine's whole line is built on it. The evaluation went from clearly winning for me to +1.07 for White (evaluations are from White's side).
White didn't play that. On move 12 they put the bishop on e4 instead, which the engine called another blunder. My rook went to e8, and on move 13 it took White's knight on e5. After a few trades I was a knight up for a pawn.
Moment 2: a knight up, and the discovered check
By move 16 the position was very good for me (-5.27). White's bishop had just taken a pawn on d5, and it stood right between White's rook on d1 and my king on d8.
That setup is a discovered check waiting to happen: whenever the bishop moves, the rook behind it gives check, and I have to answer the check instead of saving whatever the bishop attacks. I developed my knight to c6. White took it with the bishop, with check from the rook. My king went to c7, the rooks were traded on d5, and the evaluation was back to about equal (+0.10). I had given back the whole extra piece in one move. Stockfish's choice was the king move first, to c7, off the file, and only then the knight to c6.
Two blunders, both caused by the same square. That's the part of this game I most want to remember.
Moment 3: the endgame where we both blundered
After the trades it was rook and bishop against rook and bishop. My clock was the problem now: at move 25 I had about three minutes against White's eight. On move 25 I walked my king to b6 when the engine wanted a rook trade on e1 (+1.63 for White after my move), and two moves later I made the same choice again.
Taking the rook kept the position about equal. My rook move left White clearly better (+2.53). White gave some of it back on move 28 with a rook check on e5, and then on move 29 played king to b2.
Moment 4: the fork that won the game
This time I found the engine's move. My bishop came to d4 with check on the king on b2, and from the same square it attacked White's rook on e5, which had no defender. That's a fork: one piece attacking two targets, and the opponent can only save one. White's king had to move and my bishop took the rook. White's king move had turned +0.55 into -6.88, and this was the move that cashed it in. The rest took eleven more moves: my rook came round, and the game ended with my rook taking the bishop on d1 for checkmate on move 40. I had a little under two minutes left. You can replay the full game on Chess.com.
What the engine says about this win
The numbers by phase tell the story better than I can. My accuracy was 91.6 in the opening, 68.4 in the middlegame and 90.9 in the endgame. Both middlegame blunders, on moves 11 and 16, came from my pieces landing on the d-file in front of my own king. White made one blunder fewer than I did, but White's last one, the king stepping into a bishop check that also hit an undefended rook, came at the end, so the result went my way.
I won't pretend I was the better player in this game. A 147-point upset in a game with seven blunders between us mostly says that at this level the last blunder decides the game.
The lesson I took from game #91
After a queen trade in the Philidor, my king stays in the center, and the d-file becomes dangerous for me. Before I put any piece on that file in front of my king, I now ask two questions: can a rook get behind it, and is there an enemy piece already sitting between a rook and my king? If the answer is yes, the king steps off the file first. It costs one move, and in this game it would have saved two blunders.
The other half of the lesson is the fork: a check that also attacks an undefended piece is worth looking for every single move, because my opponents at this level leave pieces undefended as often as I do.
What came next
The six-game losing streak started straight after this game, and I finished the week at 872. The pattern from this game, middlegames worse than my openings, shows up across that whole week. The week before is week 2, and the whole story is on the journey page.
If you want your own games read this way, with the engine pointing at the moment and plain words for why, that's what I built Chess Coach for: here's how the analysis works, and why Chess.com's review doesn't always explain the moves.
Frequently asked questions
What happens to the king after a Philidor Defense queen trade?
In my game the queens came off on move 7 and I took back with my king, so it sat on d8 for the rest of the middlegame. Twice that square put my pieces on the same file as my king, and both times the engine marked my move as a blunder.
Should Black take back with the king or the bishop after the queen trade?
In this game Stockfish preferred the bishop recapture and labelled my king recapture an inaccuracy. A fresh check of the same position rates both moves about equal, so it was not the move that decided anything.
How did a 907 beat a 1054 in this game?
Not by playing cleanly: I made four blunders by the engine's count and White made three. The last one decided it: White's king stepped onto a square where my bishop could check it and attack the undefended rook at the same time.