DrMAL: 1988 GM World cup tournament had huge personal impact it was highlighted by Beliavsky vs Kasparov, 1988 one of greatest games ever played. But Kasparov played several other great examples to learn technique from, his games in this tounament and 28th Chess Olympiad in Greece that year were primary lessons for me to acheive candidate master the following year. I will continue posting on several as attempt to help others with similar ambition to advance in level.
This game, like Kasparov vs Campora, 1988 is simpler than others and good place to continue. From 1984 WC match Kasparov started adopting Najdorf move order (5.a6) in Scheveningen to prevent Keres (see Karpov vs Kasparov, 1985 this in fact changed whole theory of Scheveningen defense) but here against Ljubo he did not bother, perhaps he felt more confident, he knew this opening much better than anyone else anyway.
Often 6...h6 is played to stop 7.g5 but Kasparov chose developing alternative. After obligatory 7...Nd7 Ljubo tried novelty 8.Rg1 instead of usual 8.Be3 other moves such as 8.h4 have also been sucessfully played (e.g., Bronstein vs Jansa, 1979).
Kasparov had surely studied this option but in any case his
8...Be7 wa strongest as was castling right into it on move 9 (line became part of theory with transpositions to earlier lines where 10...Nde5 and 10...Nxd4 are main alternatives against 10.Qd2). Related game that year was Akopian vs Shabalov, 1988).
After long castle and start of double-edged mutual attack, played very accurately by both sides, center was fluid making position even sharper. Here, 16.Be3 was most accurate, move is in notes from that year and Houdini scores it slightly better today. But on next move no doubt 17.Qe3 was erroneous (17.Bg2 looks and scores best with 17.Ng3 also good) allowing 17...Be6! to achieve equality via 18.Nc1 occupying white N for defense. 20.Bd4 was also slight inaccuracy compared with 20.Nb3 right away now black got some edge.
In moves that followed Kasparov was simply more accurate, making small unnoticeable gains with each move, computer with top engine (Houdini) is really required to appreciate subtleties here. Nonetheless game can be understood quite simply by looking at activity of pieces after center was opened. Black position was already great but attack clearly increases in tempo with continuing initiative. After 26...b3! only good move was 27.e5 (27.Bxf8 may also survive) and 27.c3?! was decisive mistake. Of course Kasparov played winning move 27...Rcd8! then chose one of several ways to finish. It is also instructive, particularly for intermediate club player, to carefully play out various options on move 28 and after, game at this point shows many nice tactical solutions. Verifying with computer only after trying hard with real board and pieces (writing down lines as one plays on board) is only method I recommend.