First ever defeat of an IM at long time control in a match!

[Harvey Williamson]

1. Who was playing in this game?
2. When did it take place?
3. What was the result of the match?
4. What is unique about the result of the match?

1. e4 e6 2. d4 d5 3. Nd2 Nf6 4. e5 Nfd7 5. Bd3 c5 6. c3 b6 7. Ne2 Ba6 8. Bxa6
Nxa6 9. Nf3 Be7 10. Qa4 Nc7 11. dxc5 bxc5 12. Bf4 O-O 13. O-O a5 14. Rad1 Nb6
15. Qc2 a4 16. Ng3 Qb8 17. Nh5 Ra7 18. Bg5 Ne8 19. Nf4 Bxg5 20. Nxg5 g6 21.
Rfe1 Ng7 22. Nd3 Qc7 23. Qe2 a3 24. b4 Na4 25. Rc1 Rc8 26. Qd2 Qd8 27. Nf3 cxb4
28. cxb4 Rc4 29. Nc5 Nb6 30. Nd4 Rxc1 31. Qxc1 Qh4 32. Ncb3 Nc4 33. g3 Qg4 34.
f3 Qh5 35. Na5 Rc7 36. Nb5 Rc8 37. Qc3 Nb6 38. Nc6 Kh8 39. Nd6 Rc7 40. Qc5 Nd7
41. Qa5 Ne8 42. Qa8 Qxf3 43. Nd4 1-0

[Debaser]

1997 HIARCS 6.0 defeated I.M. Deen Hergott 4-2 in a match at long time controls, Ottawa, Canada (first PC in history to win a match against a FIDE I.M.)

From here http://www.hiarcs.com/hiarcs_milestones.htm

[Harvey Williamson]

1997 HIARCS 6.0 defeated I.M. Deen Hergott 4-2 in a match at long time controls, Ottawa, Canada (first PC in history to win a match against a FIDE I.M.)

From here http://www.hiarcs.com/hiarcs_milestones.htm

Yes and the games were played at a long time control but on what would be considered now antique hardware(h/w was 200Mhz Pentium MMX) -

Other Hiarcs achievements are here http://www.hiarcs.com/hiarcs_milestones.htm

Here are all the games with comments:

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "1"]
[White "Hiarcs6.0"]
[Black "Hergott, Deen"]
[Result "1-0"]
[ECO "C05"]
[PlyCount "85"]

