Author: Jon Dart
Date: 09:16:19 05/28/01
This is the game:
[Event "CCT-3"]
[Site "?"]
[Date "2001.05.27"]
[Round "?"]
[White "Arasan 6.0"]
[Black "Tinker"]
[Result "0-1"]
[ECO "A29"]
[WhiteElo "2376"]
[BlackElo "2372"]
[TimeControl "2700+10"]
1. c4 Nf6 2. Nc3 e5 3. Nf3 Nc6 4. g3 Bb4 5. Bg2 O-O 6. O-O e4 7. Ng5
Bxc3 8. bxc3 Re8 9. f3 e3 10. d3 d5 11. Qb3 Na5 12. Qa3 c6 13. cxd5
cxd5 14. f4 Nc6 15. Rb1 Qc7 16. Bb2 Bg4 17. Nf3 Qd7 18. c4 d4 19. Ba1
Bxf3 20. Bxf3 b6 21. Qa4 Rac8 22. Bxc6 Rxc6 23. Bxd4 Ng4 24. f5 Nf2
25. Qb5 a6 26. Qb2 Rh6 27. Bxb6 Rb8 28. Qe5 Qb7 29. Qxg7+ Kxg7
30. Bd4+ Rf6 31. Rxb7 Rxb7 32. Bxe3 Ng4 33. Bd4 Re7 34. Rf4 h5 35. h3
Ne5 36. g4 hxg4 37. hxg4 Rc6 38. Re4 f6 39. g5 Rc8 40. Bxe5 Rxe5
41. Rxe5 fxe5 42. Kf2 Rc5 43. a4 Kf7 44. e4 Ra5 45. g6+ Kf6 46. Ke3
Rxa4 47. Kd2 a5 48. c5 Rb4 49. Kc3 Kg7 50. c6 Rb6 51. c7 Rc6+ 52. Kb3
0-1 {ArasanX resigns}
Arasan-Tinker was an English Four Knights (A29), which is an
opening Arasan plays quite a lot.
Arasan won a pawn here with a simple tactic:
[D] 4r1k1/p2q1ppp/1pr2n2/8/Q1Pp1P2/3Pp1P1/P3P2P/BR3RK1 w - - 0 1
But Schroer thinks maybe it shouldn't have taken it. In fact
the advantage was quite fleeting as Tinker got a Knight on f2.
Arasan expected Nxd3 here:
[D] 4r1k1/p2q1ppp/1pr5/1Q3P2/2PB4/3Pp1P1/P3Pn1P/1R3RK1 b - - 0 1
No idea if that's a good idea or not. It's complicated.
The middlegame was a lot of complex manouvering.
Crafty thinks Tinker missed something at move 27:
[D] 4r1k1/3q1ppp/pB5r/5P2/2P5/3Pp1P1/PQ2Pn1P/1R3RK1 b - - 0 1
({14:+0.64} 27. ... Rb8 28. Qe5 Qb7 29. Qxb8+ Qxb8 30. Bxe3 Nh3+ 31. Kg2 Qa8+
32. Rf3 Rh5 33. Rb6 Ng5 3\
4. Bxg5 Rxg5 35. e4 Qc8 36. Kf2 $14)
({14:-0.96} 27. ... Nh3+ 28. Kg2 Qc6+ 29. Rf3 Ng5 30. Bd4 Nxf3 31. exf3 Rd6 32.
Re1 e2 33. a4 Red8 34. \
Bxg7 Rxd3 35. Rxe2 Qxf3+ 36. Kh3 Qxf5+ 37. Kg2 $17)
At move 28:
1r4k1/1q3ppp/pB5r/4QP2/2P5/3Pp1P1/P3Pn1P/1R3RK1 w - - 0 1
Arasan went for Qxg7+, forcing a series of exchanges,
winding up a whole rook down, but with a couple of connected passers
as compensation. Not a good decision, in retrospect..
Crafty thinks Qxb8+ here is much better:
({15:+1.79} 29. Qxb8+ Qxb8 30. Bxe3 Nh3+ 31. Kg2 Qa8+ 32. Rf3 Rh5 33. g4 Rh4
34. Kg3 Qd8 35. Ba7 Ng5 36\
. Rb8 Rh3+ 37. Kg2 Rxh2+ 38. Kxh2 Nxf3+ 39. exf3 $18)
Later Arasan traded its remaining pieces and it was the rook against
the passers. It still thought it was ok, but the king is behind
the passers, which still haven't advanced even a square. The
rook begain mop-up operations, and shortly after the score went
to -7 and I resigned for the program (it is reluctant to resign
while still having passers).
Crafty still thinks White is ok here:
[D] 8/5k2/p7/2r1pPP1/P1P5/3P4/4PK2/8 w - - 0 1
and recommends Kf3 or Ke3, with a + score. Crafty in fact gives
more of a bonus for the passers than Arasan does.
Tinker starting giving reasonable scores in this endgame quite a
while before Arasan, which was oblivious until near the end. As
noted elsewhere, though Tinker was having problems with thrashing
on the endgame TB probes. Arasan wasn't hitting the TBs hardly
at all (it only had 4-man tables).
I'm far from a strong endgame player myself, but I could look
at this and say, "the pawns aren't far enough advanced, the King
is behind the pawns, and he's got a whole rook. Doesn't look good."
I made a comment during the game that Arasan was too optimistic.
On the other hand, it's risky to tweak the eval to "fix" this.
One thing I can try: Arasan gives a bonus for trading material
when it has passers. It's not good to do that if the opponent
has a rook.
--Jon
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.