ex0duz: <Fusilli: Gukesh sacked a pawn as part of his overall strategy to keep the initiative. Later, for Gukesh the problem was not that the b7 pawn was poisoned, but that Ding managed to end up with a superior minor piece. His knight is so much better than Gukesh's bishop. And the pawn structure favored the maximization of the power of Ding's knight. Ding ended up with a winning attack as a result of his positional superiority.>Oh don't get me wrong, I agree with your general assessment of the game and that Ding obviously had a superior position on the board(to state the obvious lol) and that the knight was much more useful than Whites bishop by the end of the game. That is obviously undeniable, but that is also in hindsight. My point however was that white going pawn snatching(ie taking the poisoned pawn) and using 2 tempo to do so while black went king hunting also played a massive part in the result. To me this was a game of tempi, and blacks gradual accumulation of free tempi which allowed him to get a better position as you said, and ultimately pull off his attack first.
Whites first mistake and loss of tempo was 16.Rd2. After that engine gives like -0.7. But that is not really game ending in itself and just an inaccuracy. Ding then offers his own pawn on b7 with 22.Qg5. Gukesh refused to take it like I said, instead choosing to play 23.Qb5 and defend his knight. This was Gukesh/White's second mistake, and the eval goes from -0.7 to around -1.4. I guess you could count this as a loss of tempo since white could have used that tempo and taken a pawn, but at least its defending the Knight on f4 so ill let it slide as just another inaccuracy.
Now things are starting to get serious. Half a pawn(-0.7) is still manageable or drawable in many games, but a pawn and a half(-1.4) is probably going to end with a loss most of the time.
So 22.Qg5 is <"Ding sacced a pawn of his own".>
As for the <"Then few moves later he took it but that turned out to be a poisoned pawn and Ding crushed him shortly after that with great attacking move">
I was referring to Gukesh refusing to take it on move 22 when it was actually the best move and still -0.7 after taking, but instead Gukesh decided to take it on move 28 with 28.Qxb7, by which time the "ripe fruit" had already gone off and became a "poisoned pawn" hah.
Yes, whites position was already quite difficulty to play by that point and the engine gives a -2.0 eval after the recommended 28.Bf4 instead of pawn snatching.
But he indeed goes for the poisoned pawn with 28.Qxb7, and it now gives an eval of -2.6. But even without the eval, we can just try to assess the logic behind the move. By taking that pawn, Gukesh basically gives Ding 2 free moves.
First 28.Qxb7 gives Ding a free 28.Re8, and then after Gukesh plays 30.Qb5, moving his queen back to the exact same square before he snatched the pawn, he allows Ding a free 30.Na2+, leading to a forced line where black checks whites king and allows the knight and queen to infiltrate the squares around whites king, and also with the Re8 cutting off any potential escape routes for the white king.
These two tempo literally allowed Ding to turn what was "just" a -2.0 positional advantage into a crushing attack with all Dings pieces coming into play. After the forced checks and 32.Qg1, engine gives -3.5, and it was here where Gukesh blunders in a hopeless position with the move 33.Kd3.(-7 after).
Ultimately in my opinion this game was less about positional domination, and more about who had the faster attack which is usually the theme in games with opposite sided castling. White wasted at least 3 tempo, first with 16.Rd2 then 28.Qxb7, and then 30.Qb5. in this type of opposite side castling game where both sides are attacking each other's king, one tempo is huge, let alone 3. Gukesh going pawn grabbing on b7(especially after he refused it few moves earlier when it was THE best move) was not logical and in the end he paid the price for being indecisive, illogical, and wasting too many tempo on useless moves. He also sacced a pawn to gain a tempo on the queen with moves like 18.RHe1, opening the center, but in the end it's black who controls the E file with his rook while white has his confused rook on d2 still not knowing what it's doing there, and if anything gets in the way during the game and would be better placed on the back rank to stop blacks queen from infiltrating it with check near the end.
Kibitzing
