The story of the game
David, frame this one: 97.3% accuracy and not a single move the engine wants back — your cleanest analyzed game, and it took twelve moves to make the point. When 8...Qb6 aimed at your b2-pawn, you did the hardest thing in chess: nothing. 9.Nc3 just developed, and the pawn sat there as bait.
The key geometry (memorize this one)
After 9...Nxd4 10.Nxd4 Qxb2 the position looks like Black stole a pawn. It's actually a cage: 11.Ndb5 closes the queen's road home and aims at c7 — and from b5 a knight covers c7, a7 and d6 all at once. Against a king still on e8 with a rook on a8, 12.Nc7+ is a whole tactic by itself. Two days after a knight fork went against you, you delivered the same idea from the other side of the board — the fork-scan homework showing up in your play, first spotting them, now springing them.
What made it work
- You never defended b2. Defending it would have been passive and worse — the pawn was worth more as a lure than as a pawn.
- Development stayed ahead the whole way: both bishops out, castled, rook to e1 by move 8, while Black's queenside slept.
Profile check
New strengths-column entry, in ink: punishing greed with development. Your reviews' lesson has a twin now — when worse, counterplay first; when better, don't chase — develop and let the threats mature.
What to drill next
- Learn the Qb6 poison-pawn family properly (it appears in the Caro-Kann and Sicilian both) — you now own the refutation from memory.
- The b5-knight geometry — c7, a7, d6 — check for it every time an enemy king is still uncastled.