Reasoning eval · Puzzle wR7QA
1 of 3 models found the move, and said 4 false things along the way
Green is the solution, red is what the model played when it differed. Each claim below was pulled out of the prose beside it and checked against this exact position.
- Rating
- 2064(2000 band)
- Solution
- d4c3 d3c3 e5e4
- Motifs
- crushingdefensiveMoveendgamepawnEndgameshort
- FEN
- 8/6p1/p5p1/Pp2k1P1/1P1p3P/2PK4/8/8 b - - 0 45
Stated plan
March the king toward f4–g4/g3 to win the undefended h4 and then g5 pawns, creating a decisive kingside passed pawn.
Reasoning
Black's queenside pawns are all frozen (a6/a5, b5/b4, d4 blocked by c3), so the only progress comes from the king. The kingside is where Black can actually win material: h4 is undefended and g5 is only defended by h4, so marching the king via f4–g4/g3 lets Black capture h4 first, then g5, netting two pawns and creating an easy passed g-pawn (Black already has doubled g-pawns). Meanwhile White's king is too far from the kingside to intervene in time, and any White queenside break (Kc4xb5) is slower than Black's kingside harvest since Black's king starts much closer to h4/g5 than White's does to b5. Hence the winning plan begins with heading the king toward g4/g3.
Claims extracted from it
2 true · 2 false · 4 unverifiable
Hover a claim to light up the squares it is about.
- Unverifiable
Black's queenside pawns are all frozen
unverifiable
- True
h4 is undefended
defendedsquare=h4negated=true
- True
g5 is only defended by h4
defendedsquare=g5by=h4
- False
marching the king via f4–g4/g3 lets Black capture h4 first, then g5
linemoves=Kf4 Kg4/Kg3 Kxh4 Kxg5assertion=type:captures_piece piece:pawn
- False
Black nets two pawns and creates an easy passed g-pawn
linemoves=Kf4 Kg4/Kg3 Kxh4 Kxg5assertion=type:material_delta side:black min_value:2
- Unverifiable
White's king is too far from the kingside to intervene in time
unverifiable
- Unverifiable
any White queenside break (Kc4xb5) is slower than Black's kingside harvest
unverifiable
- Unverifiable
Black's king starts much closer to h4/g5 than White's does to b5
unverifiable