1. e4 e6 2. d4 d5 3. Nd2 Nf6 4. e5 Nfd7 5. Bd3 c5 6. c3 b6 {Though this move
is well known, it is considerably less popular than the more direct 6..Nc6 7.
Ne2 cxd4 8 cxd4 Qb6 followed by 9..f6 attacking the White centre. Black
forgoes the opportunity to put pressure on the White position in order to
achieve a basic positional aim in the French: the exchange of the 'bad'
light-squared bishop which is restricted by Black's pawn chain on e6 and d5.
This is rather slow as Black neglects his kingside development but as a
leading French export Lev Psakhis says, "strategic achievements can be worth
the price of a little incovenience."} 7. Ne2 {The last move in Hiarcs' book so
an interesting test of the computer's planning abilities! The game
Akopian-Hergott Belgrade GMA 1988 had continued 7 Qg4!? Ba6 Bxa6 Nxa6 9 Ne2 h5
10 Qg3 b5 11 0-0 with a slight advantage for White.} 7... Ba6 8. Bxa6 Nxa6 {
This is the first interesting point in the opening. Since the light-squared
bishop is Black's bad bishop, it is natural to assume that Hiarc's first move
out of book was a poor one, exchanging White's 'good' bishop ( the one on the
opposite colour to his own pawn chain) for Black's bad bishop. Well, 8 Bb1
would have avoided the exchange of bishops and is fairly reasonable, but
probably 8 Bxa6 is the best move. Confused? Well the point is that when
assessing an opening, we have to balance achievements that will be useful in
the long term with the problems that we have to face right now. The exchange
of light-squared bishops is something that Black will enjoy in the endgame. It
is an investment for the future. However, in the present, Black has a few
worrying problems. The first is the time that he has spent achieving this aim:
the net result is that the knight is offside at a6, generally contributing
very little to the game. The second is that Black has done nothing to break
White's main advantage in the French which his pawn wedge on d4 and e5. We see
this structure so often that it is easy to forget its strengths: The pawn on
e5 gives White a central and kingside space advantage; it provides an outpost
on d6 for a White minor piece; and most importantly of all, it prevents
Black's king's knight from coming to f6. This means that Black's kingside
lacks its main defender: without a knight on f6, White can bring a queen to g4
or h5 and attack h7 and f7 in combination with a knight on g5. The final point
is a rather strange one: the 'bad' bishop is often Black's most crucial
defender. The bishop on c8 was in fac the best defender of the e6 pawn. Its
protection of the e6 pawn allowed Black to consider moves such as ..f6
removing White's space gaining e5 pawn, or ..f5 to gain some space for Black
on the kingside and thus reduce White's advantage in this area. Black knew
that if White were to open up the e file by taking on f6, then the e6 pawn
would be well defended by the 'bad' but crucial bishop on c8! Moreover the
presence of such a bad bishop makes it strangely more difficult for White to
blast open the position as this will only increase the activity of the
light-squared bishop by opening diagonals for it. A similar case arises in the
King's Indian, this time with the dark-squared bishop: After 1. d4 Nf6 2. c4
g6 3.Nc3 Bg7 4. e4 d6 5. Nf3 0-0 6. Be2 e5 7 d5, Black has blocked in his
dark-squared bishop behind his e5 pawn. However this makes it less tempting
for White to try and storm the Black position by playing for f4-f5 as this
would allow Black to play ..e5xf4 and gloriously activate the bishop on g7
along the a1-h8 diagonal. Bishops, bad or otherwise are always useful since
they impart solidity to a pawn centre. Thus we see that White's 8th move is
absolutely correct, providing that he realises one crucial point: that the
present is his big chance to make the most of his advantages. The main line is
very logical: 9.0-0 b5 10 Nf4 Be7 11 Qg4 g6 12 h4! as in Kotrinias-Short Novi
Sad Olympiad 1990 and more recently Emms-Straeter Hastings open 1996 when 12..
Bxh4 13 Nxe6! fxe6 14 Qxe6+ would pick up the loose knight on a6. In this way
White manages to create a big initiative on the kingside.} 9. Nf3 $6 Be7 10.
Qa4 $6 {Rather chasing the knight to where it wants to go.} 10... Nc7 11. dxc5
$5 bxc5 12. Bf4 $6 {12 Qg4 transferring the queen to the kingside (via a
slightly complicated route!) was more incisive.} 12... O-O 13. O-O {Annotating
games played by computers needs a slightly different approach to annotating
games played by people. To be honest, White's opening play has been quite poor:
Black has been able to develop his kingside satisfactorily and White's pieces
dont make a particularly active or coordinated impression. I have seen d4xc5
played by strong players in such a position but then it is usually followed by
c3-c4 to put pressure on the Black centre. Here the plan is not possible, as
the misplaced queen on a4 leaves White vulnerable to the fork ...Nd7-b6 if he
plays c4. If this were a human game, I would conclude that White was quite a
weak player and would expect a rather weak continuation of the game
culminating in a tactical blunder around move 29. 0-1 no problem! But here of
course, this does not happen: from now on, White plays extremely sensibly and
patiently! This is one of the psychological problems that humans face when
playing computers. Experienced players tend to adjust their strength to the
strength of their opponents: from the opening they can gauge the ability of
their adversary and somehow concentrate accordingly. A human who played as
Hiarcs did here would get little respect and the expectation of a quick win
would be in the air! The concentration level drops a little, but this doesn't
matter because the human opponent will soon make a concrete tactical mistake
that will be refuted. But the computer never does! In general, from
strategical openings that they do not know, even the best computers will not
be able to gain an advantage and will probably concede even a slight
disadvantage, but nowadays that is ALL they will give away. From this
situation, they will play sensibly, forcing nothing and giving nothing away,
