"A bit like Ian Nepomniachtchi after his loss to Carlsen in game 6 of their World Championship match, the long loss may have broken Keymer’s spirit, and he was completely outplayed and outclassed in game 4."
A possible interpretation, but maybe it didn't really matter _how_ Keymer had lost game 3, only _that_ he lost it to be in a must-win situation. He had to "take risks" in game 4, and it backfired completely. As often in such situations, games like Erigaisi-Le are relatively rare. Then there's also no need to discuss what is more tiring: a rapid game lasting about an hour with the next game starting minutes later, or a classical game lasting many hours with time to eat, sleep and try to recover before the next game.
"Carlsen refutes the idea at the board" - maybe, but it's also conceivable that he had at least a quick look at the consequences of the white pawn grab before. He had this variation three times before, with black against Artemiev as mentioned in the game notes, with white against Mamedyarov and Dubov. Funny that he had lost with white in rapid against Dubov also due to a bad pawn grab, under different circumstances: 31.Qxa5?? was well in a "middlegame-like (king safety mattered) endgame with queen and one rook for both sides".
It depends on whether 9.Qa4+? is dismissed immediately on human grounds or engine verdict of -1, or whether you check it anyway - being ready for it with black or even checking whether it may still work with white. The entire concept for black also wasn't too different from what black did in the three earlier games.
"A bit like Ian Nepomniachtchi after his loss to Carlsen in game 6 of their World Championship match, the long loss may have broken Keymer’s spirit, and he was completely outplayed and outclassed in game 4."
A possible interpretation, but maybe it didn't really matter _how_ Keymer had lost game 3, only _that_ he lost it to be in a must-win situation. He had to "take risks" in game 4, and it backfired completely. As often in such situations, games like Erigaisi-Le are relatively rare. Then there's also no need to discuss what is more tiring: a rapid game lasting about an hour with the next game starting minutes later, or a classical game lasting many hours with time to eat, sleep and try to recover before the next game.
"Carlsen refutes the idea at the board" - maybe, but it's also conceivable that he had at least a quick look at the consequences of the white pawn grab before. He had this variation three times before, with black against Artemiev as mentioned in the game notes, with white against Mamedyarov and Dubov. Funny that he had lost with white in rapid against Dubov also due to a bad pawn grab, under different circumstances: 31.Qxa5?? was well in a "middlegame-like (king safety mattered) endgame with queen and one rook for both sides".
It depends on whether 9.Qa4+? is dismissed immediately on human grounds or engine verdict of -1, or whether you check it anyway - being ready for it with black or even checking whether it may still work with white. The entire concept for black also wasn't too different from what black did in the three earlier games.