Having already effectively ensured the $65,000 first prize after just three games, it's quite possible that Kasparov slightly lost interest. He may also to some extent have been affected by the current imbroglio over his projected "title defence" against Alexei Shirov (of which, while there are various rumours about possible sponsors, nothing concrete has yet emerged), though surely not as much as poor Alexei in Spain. But there is also a long-term underlying pattern at work.
Kasparov has always operated in bursts, with periods of phenomenal energy interspersed with calmer moments. All chess players need some "permission" from their opponents in order to win games and Timman's excellent opening preparation denied Kasparov this. The critical moment came in the fifth game after Timman, despite being White, had had to work a bit to hold the fourth.
The opening was quite similar to the third game, but this time Timman got what he was clearly angling for: a repeat of Kasparov's win last year, during his simultaneous against the Argentine team, no less, against veteran grandmaster Oscar Panno.
There Kasparov had uncorked the phenomenal 11.Qd3, the move which won the vote of a panel of experts as the most important theoretical novelty of volume 70 of the prestigious Belgrade-based periodical Chess Informant. The point is to keep the queen on the d file so that if 11... Nxg5 12 h4 g6 13 hxg5 Qxg5 White can take with his queen on d6, rather than with the rook as would be the case if she had retreated to c2.
Panno had played 16... Nd7, and after 17 dxe6 Bxe4 18 e7! Re8 19 Rxd7 he got well beaten. In his notes to that game, Kasparov dismissed Timman's choice of 16... exd5 with the single move 17 Nc3, but that proved to be far too simplistic. After 17... Nc6! 18 Bxd5 Na5, both sides had somewhat bad pawn structures, and Timman drew without difficulty.
White: Gary Kasparov
Black: Jan Timman
Prague (Game five) 1998