providing no concrete opportunities for finishing off the game easily that the
human player so desperately desires! This ability to wait, playing reasonable
moves in fractionally worse positions without giving a single conrete tactical
chance is in my opinion the most important reason why computers have become as
strong as they are. The rest of the game does make a strange impression: White
waits, tempts Black forward and then in Black's time trouble mops up the game
with some virtuoso knight play!} 13... a5 14. Rad1 Nb6 {Both these moves are
very reasonable but they seem a little carefree for my liking! Despite his
loss of time in the opening, White's kingside chances are still there. I would
prefer the conservative 13..Re8 14 Rad1 Nf8 securing the Black kingside and
only then looking for some queenside play. By leaving his kingside so free of
minor piece protection, Black is just asking White to get active over there!}
15. Qc2 a4 16. Ng3 Qb8 {Alertly spotting White's threat of 17 Ne4 heading
either for the d6 outpost or for g5, attacking h7. However, now the black
queen is forced away from defending the kingside, giving White even more
encouragement over there!} 17. Nh5 {
17 Bg5 is also interesting, not allowing Black's defence in the game.} 17...
Ra7 18. Bg5 $6 {18 Ng5 must be better and should give White a nice initiative.
if 18..Bxg5 then 19 Bxg5 Nc4 and a choice between 20 Nxg7, Be7 or Bf6} 18...
Ne8 $1 {Just in time! Black has made progress on the queenside, but the rest
of his position looks a little uncoordinated and his kingside is under a
little pressure. He still has his little edge but he has a bit of regrouping
to do.} 19. Nf4 Bxg5 20. Nxg5 g6 21. Rfe1 Ng7 22. Nd3 {Black has re-established
control over the position as 22 Qd3 or 22 Rd3 would be met simply by 22..h6
driving the knight from g5.} 22... Qc7 23. Qe2 a3 $6 {
As Eric points out in his magazine, 23..c4 first followed by ..a3 was better.}
24. b4 Na4 25. Rc1 {I'm not quite sure what is going on in this position, but
I do know that black definitely did not want to allow this type of situation.
I have lost many blitz games against Genius from this type of position. White
has too many possible threats: he can soften up the kingside by transferring
his queen to the h-file or he can break on the queenside with c4 using the
Black queen's vulnerable placing opposite the rook on c1. In increasing time
trouble, this is a horrible one to defend against a machine.} 25... Rc8 26. Qd2
$1 {A good defensive move and also a source of worry for Black since once the
White knight retreats from g5, White will always have the chance of attacking
with Qh6 and Ng5.} 26... Qd8 27. Nf3 cxb4 $2 {Definitely a mistake: unless
this leads to a concrete result, Black should not be opening the lines as this
just feeds the tactical posibilities open to White and gives Black another
front to protect. My feeling is that Black has not quite re-adjusted to the
position and was still just playing for the win even though the position has
become murky and is not in his favour any more.} 28. cxb4 Rc4 29. Nc5 $1 Nb6
30. Nd4 Rxc1 31. Qxc1 Qh4 $2 {Another very agressive but misplaced move that
shows that Black still has not realised that the position is against him. The
rest from Hiarcs is just a joy! A flurry of very precise and very unpleasant
knight moves.} 32. Ncb3 Nc4 33. g3 $1 Qg4 {
The queen cannot return to cover the back rank with 33..Qd8 as 34 Nc6 wins.}
34. f3 Qh5 35. Na5 Rc7 36. Nb5 $1 Rc8 37. Qc3 Nb6 38. Nc6 Kh8 39. Nd6 Rc7 40.
Qc5 Nd7 41. Qa5 Ne8 42. Qa8 Qxf3 43. Nd4 {A very fine finishing sequence that
any human player would have been very proud of! My feeling is that Hergott
lost this game by underestimating the machine: when things went slightly wrong
after his hasty 23rd move, he lashed out, opening lines (27..cxb4) and
abandoning his back rank (31..Qh4). By playing antipositionally, he did the
computer's job for him. It is something that every chess player has done when
suddenly confronted by a problem against a player that he expects to beat:
unwillingly to accept the fact that he is worse, the player launches an attack,
seeking to change the course of the game not by the objective strength of his
moves, but by their aggressiveness and the force of his desire for victory. In
this respect we have much to learn from computers in the way that they defend
some inferior positions, concentrating not so much on finding the best plan of
defence but looking instead to play reasonable positional moves that are
tactically watertight at a fairly brisk pace. What also struck me in this game
was that computers still have to develop a feel for CREATING an initiative.
For example, once Black had opened the lines for White with 27 ..cxb4, Hiarcs
played at a very high standard and it would have needed an immensely
impressive defence to save the game, especially with time becoming a little
short. However we can contrast this with the opening where White needed to
sense that White's initiative was going to happen on the kingside: twice,
Hiarcs passed up the opportunity to transfer his queen to this critical area.
The problem I would imagine in programming must stem from the strange way in
which computers regard tempi. To a human player, losing a tempo by moving a
piece back to where it came from is a traumatic event: it shows a lack of
foresight, a mistake in calculation. This feeling is what drives humans to
play dynamic chess, always seeking to go forwards and justify the moves that
went before. For computers none of this applies as each position is treated on
its own merits and not seen necesarily as part of a great whole. One of the
most frustrating feelings about playing a computer is that it will take a move
back without emotion or distress! Indeed the distress is all yours when you
realise that though the move is not frightening, it is not REFUTABLE. In
general a human player can only take so many shocks or frustrations before he
starts feeling vulnerable, and somehow the computer has a way of provoking a
few too many for comfort!} 1-0

