[Event "U.S. Championships"] [Site "St Louis"] [Date "2023.10.05"] [Round "2"] [White "Eswaran, Ashritha"] [Black "Tokhirjonova, Gulrukhbegim"] [Result "0-1"] [WhiteElo "2273"] [BlackElo "2350"] [Annotator "Lang,JJ"] [Variant "From Position"] [ECO "?"] [Opening "?"] [FEN "1rk5/4b3/2p3p1/2p1p3/p1PrP1pP/PbR1KP2/1B1N4/7R w - - 0 41"] [SetUp "1"] [Source "https://lichess.org/study/b0nfXV2i/hA981NN0"] [Orientation "white"] { [Annotations by JJ Lang] Right after reaching the time control, and after defending a cramped position for the majority of the game, White makes an inaccuracy that portends what comes next. } 41. Nxb3?! gxf3! 42. Kxf3 (42. Nd2 Rxb2 { is the point, explaining why there was no need to recapture the knight immediately. }) 42... axb3 43. Rcc1 Rd2!? { Practically, this is a very testing move that ended up winning the game. But, interestingly enough, the engine is convinced that Black's most promising try is a very Botvinnik-ian Exchange sac!! } (43... Kb7!! 44. Bxd4 cxd4 45. Rb1 Ka6 { when White's four weak pawns, coupled with Black's two passers, provide Black with excellent winning chances. }) 44. Bxe5?? (44. Bc3! { and Black's rook cannot stay on the second rank, as after } 44... Ra2 (44... Rd8 { Of course, Black need not stay on the second rank. But her position is less venomous now: } 45. Bxe5 Rb7 46. Rb1 Rd3+ 47. Ke2) 45. Bxe5 { things are equal: } 45... b2 46. Rb1 Rb3+ 47. Kf4 { Note that, in the game, Black's rook was able to swoop to g2 and prepare ... Be7xh4 with the idea of ...g6-g5+. Shutting the rook out first was essential. }) 44... b2! { With the rook on d2 (rather than a2 after 44. Bc3!), Black can push her pawn without shutting the rook out of the game. } 45. Rb1 Rb3+ 46. Kg4 Rg2+ 47. Kf4 Bxh4 48. Bg7 (48. Rxh4 g5+ $19) 48... Kd7 { Short of giving up an entire rook by capturing on b2, there is no way to stop Black's king from constructing a mating net. } 49. Rbd1+ Ke6 50. Bh6 g5+ { 0-1 Black wins. } 0-1