Author: Amir Ban
Date: 13:43:31 08/30/98
Go up one level in this thread
In the first game the issues were relatively simple. The second game was a
different matter. The game was very difficult for both sides. Junior needed a
lot of tactical power and understanding of the position just to stay out of
trouble in this game. It ultimately failed in this, but in analyzing the game
(and I'm not done with this yet), I'm wondering how it managed to get so far.
The game was not easy for Yudasin either, and he was not hiding this fact, and
was of the opinion that many other GM's would find the game difficult. He had,
though, the advantage of playing one of his favorite setups, which, as he told
us, he had played and won against Salov, A.Greenfeld and others.
After 7.Nc3 Junior was out of book. 7.0-0 is the more usual move, and indeed
Yudasin confessed to have mixed up the order of moves. Before playing 7...Qc7
Junior considered for a long time the weird plan 7...Bd6 8.e5 Bb8.
10.Nxc6 was clearly intended as a provocation. At shorter time controls, Junior
would have fallen and played the highly problematic Bxh2+. On the next move,
Yudasin showed with 11.Qh5 that he knew this opening well and was going exactly
where he wanted to. 11... Be5 is necessary to prevent e5, which would make
black's position already hopeless.
Junior probably showed good judgement in refusing to castle here, but not
castling is also a problem of course. In considering 13...g6 Junior was briefly
excited at the prospect of 13... g6 14.Qh4 d5, and apparently gaining a big
advantage because of the threat to the pinned Nc3, before realizing 15.exd5 cxd5
16.Nxd5 that it is in fact the black Be5 that is pinned. Tricky :( After 14...d6
Junior expected 15.Bb2 Qe7 16.Qxe7+ Kxe7 with slight advantage for white.
Yudasin said he would have played this if not for the king's position on e7
which would be fortunate for black and make this hard to exploit.
On move 19 everybody on ICC (and Junior too) expected the GM to play f4. Junior
expected 19.f4 Bg7 20.e5 and evaluated about -0.25. Yudasin gave this move a lot
of thought, and after 19.Qg4 we told him everyone expected f4. He smiled and
said: "Of course they do, but I have my style".
20...h4 caught Yudasin off guard. He was clearly surprised, said so, and later
in the game complained several times that his 20.Qg5 is inaccurate and he should
play Qh4. What greatly astonished me is that instead of agonizing for half an
hour (he really had plenty of time), he played 21.h3 in a matter of seconds,
with a smug comment that no computer would ever understand this move.
Junior's time to blow a big chance came two moves later, when it played
22...Rh5?, preferring it narrowly over 22...Qc5. Yudasin immediately pointed out
that Rh5 is much more effective after Qc5, which makes the black queen active
and pushes the white queen to the inactive g4 square. The reason Rh5 was
actually played is probably related to some deep tactics miscalculated. It
expected to answer 23.Qe3 with 23...c5 but later greatly lowered its opinion of
this move and was forced into the silly-looking 23...Rb7 (to stop the white
queen from penetrating).
I noted that several people here thought 25...Bh8 to be the losing move. Neither
we nor the GM thought so during the game. Indeed, Bxf4, which was almost played
by Junior was the real loser. Shay commented while waiting for this move that if
it is played, we may as well resign and go to bed. Yudasin agreed. Without the
dark-squares bishop black is lost. Of course 25...Bg7 is better, but Junior
expected the continuation 26.b4!? Bxc3 and so didn't care about the bishop's
"temporary" parking at h8. I doubt Yudasin considered b4, because next move he
was mildly surprised by 26...Qb4, which is of course necessary to prevent b4
which is now fatal.
The next moves till 31...d4 I thought was some nice regrouping by Junior,
incidentally managing to avoid some cheap tactical thrills of grabbing a pawn
and probably losing immediately. 34...Bg7 should probably have been played one
move earlier, and white may have missed 34.Bc1 a4 35.Ba3 Qg5 36.f4.
After this the GM played a small waiting game. He had a time advantage, and
didn't mind waiting to see what the computer wanted to do. Going into the faster
pace of the second time control segment, Junior immediately delivered by making
the losing error 41...Bb7? As Yudasin said, this looks like a direct
provocation: Come on, play b4, make my day.
Junior, though, did not expect b4, and didn't see until 44...Qxd3 that it was
losing. Why ? Yudasin did not understand what he did to cause the program to
collapse, and played the last winning moves with great caution, though Junior
had no more resistance to offer of course.
The reason was that white was actually offering here two rooks for a queen
(black can play 45...Qxb1+ rather than Rxb7). Accepting this exchange is quite
hopeless, since with one of the rooks on h5, the passed a-pawn would win
quickly. This was so obvious to the GM, that he was not even conscious of making
an offer of material. Junior became aware that this "winning of material" leads
to a quick loss only when it was too late, since the a-pawn is unstoppable.
As I said, I only wonder that it didn't fall for something like this sooner.
Chess is really a tough game for computers.
Amir
[Event "ICC 120 0 08/27/1998"]
[Site "Internet Chess Club"]
[Date "1998.08.27"]
[Round "-"]
[White "*Yudasin"]
[Black "*Junior"]
[Result "1-0"]
[WhiteElo "0"]
[BlackElo "0"]
[ECO "B30"]
[NIC "SI.31"]
[LongECO "Sicilian: Nimzovich-Rossolimo attack (without ...d6)"]
1. e4 c5 2. Nf3 Nc6 3. Bb5 e6 4. O-O Nge7 5. Re1 a6
6. Bxc6 Nxc6 7. Nc3 Qc7 8. d4 cxd4 9. Nxd4 Bd6 10. Nxc6 bxc6
11. Qh5 Be5 12. Re3 Rb8 13. b3 g6 14. Qh4 d6 15. Rf3 Qa5
16. Bd2 Qa3 17. Rb1 h6 18. Rd3 a5 19. Qg4 h5 20. Qg5 h4
21. h3 Ba6 22. Rf3 Rh5 23. Qe3 Rb7 24. Qe1 d5 25. Bf4 Bh8
26. e5 Qb4 27. Na4 Bb5 28. c3 Qe7 29. Nb2 c5 30. c4 Bc6
31. Re3 d4 32. Re2 Ra7 33. Nd3 Ra8 34. f3 Bg7 35. a3 Bf8
36. Kh1 Qb7 37. Kh2 Qa6 38. Rc2 Qc8 39. Kh1 Qc7 40. Qf2 Rb8
41. Rcb2 Bb7 42. b4 cxb4 43. axb4 Qxc4 44. bxa5 Qxd3 45. Rxb7 Rxb7
46. Rxb7 Be7 47. Rb8+ Bd8 48. Qb2 Qf1+ 49. Kh2 Qa6 50. Bd2
{Black resigns} 1-0
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.