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "2"]
[White "Hergott"]
[Black "Hiarcs"]
[Result "1/2-1/2"]
[ECO "E04"]
[PlyCount "150"]

1. Nf3 d5 2. g3 c5 3. Bg2 Nc6 4. d4 e6 5. O-O Nf6 6. c4 dxc4 {After a slight
detour, we are back into the Catalan opening, an opening with which Hergott
has a great deal of experience.} 7. Qa4 {
Ne5 has been much more popular recently but the text also contains some venom.}
7... Bd7 8. Qxc4 b5 9. Qd3 Rc8 10. dxc5 Bxc5 11. Nc3 b4 12. Ne4 Nxe4 13. Qxe4
O-O {Though this plan of development has been used by many strong players, it
does leave a very strange impression on me. The main problem I think is that
both Black's bishops are misplaced: the bishop on d7 is vulnerable on the open
d-file when it would rather be on the long diagonal on b7, while the dark
squared bishop is rather loose on c5 and would rather be on f6 attacking the
weak pawn on b2. Black's b pawn is also exposed on b4; though it gains
queenside space and ties down the White a pawn, it provides a nice target for
Whites's pieces. Moreover its absence from b7 means that the knight on c6 is
loose on the open c file and that Black has some light square weaknesses on
the queenside. All in all, if I were playing this position with Black for the
first time, I would always be worried that I might drop a piece somewhere
since none of my minor pieces have found secure posts! Instead of Black's last
move, 13..Ne7 14 Ne5 Bb5 has been more customary to free the light-squared
bishop and at least activate it a little, but I don't believe that Black has
full equality even here.} 14. Rd1 Qe7 15. Bg5 f6 16. Bh3 $1 {A nice move,
preventing 16..fxg5 due to 17 Rxd7 Qxd7 18 Bxe6+. More importantly however,
White stops Black from trying to activate his light-squared bishop in the
future either by ..e6-e5 and ..Bd7-e6 or by ..Bd7-e8-f7 or g6.} 16... Nb8 $5 {
A very interesting and ambitious idea, dealing with the threat of 17 Rxd7 by
17..Nxd7 18 Qxe6+ Rf7! and freeing the bishop on d7 to activate itself
somewhere on the e8-a4 diagonal. The problem of course is that the knight is
very passive on b8, unprotecting the b pawn and losing influence over the
central dark squares. 16.. Kh8 17 Bc3 is also very unpleasant however: as a
Black player in such situations, you always have a feeling of helplessness. It
is very easy to see the moves that White is going to play to improve his
position - a rook on the c file, maybe double on that file - but as Black, it
is hard to come up with anything that does not worsen your position.} 17. Be3
f5 {Another unwanted but necessary move to defend e6.} 18. Bxc5 Qxc5 19. Qf4
Qe7 20. Rac1 Ba4 $6 {So far I think that Hiarcs had defended excellently: it
has conceded a few squares to White but that is not its fault, it's the
position's fault! In every other respect, Hiarcs has done a good job: it has
kept a reasonable hold over the open files and it has allowed its bishop to
get active. However, here it makes a crucial mistake: by forcing the White b
pawn to b3, Black does force White to weaken his dark squares on the queenside,
but this is not really of any use to Black since his pieces are too inactive
to take any advantage from it: he would need his knight to be on d5 for it to
matter. However, this helps White by protecting the a pawn against any attack
along the g8-a2 diagonal: in fact it gives White a useful extra tempo. I think
that the immediate 20..Bc6 21 Ne5 Bd5 (attacking a2) was stronger as 22 Rxc8
Rxc8 23 Bxf5 Rf8! is pleasant for Black, while if 22 b3 then Black has at
least an extra tempo to do something useful. Now the position becomes REALLY
grim!} 21. b3 Bc6 22. Ne5 Be4 23. Bg2 $1 {An excellent move exchanging off
Black's only active piece for the only Whbite piece that has become inactive!}
23... g5 {
Probably not a bad move! It at least gains some space on the kingside.} 24. Qe3
Bxg2 25. Kxg2 Rfe8 26. Rd4 h6 27. Rdc4 $1 Rxc4 28. Rxc4 Kh7 29. h4 $1 g4 30.
Qc1 Na6 31. Rc6 {Good of course but I quite like the simple 31 Kg1 just
avoiding any tricks on the a1-h8 diagonal. White's position is so good that he
doesn't need to - in fact he MUSTN'T rush to capitalise on it.} 31... Qg7 $5 {
Very ingenious and an annoying move to meet in time trouble.} 32. Nxg4 $2 {
If this works tactically then it is a fine spot to round off the game quickly.
If it doesn't work however, then it is simply a bad positional move as White
trades off his excellent knight on e5 for Black's sad specimen on a6. There
may be more trenchant ways to play but I would like to play simply 32 Nd3
attacking the b4 pawn and if 32..Qb7 then 33 Kg1, consolidating just for a
moment before the final invasion. There is no question here that White is
winning.} 32... Qxg4 33. Rxa6 Qe4+ 34. Kg1 Kg6 35. Qc7 $6 {This is all very
unfortunate for Hergott since all his mistakes stem from missing just one move
from Hiarcs. If he realised this then I am sure he would have played 35 Qd2
here, intending consolidation first with f2-f3 and Kg1-f2 before proceeding.
Incidentally 34..Qxe2 was better than 34..Kg6, not allowing this chance.} 35...
Qxe2 36. Rxa7 Kh5 $3 {Wonderful! 37 Qf7+ Kg4 Qxe8 Kh3!! wins!} 37. Kg2 $5 Qe4+
38. Kh2 Rg8 39. Qd6 $2 {This must be the final mistake. Eric's suggestion of
39 Qb7 regaining control of the h1-a8 diagional is absolutely right! It would
be a lot more difficult than before - maybe White is not actually winning any
more- but White would still have all the chances.} 39... Qe2 40. Kg2 Qe4+ 41.
Kh2 Qe2 42. Qd4 Rg4 43. Qe3 {White should now take the draw that was on offer
after 43 Qc5 Rxg3!! 44 Kxg3 Qg4+ and perpetual. After 43 Qe3 he has to suffer
for a little while, before getting the draw.} 43... Qxe3 44. fxe3 Re4 45. Rb7
Rxe3 46. Rxb4 Re2+ 47. Kh3 Rxa2 48. Rb5 Re2 49. g4+ Kg6 50. gxf5+ exf5 51. b4
Re3+ 52. Kg2 Rb3 53. Kh2 Kh5 54. Rxf5+ Kg4 55. Rf2 h5 56. Rg2+ Kxh4 57. Rg8
Rxb4 58. Rg2 Ra4 59. Rg8 Rc4 60. Rg7 Rc3 61. Rg8 Rd3 62. Rg7 Rf3 63. Rg8 Rc3
64. Rg7 Rh3+ 65. Kg2 Ra3 66. Kh2 Rb3 67. Rg8 Rb5 68. Rg7 Ra5 69. Rg8 Re5 70.
Rg7 Ra5 71. Rg8 Re5 72. Rg7 Re3 73. Rg8 Rf3 74. Rg7 Rb3 75. Rg8 Rf3 {A very int
eresting game. Hergott played most of it beautifully in fact: the opening
choice was excellent and the middlegame play was simple and very effective.
Hiarcs found an excellent defensive plan on move 16, but a silly 20th move
cost Hiaarcs a vital tempo for the defence. Why didn't Hergott win? Simply
because of the refusal of the computer to offer him a concrete tactical win.
There were plenty of positions on offer that were probably or definitely
winning in the LONG RUN, but never anything immediate. Finishing off better
positions is one of the most difficult skills for all chessplayers- you always
really appreciate it when someone makes an early mistake that takes away the
need for a long period of care and accuracy to nurse an advantage home. But
for game after game, the computer will offer you nothing clear and always ask
you to make 70 moves to win a won game rather than the 30 or 40 moves it would
take against most human players. The moral is clear: to beat a computer, any
chessplayer really needs excellent technique and vast amounts of patience when
in a clearly better position, because the computer will never give you the
game on a plate- you always have to go out and win it by your own efforts!}
1/2-1/2

