Stories have a way of writing themselves. As an example, I give you two games from the 7th round, Gelfand-Kramnik and Carlsen-Radjabov.
I’ll start with the Gelfand-Kramnik game. Here’s the position after Kramnik’s 18…Nf6-e8.
rq2n1k1/1b3ppp/pp1bp3/8/3PN3/3B1NP1/PP2QP1P/2R3K1 w - - 0 19)
ChessVibes’s writeup has the following: “For a moment Kramnik was in big trouble, but he escaped with a draw when his opponent Boris Gelfand of Israel refrained from playing actively on move 19.”
That’s true – White can win with the very nice 19.Neg5! g6 20.Nxf7! Kxf7 21.Ng5+ Kf6 22.Qxe6+ Kxg5 23.Qh3!! (the only move to win). Of course, White’s work is not yet done, for example 23…Kf6 24.Qxh7 Bb4 (covering e1, freeing the Qb8 to come into the game), he should find 25.Rc2!, winning.
There are other lines to consider too, such as 19.Neg5 h6 20.Bg6!!, or 19.Nfg5 h6 (19…g6 20.Nxf7 would transpose to the above line) 20.Qh5 hxg5 21.Nxg5 Nf6 22.Qxf7+ (which leads to a strange pawns vs piece endgame, with some other material on the board).
Is White objectively winning after 18…Ne8? Yes. Is it trivial to find? Not especially, I think. It’s not hard to see that a knight jump to g5 is possible; and it’s not hard to keep looking (or as Mark Crowther puts it at TWIC, “Gelfand could have played it and tried to make it work move by move”), but it helps to know the outcome in advance when you go down that route.
Meanwhile, the same ChessVibes writeup says about the Carlsen game, “Against Teimour Radjabov of Azerbaijan, Carlsen needed to sacrifice an exchange to wear [sic] off dangerous threats against his king, which proved to be sufficient.”
Ah yes, the sufficient exchange sacrifice to ward off dangerous threats.
3r1r1k/p3q1pp/2pb4/2p1p3/2P1Q3/BPNP1pP1/P2R1PnP/4R1K1 b - - 0 23)
Carlsen got sidetracked on the queenside and found himself allowing …f3/…Ng2, leaving Black with a won position. It’s most simply winning with 23…Nxe1 24.Qxe1 Bc7! – a very nice redirection with a piece that hasn’t been terribly useful so far in the game. The problem with taking the exchange on e1 or e3 (as in the game) is that White’s knight can become a rather nice blockader on e4 or f2. But with …Ba5 in the air, White doesn’t have enough time to clear the diagonal and install the knight safely on e4. Once the knight is traded though, there’s no holding Black back – the pawn on f3 is a monster, the d3-pawn will fall eventually, and Black’s rooks will get into White’s position.
As played, 23…Qe6 forced White to play 24.Re3, after which 24…Nxe3 25.fxe3 saw White again trying to set up a blockade, this time on f2 with Nd1-f2. Whether or not this is winning for Black is hard to say; probably it is in the long-term, but again, it’s not trivial. Radjabov thought he saw a way forward with the thematic 25…f2+ (opening lines for the rooks, hence thematic) 26.Rxf2 Rxf2 27.Kxf2 Rf8+ 28.Ke2 Qh3, but 29.Qh1! was a cold shower after which White is holding.
Both games were winning for one side. Both games ended up drawn. Neither was trivial to win, with the winning lines being relatively more tactical and positional, respectively. But while Radjabov missed the …Bc7-a5 positional win, he may have been able to win as well without forcing things with …f3-f2+.
But while Kramnik was lucky to survive, the Carlsen game instead becomes, as TWIC puts it, “[a] fascinating yet flawed game that shows just how confident Carlsen is in his own abilities and perhaps provides a better insight into the way Carlsen tries to drag his opponent into trouble than games where this works out.”