The story of the game
David, no single move lost this one — that's the honest headline. By move 7 you were already around 60/40 against without anything the engine calls a blunder: the double-fianchetto setup handed James a free, full-width center, and the moment 8.e5 hit your knight the squeeze never let go. Meanwhile your defense from move 22 onward was genuinely good — 90.5% accuracy in a lost position. This game was decided in the first eight moves, not the last ten.
The key moments
7...Nf6 — the real culprit (39% → 27%). The knight walks straight into 8.e5 with tempo, and after 8...Nh5 it's stranded on the rim right next to a g-pawn that dreams of g4 — which is exactly how it ended: 11.g4 and 12.gxh5 won the piece for a pawn. The engine wanted 7...d6 first: blunt the e5 push before developing the knight, and ...Nf6 comes a move later in total safety. The rule this game teaches: when an enemy pawn duo is ready to roll forward with tempo, take away its best square first, develop second.
6...Bxc3+ is playable — but know what you're signing. The engine rates it as its #2 choice (a cost of barely 1.5%), so this wasn't the mistake it might look like. What it did do was sell your dark-square insurance: Bh6 came with tempo, the long diagonal became James's property, and your king never found a safe address. If you give up the fianchetto bishop, the compensation has to be concrete — doubled c-pawns to attack, fast queenside play. It never arrived, so the trade was all cost.
17...Kf8 — the fork pattern again (21% → 14%). After 17.Qa4+ the engine prefers 17...Kd8: the material loss is similar, but the king stays connected to its pieces on the queenside. 17...Kf8 conceded the same exchange to 18.Nd7+ and left the king to face 20.Rg1+ alone in the corner. Across your ten analyzed games, allowed knight forks are your single most-flagged pattern — 24 times. Before every king move: name the squares a knight could check you from. Here d7 was shouting.
What went right
- 22...Qh3 and 23...Qxh2: down roughly a rook's worth of material, you found the one plan on the board that creates real problems, and defusing it cost James several win-percent and half his clock. That practical stubbornness is a real strength — it just deserved a better position to work with.
- You never hung a piece outright. The loss was structural, and structural losses are the most fixable kind.
Profile check
This game is your engine profile in miniature: the fork flag (17...Kf8), the middlegame drift (steady bleed, no blunder), and opening generosity (bishop plus center for nothing concrete). It also confirms the listed strength — creative, stubborn defense under pressure.
What to drill next
- Knight-fork scan on every king move — one second: "where can a knight check me from?" It's the highest-value habit in your data.
- Play a set of games where you keep both fianchetto bishops past move 15 — feel what the dark squares are worth before you ever sell one.
- Rim-knight alarm: before any ...Nh5-style move, ask "what kicks it, and where does it go next?" If the answer is "nothing and nowhere," find another square.