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "3"]
[White "Hiarcs"]
[Black "Hergott"]
[Result "1-0"]
[ECO "A56"]
[PlyCount "97"]

1. c4 Nf6 2. d4 c5 3. d5 e5 {A slight surprise from Hergott as although he
plays a wide variety of offbeat systems against d4 lines, I have no previous
games of his with the Czech Benoni. It is generally regarded as a bit passive,
but it does instantly provide the blocked positions to which computers are
less suited.} 4. Nc3 d6 5. e4 Be7 6. Nf3 Nbd7 7. Be2 Nf8 8. O-O Ng6 {I'm a litt
le surprised that this scheme of development was programmed into Hiarcs' book.
White has 2 general methods of combatting the Czech Benoni: either, White
fianchettoes his bishop with g3 and Bg2, and castles and then transfers his
knight to d3 to support f2-f4, or he exploits the fact that the centre is
closed and plays to gain vast amounts of space on the kingside with h2-h3 and
g2-g4. 7 Be2 and 8 0-0 is a bit of a 'halfway' plan: it can do no harm to
develop in this way, but it does not make the most of White's possibilities.}
9. Be3 h6 $6 {This is a little careless as it breaks the cardinal rule of
playing computers in blocked positions: don't make the correct positional plan
needlessly tempting by giving it a tactical basis as well! 9..0-0 was simple
and best.} 10. b4 $1 b6 11. bxc5 bxc5 {
By his mistake on move 9, Black has given Hiarcs the open b file for 'free'.}
12. Rb1 O-O 13. Qa4 Nh7 14. Qc2 $2 {A needlessly passive move. Though Black's
proposed kingside expansion with ..Ng5 and ..Nf4 may seem a little threatening,
diverting all these pieces to the kingside will inevitably give White extra
chances against Black's weakened queenside. I would prefer a calm
consolidating move such as 14 Kh1?! when 14..Ng5 15 Nxg5 Bxg5 16 Bxg5 Qxg5 17
Qc6 is very good for White or a move such as 14 Rb2 threatening 15 Rfb1 and 16
Rb8 invading on the open b-file.} 14... Ng5 15. Nd2 $6 Nf4 16. Bxf4 exf4 {
In just a few moves, Black's position has gone from quite unpleasant to
better! White's last move opened the h8-a1 diagonal for Black's dark squared
bishop which is usually a problem piece in this line, so Black has achieved
quite a lot so far.} 17. Rfc1 $5 {Not a particularly purposeful move, but a
very solid defensive move, protecting White's loose knight on c3.} 17... Bf6
18. Bd3 $5 {A very cheeky idea, threatening 19 h4 Nh7 20 e5 winning a piece!}
18... Be5 $6 {This is needlessly commital. Simply 18..g6 would have dealt with
White's threat, leaving White with a rather planless position.} 19. h4 $1 {
Sharp and best: this is White's chance to confuse the situation.} 19... Nh7 20.
Nf3 Qe7 21. Nxe5 Qxe5 $2 {21..dxe5 was a must here when, to be honest, I quite
like Black's practical chances - though I don't think that he is actually
better any more. But after 22 g3 g5!?, 22 Na4 Qxh4 23 Nxc5 (f3 is best) f3; or
22 f3 Nf6! (22.. Qxh4 23 Qf2! is annoying for Black) I quite fancy Black's
chances of making some chaos on the kingside. In general from such positions,
I have quite a good record against computers as they tend to react a trifle
slowly to the gathering clouds on the kingside, concentrating just a little
too much on their queenside achievements. The point of such positions is that
Black's weaknesses on the queenside are permanent and will remain whatever he
does. Consequently, the only way that Black can 'defend' these weaknesses is
to make sure that they don't matter by mating White on the kingside!
Consequently the best way to handle such positions as White is to concentrate
very hard on securing your kingside position, confident that once you have
weathered the storm, Black's queenside problems will still be waiting for you!}
22. Nb5 $1 {A fine tactical point that Hergott had no doubt missed.} 22... Bd7
23. Nxd6 $1 Qxd6 24. e5 Qxe5 25. Bxh7+ Kh8 26. Be4 {
And all of a sudden, Black is worse and faces an unenviable task!} 26... Rab8
27. Re1 Qf6 28. Qd3 $1 {A very nerveless and strong idea. After 28..Qxh4 29
Rxb8 Rxb8 30 Qa3! Re8 (30..f5 31 Bf3! (31 Qxa7 Rb2!) is good for White) 31
Qxa7 Qe7 32 Re2! to unpin with Bf3, White stands very well.} 28... Qa6 29. Qc3
$1 {A very good move, taking control of more central dark squares while
preventing 29..Qxa2 due to 30 Ra1!} 29... Rfe8 30. Rxb8 Rxb8 31. Bd3 Qd6 32.
Qe5 {This is a touch too direct for my liking. I would really like to soften
up the Black position a little more before heading for the ending. I quite
like 32 Bc2 Re8 33 Rb1 Bc8 34 Qd3 f5 35 Qf3, seeking to 'tickle' Black a bit
to see if he can be forced into a mistake.} 32... Qxe5 33. Rxe5 Re8 $4 {
A losing blunder. 33..Kg8 is worse for Black, but winning it would not be too
easy for White by any means since there is no immediate way to break through.
The text however just loses a piece, and all hope of saving the game. We see
here an example of what I said earlier: the human player, discouraged by the
turn of events, makes a blunder and hands the game over on a plate. Hiarcs
would not have donethat!} 34. Rxe8+ Bxe8 35. Bf5 g6 36. Bc8 Kg7 37. d6 Kf6 38.
d7 Bxd7 39. Bxd7 Ke5 40. Be8 Kd4 41. Bxf7 g5 42. hxg5 hxg5 43. Kh2 a5 44. Kh3
a4 45. a3 $1 {No panic! [Tryfon notes that Matthew is also a Hitchhikers guide
to the galaxy fan!]} 45... Kc3 46. Kg4 $1 Kb3 47. Be8 Kxa3 48. Kxg5 Kb3 49.
Bxa4+ 1-0

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "4"]
[White "Hergott"]
[Black "Hiarcs"]
[Result "1/2-1/2"]
[ECO "A11"]
[PlyCount "72"]

1. Nf3 d5 2. g3 c6 3. Bg2 Nf6 4. O-O Bg4 {A very solid and well respected way
of playing against White's flank opening. Black activates his light squared
bishop outside the pawn chain before setting up a solid light-squared centre
on c6, d5 and e6 to reduce the activity of White's bishop on g2.} 5. c4 e6 6.
b3 Bd6 7. Bb2 O-O 8. Ne5 Bf5 9. d4 $6 {Hergott has played this setup in a
number of games, but experience from both sides of the board has convinced me
that this is the wrong approach for White and that the standard plans with
d2-d3 are his only chance for a small edge. Essentially, the problem with this
line is that White's main chance to gain territory is to play e2-e4 at some
stage, but not only does this take a lot of organising, but also it will
necessarily involve the exchange of a few minor pieces which substantially
hurts White's prospects of gaining anything from the opening.} 9... Nbd7 10.
Nd2 Rc8 11. Re1 Be7 $1 {A very sensible move, looking to play ..Nd7xe5
removing White's advanced knight at some stage.} 12. a3 b6 {Slightly strange,
weakening the queenside for no real reason, but it does not ruin anything.} 13.
Rc1 Nxe5 14. dxe5 Ne4 15. cxd5 cxd5 16. Rxc8 Qxc8 {
The position is absolutely equal.} 17. Nf3 Qb8 18. Qa1 Bg6 19. Rc1 Bh5 20. Nd4
Bg5 21. f4 Be7 22. Bxe4 dxe4 23. Nc6 Qb7 24. Nxe7+ Qxe7 25. Rc2 Qd7 26. Bd4 Rd8
27. Kf2 f5 28. Be3 Qe7 29. Qc1 Qe8 30. Rc7 {White has made maybe a little
progress here, but the text loses control by allowing Black an immediate
tactical solution to the position. 30 a4 (stopping ..Qe8-b5) would give White
a tiny edge since his pieces are slightly better placed: he controls the open
c file and his bishop is well placed on e3, eyeing the Black queenside pawns.}
30... Bxe2 $1 31. Kxe2 Qh5+ 32. Ke1 Qxh2 33. Rc8 Qh1+ 34. Kf2 Qf3+ 35. Ke1 Qh1+
36. Kf2 Qf3+ 1/2-1/2

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "5"]
[White "Hiarcs"]
[Black "Hergott"]
[Result "0-1"]
[ECO "C11"]
[PlyCount "141"]

