MNb, 

This time, it was a pure typo.  Thanks for pointing it out though!

Here is the game Van der Wiel - Timman:

[Event "Brussels SWIFT"]
[Site "Brussels"]
[Date "1986.??.??"]
[Round "?"]
[White "Van der Wiel, John TH"]
[Black "Timman, Jan H"]
[Result "0-1"]
[ECO "C06"]
[WhiteElo "2555"]
[BlackElo "2645"]
[PlyCount "110"]
[EventDate "1986.03.??"]
[EventType "tourn"]
[EventRounds "11"]
[EventCountry "BEL"]
[EventCategory "13"]
[Source "ChessBase"]

1. e4 e6 2. d4 d5 3. Nd2 Nf6 4. e5 Nfd7 5. c3 c5 6. Bd3 Nc6 7. Ne2 cxd4 8. cxd4
f6 9. Nf4 Nxd4 10. Qh5+ Ke7 11. exf6+ Nxf6 12. Ng6+ hxg6 13. Qxh8 Kf7 14. Qh4
e5 15. Nf3 Nxf3+ 16. gxf3 Bf5 17. Bxf5 gxf5 18. Bg5 Qa5+ 19. Kf1 g6 20. Bxf6
Qa6+ 21. Kg2 Qxf6 22. Qxf6+ Kxf6 23. Rac1 Bd6 24. Rhd1 Ke6 25. Rc3 g5 26. Rdc1
Kd7 27. Rb3 b6 28. Rd3 d4 29. h4 gxh4 30. f4 Rg8+ 31. Kh3 Ke6 32. fxe5 Bxe5 33.
Re1 Kd5 34. Rf3 Rg5 35. Kxh4 Rg4+ 36. Kh3 Re4 37. Rg1 f4 38. Rb3 Re2 39. Rg5
Kc4 40. Rg8 Bd6 41. Kg2 Bc5 42. Rg4 a5 43. Kf1 Rc2 44. Rxf4 a4 45. Rc3+ Rxc3
46. bxc3 Kxc3 47. Re4 a3 48. f4 d3 49. Ke1 Kb2 50. Kd2 Kxa2 51. Kxd3 Kb3 52.
Rc4 Bb4 53. Rc1 a2 54. f5 b5 55. f6 Kb2 0-1

I think the position after 26...Kd7 is worth a diagram:



Obviously Timman did not think this is that unpromising for the 1st player, as he went into this exact line again against Morozevich in 2000, but this time as White! (The game differed with 18...Rc8?) It can be found here commented by NM on chesspublishing, in which the game Van der Wiel - Timman is mentioned in the notes. I allow myself to quote  from this game: (from diagram position) 27. Rb3 b6 28. Rd3

Quote:


If this is indeed true and black cannot go for more without doing any dubious moves, then  this certainly looks like an unpleasant line to play against people eager to draw!

Obviously we must look closer on the 28...d4 29.h4-line to investigate if there is anything better, or if there is anything bad with it all. But seeing Timman`s comment after the game make me think that he was not entirely happy with his position during the game.

"Van der Weil"
Ouch, Dutch is such a difficult language. The GM's name is Van der Wiel.
Wiel means wheel.
I agree, that Van der Wiel-Timman is the model game.

Klick,

My source is Psakhis' 2003 book on the Tarrasch:

Following the game you cite above, Psakhis mentions

22. Qa4

(22. Qf6+ looks dubious according to Psakhis who cites Van der Wiel - Timman, Brussels 1986)

22. ... Qc6
23. Qb3 

... and now, instead of Bh6 as in Belov - Kozyrev, St. Petersburg 2001 or Belov - Smikovsky, Berlin 1998, Psakhis suggests ....

23. ... Bc5 as in Kurnosov - Iljushin, Samara 2002

24. Rac1 b6
25. Rhe1 Kf6
26. Qe3 Qd6
27. Qh6 a5
28. Rg1 Qf8
29 Qh4+ g5
30. Qa4 Qe8

P.S.: I believe that chesspub only mentions the Smikovsky but not the Kozyrev game

I prefer 3...Nf6 for Black, but 3...c5 and even 3...a6 have been effectively played in the past.  Klick, your question about 9.Nf4 had been answered pretty definitively several years ago:  White gets a small material plus for a large positional deficit.  However, recent games suggest that the position may yet be playable for White after all.  I personally still prefer Black, but now have to do more homework to find specific, concrete lines to prove it.   

I used to love playing against 9.Nf4 because it required an understanding of the position rather than specific moves, but now even this variation is beginning to get fritzified.  The final conclusions on this line will probably never be completely resolved.  But I believe in the old Steinitzian idea that if you haven't made any opening mistakes, any attack against you should be doomed.  (Yeah, I like to defend!)

I would have thought that the traditional view is that 3...c5 is the best.  For instance, Raymond Keene annotating Korchnoi's 3...c5 in one of his candidates matches against Karpov:  "No nonsense from Korchnoi.  Other 'interesting' moves are 3...Nf6 and 3...Nc6, but White is better in both cases."  It seems to me that 3...Nf6 has gained some theoretical ground in the years since then, but this may be the first time I've ever seen the opinion that it is better than 3...c5.  

The 3...Nf6 lines has traditionally been black`s move, although the Guimard is interesting. In many lines of the Nf6-Tarrasch, black gains dynamic play with good winning-chances after the exchange-sac on f3. Excellent if you want to try for a win! M.Gurevich has tremendous results with black in this variation. Take a look at some of his games and you will se what I mean.

However, is there any consensus on the 9.Nf4-move? How can black best play for a win without going into dubious territory with for instance 15...e4 in the mainline? I`ve found it a little hard to maintain reasonable winning-chances when playing against lower rated opponents....

Almost with 5.Nxe4 dxe4 6.Bc4 c5 (a6 7.a4 is no improvement) 7.d5 Nd7 8.dxe6 fxe6 9.Nh3!