1. e4 e6 2. d4 d5 3. Nc3 Nf6 4. e5 Nfd7 5. f4 c5 6. Nf3 Nc6 7. Be3 Qb6 8. Na4
Qa5+ 9. c3 c4 $5 10. b4 Qc7 {As a White player, you always feel that it is a
trifle unsporting when someone plays such a variation against you: you want
open lines, a rapid attack against the Black king, glory or bust.. and instead
you have to face this war of attrition! I don't have any examples of Hergott
playing this position before though 3..Nf6 seems to be a favourite way of his
to meet 3 Nc3 in the French, and he has played 7..Qb6. In fact, it has been
noticeable throughout the match that Hergott generally played his favourite
openings and didn't just play anything just to try and confuse Hiarcs. This
actually makes Hiarcs' success all the more impressive since it beat Hergott
from positions in which Hergott has had virtually a lifge-time's experience!
This particular variation is a favourite of the Russian grandmaster Evgeny
Bareev who is the most consistent exponent of the French at the highest level.
By blocking the centre and the queenside, Black removes most of the chances of
creating play there which leaves only the kingside which Black intends to seal
with ..f7-f5.} 11. Be2 Be7 12. O-O {Against Psakhis in Moscow 1994, Short
played the interesting 12 a3!? which can best be described as a constructive
waiting move, shoring up the queenside while waiting for Black to commit
himself. After 12..f5, he played 13 Rg1 and achieved a rapid g2-g4 to open the
kingside a little, without having committed his king to the kingside.} 12... f5
13. exf6 $5 {A new move amazingly! It does seem like a reasonable try, opening
up the position a little. 13 g4!? has been the most violent alternative. For
example in Franzoni-Kortchnoi Swiss Championship 1994, the game carried on 13..
fxg4 14 Ng5 Nf8 15 Bxg4 h6 16 Nh3 g6 17 b5 Nd8 18 Bf3 b6 19 Nb2 which looks
ugly for both sides to be honest, but shortly ended in a draw. The big problem
with 13 exf6 is that it does give Black a path to activate his light-squared
bishop via d7-e8 to g6 and then to f5 or e4} 13... Bxf6 14. Ng5 Nf8 15. Qc2 Ne7
$1 {Good patient manoeuvering: the knight eyes the f5 square and covers the
kingside light squares as well.} 16. Bg4 b6 17. Rae1 h6 18. Nf3 {18 Bh3 seems
more logical, covering the h2 pawn and thus allowing White to meet 18..hxg5
with 19 fxg5 regaining the piece. In this way, White keeps the pressure on the
e6 pawn and makes it harder for Black to develop: for example, after ..Bd7, ..
0-0-0 cannot be considered immediately due to Ng5-f7 forking the rooks on h8
and d8. Since White has chosen such an active plan with 13 exf6, he must keep
the pressure on Black to have a chance of an advantage.} 18... Bd7 19. a3 {
Around here, I would really like White to be regrouping his offside knight on
a4 to somewhere useful such as g4 via b2-d1-f2. It may seem like a rather slow
trip but the position is so static that nothing is going to happen in a hurry.}
19... O-O-O 20. Ne5 Bxe5 21. fxe5 {Again, I'm not too happy with this decision.
White has gained the 2 bishops but by blocking the half-open e file, White
makes Black's biggest weakneess- the pawn on e6- considerably safer.} 21... Be8
$1 22. Nb2 Bg6 23. Qa4 Nf5 24. Bf4 Qd7 25. b5 $2 a5 $2 26. bxa6 $6 Qxa4 27.
Nxa4 Kc7 {A sudden change in the position and one that is definitely against
Hiarcs. The point is that before move 25, though Hiarcs had not played the
very best positional moves, there was only one filee for invasion- the f file-
which by itself could never prove enough for Black to drum up anything really
meaningful. However, now Black has the half-open a file and a target on a3;
coupled with the bishop on g6 which eyes the b1 square and thus dissuades
White from contesting the b file should it ever be opened, this gives White an
unenviable task.} 28. Nb2 Ra8 $1 29. a4 Rxa6 30. Rf3 {A strange move but White
has a rather nasty dilemna here: if he tries to go passive, then Black will
calmly improve his position, bringing his knight on b3 via d7-b8-c6-a5. My
favourite idea would be to try and distract Black a bit by getting some
activity on the f file also, but to do this I feel that White has to exchange
off the light-squared bishops. Perhaps 30 Bd1 Nd7 31 Bc2 to play Bf4-c1 and a
quick g2-g4 is the best, but it still is not a great thrill to defend.} 30...
Nd7 31. g3 $6 Rb8 32. Ref1 b5 $1 {A nice idea, removing White's weakness but
opening up 2 files for Black's rooks.} 33. axb5 Rxb5 34. R3f2 Rb3 35. Nd1 Kd8
36. Be2 Ne7 37. g4 Nc6 38. Bg3 {I would prefer 38 h4 looking for h4-h5 and
g4-g5 to create some more space on the kingside.} 38... Na7 39. Bh4+ Kc7 40.
Rf4 Nb5 41. Be1 Ra1 42. R4f2 Kc6 43. Bd2 Nb6 {
Black turns the screw. This knight is heading for a4 to attack the c3 pawn.}
44. g5 h5 45. Be1 Na4 46. Rf8 Naxc3 47. Bxc3 Nxc3 48. Nxc3 Rxf1+ 49. Bxf1 Rxc3
{Black is just winning.} 50. Rc8+ Kb6 51. Rb8+ Ka5 52. Ra8+ Kb4 53. Rb8+ Ka3
54. Ra8+ Kb2 55. Ra6 Bf5 56. Be2 g6 57. Kf2 Rh3 58. Kg2 Rh4 59. Bf3 Rxd4 60.
Kf2 c3 61. Rb6+ Kc1 62. Ke1 Rd2 63. h4 Rb2 64. Rc6 c2 65. Bd1 Kb1 66. Bxc2+
Rxc2 67. Rd6 Rc4 68. Kd2 Rxh4 69. Rb6+ Ka2 70. Ra6+ Kb3 71. Rc6 {A very nicely
controlled game from Hergott. The difference from the previous game was the
patience he showed and the time he gave himself to play the game: first of all
in the opening, he quietly regrouped his pieces, covering his weak points, and
made sure that he did not give Hiarcs anything concrete to grab hold of. Then
in the winning process, unlike game 2 for example, he concentrated all the
time on playing good safe positional moves: even if they don't win the
quickest, they always win the surest! But despite all that fine play, it still
takes 70 moves to put Hiarcs away! 3-2 down with just the last game to play,
Hergott knows that he will need a similarly big effort just to tie a match I'm
sure he expected to win quite easily. This is never an easy psychological
situation: adjusting your sights down-ward, swallowing your pride and fighting
like a demon for something that you would have never accepted in a million
years if it had been offered to you before the match! As a small digression, I
do wonder whether something of this kind happened to Kasparov when playing
Deeper Blue: all the work, the preparation and the pride he had put at stake-
2.5-2.5 and fighting to save the match by drawing as Black just wasn't what he
had envisaged. It wasn't the triumphant situation that he felt all his efforts
deserved, and so somehow he couldn't bring himself to give the game his full
undivided attention. That at any rate is the only way that I can explain
Gazza's mixing up of moves (7..h6 instead of Bd6) in the fateful 6th game.} 0-1

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "6"]
[White "Hergott"]
[Black "Hiarcs"]
[Result "0-1"]
[ECO "D23"]
[PlyCount "116"]

1. d4 Nf6 2. Nf3 d5 3. c4 dxc4 4. Qa4+ $5 {A very solid opening from Hergott.
The usual choice here is the very sharp 4 Nc3 when Black can try to hang on to
the pawn with 4..a6 leading to crazy complications. Though Hergott had played
4 Nc3 as White before, this would not fit in with his match strategy as White
in which he tried to play solidly and carefully, with reasonable success.} 4...
Nc6 $1 {The most active and best move.} 5. Qxc4 Be6 $5 {5..Nd7!? intending ..
Nd7-b6 and then ..g7-g6 and ..Bf8-g7 to put pressure against the d4 pawn in a
sort of Gruenfeld formation is also quite reasonable.} 6. Qa4 Bd5 $5 {
This leads to a sort of Chigorin-type position which is one of those aquired
tastes! You give up both bishops and the centre, but in return you give White
some pawn weaknesses. Tony Miles loves this sort of thing but personally, I'd
still go for 6..g6!} 7. e3 e6 8. Nc3 Bb4 $5 9. Bd2 Bxf3 10. gxf3 a6 11. a3 Bxc3
12. bxc3 {I imagine that Hergott must have been feeling quite confident at
this point: the reason is not so much to do with the position as with the
nature of the stuggle. After all, he played a quiet, innocuous continuation
and within 12 moves, Hiarcs has given him both bishops and a fairly imposing
pawn centre! This sort of thing makes you feel that you made the right
decision: why take risks yourself to unbalance the position if your opponent
is going to do all that for you? As for an objective assessment of the
position, I would say that it is easier for White to play than Black, but the
latter does have definite chances. The undisputed expert in this type of
position is the young Russian Alexander Morozevitch who handles the 2 knights
with all the skill of his hero Chigorin. Black's main idea is to 'dilute' the
White centre by striking at d4 first with ..e5 and then after moving the Black
knight from c6, by ..c5. In order to counter this White often moves his e pawn
forward so that after ..e5, White can play d5 and establish an advanced pawn
chain. However Black then says, "Aha, I've tempted you forwards: this leaves
weaknesses in your position". For example, when the e pawn moves to e4, then
f4 is weakened. Therefore after ..e5 d5, I will play my knight from c6 to e7
to g6 and eye this weakness!". I have studied many games of this type but even
now I'm not really sure what the truth is. The reality is that such systems
as the Chigorin are particularly effective as 'psychological' openings since
they create positions that are entirely new and that require different
stratagems to normal openings. In such a case, the Black player will gladly
take on a position that he knows is a little worse for him, knowing that this
is compensated by the extra imagination that White will require to deal with
these new problems.} 12... Qd5 {12.. 0-0 had to be better, aiming for ..e6-e5.}
13. Rg1 $1 {And not e4? Nxe4! 14 fxe4 Qxe4+ picking up the loose rook on h1.}
13... O-O {13..Qxf3 14 Bg2!} 14. e4 Qd6 {!?!? Utterly outrageous! This is one
move that I could confidently say, no grandmaster in 100 would consider! I
don't actually think that is a very good move but it is typical of the moves
that we will be forced to anticipate as computers become stronger and we play
more games against them. Black gives up a piece? And White has the 2 bishops
on top?? Black does get three pawns I know, but in the ending, 2 bishops can
make light work of 3 pawns in just a few moves. The beauty of this move
however is the psychological effect that it has since as with so many computer
moves, it affects what for human players would be the natural course of the
game. I'm sure that Hergott was looking forward to a kingside attack, using
his 2 bishops and the extra space that his centre gives him after a normal
move such as 14..Qh5. Now he is forced to play this position with an unusual
material balance and where his centre has disappeared- a position that is sure
must be completely winning and yet still requires a great deal of care,
especially against this particular opponent. It doesn't matter that this
position is better than if Black had played 14.. Qh5; the point is that it
does not fall within the range of normal experience and so it cannot be played
on auto-pilot: it requires good thought and great care and an awareness of
unusual factors. And for a crucial game like this where part of your energy is
devoted to keeping hold of your nerves, any extra effort likke this is not
appreciated!} 15. e5 Nxe5 16. dxe5 Qxe5+ 17. Be2 Qxh2 18. O-O-O Rfd8 19. Qf4 $1
Qxf4 20. Bxf4 Rxd1+ 21. Rxd1 Nd5 22. Be5 $2 {A bad sign. Converting an
advantage needs patience and accurate short-term calculation. This just tries
to force matters when there is no need to do so. Simply 22 Bd2, holding the
fort temporarily, aiming to force the knight back with c3-c4 before reemerging
with Bd2-f4 was better. Black will not be able to hold out against a rook and
2 bishops though it may take some time! The text shows that Hergott knows that
he is winning and just wants to force it through and get it over with.
Unfortunately, this is when computers like Hiarcs suddenly jump to 2700
strength!} 22... f6 23. c4 Re8 $1 {A very nice move and undoubtedly a nasty
shock for Hergott. Remember, he HAS to win this game!} 24. cxd5 $6 {Retreating
is always hard, but I think I would have preferred 24 Bb2 Nf4 25 Bf1. The 2
bishops aren't so wonderful any more, but give them a glimmer of chance and
they will be at your throat again! The text increases the chance of a draw by
allowing not only White's bishop pair to be destroyed, but by swapping off a
pair of pawns as well, bringing black closer to his dream of R vs R and B!}
24... exd5 25. Rxd5 Rxe5 26. Rd8+ Kf7 27. Bd3 g6 {White has lost all his
advantage since Black's position is very harmonious and Whites own pawns are
rather sad specimins. Even more unfortunately we are now in the type of ending
that strong computers play absolutely superbly: where they have pawns versus
a piece.} 28. f4 Re8 29. Rd4 Re7 30. Kd2 h5 31. Rc4 f5 32. Rc1 Kf6 33. a4 a5 $1
{The rest is rather sad for Hergott, and I dont really want to comment on it.
I have also known that feeling, usually after a massive time scramble, of
seeing the ruins of a promising position and knowing that the point should
have been mine... if only at the critical moment!} 34. Rb1 c6 35. Rc1 Rh7 36.
Bf1 h4 37. Rc3 Rd7+ 38. Ke3 Rd1 39. Bg2 Rb1 40. Rd3 Ke7 41. Rd2 Re1+ 42. Kf3 b5
43. axb5 cxb5 44. Rd5 Rb1 45. Ke3 a4 46. Re5+ Kf8 47. Bc6 h3 48. Bxb5 a3 49.
Re8+ Kg7 50. Ra8 Rxb5 51. Rxa3 Rb1 52. Ra7+ Kh6 53. Ra8 Kh5 54. Rh8+ Kg4 55.
Rh6 Rb3+ 56. Ke2 Kxf4 57. Rxg6 Rb2+ 58. Kf1 Rb1+ 0-1

[Steve B]

yes an historical moment in chess history for sure
Hallsworth published an entire special issue devoted to this match alone

the issue is now somewhat of a collectors item

History In The Making Regards
Steve

[Debaser]

In this page are the dates of the games, from 7-19 april 1997, search for hiarcs as the page is long

http://www.geocities.jp/chess/tour1997.htm


Thanks for the commentaries, but who was the annotator?

[Harvey Williamson]

Thanks for the commentaries, but who was the annotator?

GM Matthew Sadler

[JEB_Tucson]

Outstanding! It's like reading a good book.

[Andrew]

Does anyone still have the Arno Nickel game? This was one of several a few years back where the computers were given the names of planets during
the game to hide their identity. I seem to remember Hiarcs won.
Also Arno posted a heap of analysis somewhere...

Andrew

[thierrycatalan]

Festival international des jeux à Cannes ( France) , february 1987,

the computer HITECH wins a match 2.0 vs Joel Lautier ,

World's Champion - 14 ans (elo 2250)

[mclane]

ok, no IM but nearly... (tournament time control) 1984

Weiß: Mephisto Exclusive S (Level 7)
Schwarz: Frank Grzesik (Elo 2285)

1.e4 Sf6 2.Sc3 d5 3.exd5 Sxd5 4.Lc4 Sb6 5.Lb3* Sc6 6.Df3 e6 7.Sge2 Le7 8.d3 0-0 9.0-0 Sa5 10.Lf4 Sxb3 11.axb3 Sd5 12.Lg3 a6 13.Sxd5 exd5 14.c4 dxc4 15.dxc4 c6 16.Tfd1 Db6 17.Ld6 Lxd6 18.Txd6 Te8 19.Sg3 Dc5 20.Tadl Dg5 21.T6d4 Le6 22.Se4 De7 23.h3 h6 24.Sd6 Ted8 25.Dd3 Td7 26.c5 Tad8 27.b4 Dg5 28.De3 Dxe3 29.fxe3 Tb8 30.T1d3 Te7 31.Td2 Kf8 32.e4 g6 33.Tf2 Kg7 34.Tf1 Td8 35.b5 axb5 36.Sf5 + gxf5 37.Txd8 fxe4 38.Td6 f5 39.b4 h5 40.Tf4 Kf6 41 .Kf2 Ke5 42.Th4 Lf7 43.Ke3.Le8 44.Td8 Lg6 45.Tf4 Tg7 46.Tf1 Lf7 47.Td7 h4 48.Kf2 f4 49.Kel e3 50.Te7+ Kf5 51.Txb7 Th7 52.Td7 Lg8 53.Td4 Ld5 54.Tfxf4+ Ke5 55.Ke2 Tg7 56.Tg4 Ta7 57.Kxe3 Ta2 58.Td2 Ta3+ 59.Kf2 Tb3 60.Txh4 Lc4 61.g3 Txb4 62.Th5+ Ke4 63.Th6 Tbl 64.Txc6 Tf1 + 65.Kg2 Ke3 66.Tb2 Tel 67.Td6 Ld3 68.g4 aufgegeben, da Mephisto auf den kleinen Schwindel Le4+ nebst Th1 matt nicht hereingefallen ist.


Grand-Prix" Nr. l war der rumänische Großmeister Florin Gheorghiu, der 5,5 Punkten aus sechs Partien erspielte.
Der mitspielende Mephisto Exclusive S, ein Seriengerät mit 68000 Prozessor, 12 MHz und dem neuesten Nitsche-Henne-Programm, erzielte drei Punkte und damit 50%, ein Ergebnis, das jedem Schachcomputer zur Ehre gereichen würde. Gekrönt wurde der Erfolg jedoch durch einen sensationellen Sieg über den Berliner Stadtmeister Frank Grzesik, den wohl stärksten Spieler, der je unter Turnierbedingungen gegen einen Mikrorechner eine Partie verloren hat.

it was the first Mephisto GrandPrix. winner of it was romanian GM Florin Gheorghiu with 5.5 from 6.
Mephisto III S 68000 (normal dedicated unit with 12 mhz) made
50% or 3 points.
Frank Grzesik was at this time best player in Berlin.