Valvo - Bono, game 1 summary (long)

[John Richards]

Here is a summary of game 1 of the Valvo-Bono match. All comments from
the Bono team are prefixed by [B]. All comments from the Valvo team
have [V] in front. All diagrams have been stripped out.

I have done a sub-editing job. I have removed bits and done some editing to
clarify or correct typos. I hope the two teams are not unhappy
with the results.

Thanks to both teams for sharing this game with us.

Game 1
Valvo Team - Bono Team

1. e4 c5
2. Nf3 d6

[B] The team didn't have any strong feelings about a defense. I spent
many years playing the Philidor's Defense and I'm probably one of the
few people still alive who likes that opening. I think the team as a
whole has more knowledge about the Sicilian than any other defense to
1. e4, though.

3. d4 cd
4. Nd4

[V] This game has yet to take shape. Black will probably play 4...Nf6 and we
will reply 5 Nc3. Black then can choose between 5...a6, 5...e6, and 5...Nc6.

4. ... Nf6

[B] We still haven't decided what variation of the Sicilian to play. Bill
Kelleher is very knowledgable about the Najdorf (5. Nc3 a6). I know
quite a bit about the Dragon (5. Nc3 g6). Either of these can lead to
a very wild and tactical game. This probably isn't ideal when one
member of the opposing team is a calculating machine, but as a group we
decided that the best approach to this match would be to try to make
the games fun first and then try to win later!

5 Nc3

[V] I think it is unlikely that we will see 5...Nc6 now as they could have
tried the 2...Nc6 move order to get there. Intuitively, I expect a
Naijdorf. A European player would play 5...e6, but a European player
would have used the 2...e6 move order. Maybe they are aware I often
play the Dragon and intend to use it against me.

They must not have much experience against strong computers or I would
think they would try something more solid and positional. From my
experience with DT, I KNOW they will always be wondering if they have
missed something and that will be to my psychological advantage.

5. ... a6

[B] The Najdorf Sicilian. As I mentioned in my previous note, this is one
of Bill Kelleher's favorite openings. I haven't played it in about 12
years but Adrian Casillas plays it from time to time. To the
uninitiated the move ... a6 looks a little premature. Intuitively, it
would seem that black should be completing his king side development.
Instead, black launches into his queen side counterplay at the first
possible moment.

The Najdorf is one of the most analyzed openings of the last thirty
years. At various times it's seemed like the whole opening has been
refuted, but new resources for black keep getting found. I'm not going
to even try to describe the possibilities here, but white's main
alternatives are 6. Be2, 6. Be3, 6. f4, 6. a4, 6. Bg5, and 6. Bc4.

6. Bc4 e6

[B] 6. Bc4 was a small surprise for us. We expected 6. Bg5, which is
another very aggressive and tactical alternative. I personally don't
know too much about the 6 Bc4 variations, but I know they've been
popular recently in the Informants. 6. Bc4 against the Najdorf can
transpose into the Sozin Attack (1. e4 c5 2. Nf3 Nc6 3. d4 cd4 4. Nxd4
Nf6 5. Nc3 d6 6. Bc4) if black plays an early ... Nc6. Black can try
for more if he wants by playing ... b5 and ... Nd7-c5. We'll spend
some more time looking at recent games before deciding which variation
to try.

The move we played, 6. ... e6, is the normal reply to 6. Bc4. Another
possibility was 6. ... Nbd7, but we rejected that as being unsound
looking and too commital. White usually plays 7. Bb3 after ... e6.

7. Bb3

[V] All standard so far. Najdorf once played 6...b5!? against Fischer, but for
the past 20 years the text moves have been automatic. 95% of the time, Black
now plays 7...b5 and play continues 8 OO. In Fischer's day 7...Nc6 was played
once in awhile and white can choose between Fischer's king side castling
and 8 Be3!? attempting to transpose into the Velmirovic Attack.

7. ... b5

[B] We spent a fair amount of time trying to decide between this move and
either 7 ... Be7 or 7 ... Bd7 with the idea of transposing into a normal
Sozin Attack. After looking at some games where black played 7 ... b5,
we decided that black's position seemed to be OK. In the latest
Informant there were a number of games with the variation 7 ... b5 8 O-O
Be7 9 Qf3. White won most of them but the opening didn't seem to be the
reason why. Ehlvest - Timman went 9 ... Qc7 10 Qg3 b4 11 Na4!? Nc6 12
Nxc6 Qxc6 13 f3 O-O 14 Bh6 and now according to Ehlvest black could have
played 14 ... Nh5 with equality. (15 Qg4?! g6! 16 Bxf8 Kxf8 is good for
black because the white queen is in danger.)

8. O-O

[V] While not the most complicated line in this variation, it can
produce some hair raising lines. White's basic theme is f2-f4
followed by e5 if possible. Sometimes white will play f5 to
control the d5 square. Many times black will play Bb7 (or Nbd7
blocking the bishop's control of e6) and white will sacrifice
a piece (bishop or night) on e6.

Black's two main moves at this point are 8...Bb7 and 8...Be7. The
latter is currently considered superior. After 8...Bb7, white
usually plays 9 Re1 intending 10 Bg5 and piece play.

8. ... Be7

[B] 8. ... Bb7 was the alternative, but it's very dangerous. White often
gets the opportunity to sacrifice on e6 then. We're expecting white to
try Qf3 or f4 now.

9. Be3

[B] None of us had ever seen this move before. We'd spent many hours
looking at 9 Qf3 and 9 f4 varaiations and had decided that black was OK
in the main lines. We found 9 Be3 mentioned in ECO but in the game
referenced neither side played exceptionally well. Castles seemed like
the most natural reply.

9. ... OO
10. f4

[B] This move looks kind of scary, but it would have been more scary a move
ago. As nearly as we can figure white hasn't gained anything with Be3.
It doesn't help support either of white's pawn breaks (e4e5 and f4f5),
and it gives black an extra tempo to develop. We've looked at lots of
varations and we're pretty happy with what we've seen.

[V] This, for better or worse, is what I have been aiming
for. Black has several possible defenses: those with
b4 and those without. There is very little book on
this obscure line. Kavalek, in 1974, recommended 10...Bb7,
but 11 e5! de (11...Ne4 12 Ne4 Be4 13 Qg4) 12 fe Nd5
(12...Ne4 13 Ne4 Be4 14 Qg4; 12...Nd7!? 13 Qh5!) 13 Bd5 Bd5
14 Nd5 Qd5 15 Nf5!! Nc6! (15...Qb7 16 Qf3!) 16 Qd5 ed 17 Rad1
Rad8 18 Ne7 Ne7 19 Bc5 Rfe8 20 Bd6! +- (Valvo-West 1986).

Usually recommended (Nunn) is 10...b4 11 Na4 Bb7 (11...Qc7!?;
11...Ne4? 12 f5! d5 13 fe fe 14 Nf5!! Nc6 15 Nb6 Rb8 16 Nd5 +-
from a 1974 ICCF correspondence game of mine) and most opening
authorities give black the nod. If this position is reached, I
would try an experimental idea of mine: 12 e5!? (12 Qe1?! Be4!)
de 13 Ne6!

Another black try is 8...Qc7 with the idea 9 f5 b4 (9...e5
10 Nde2 Bb7 11 Nd5 Bd5 12 Bd5 Nd5 13 Qd5 Nd7 14 c4! += from
a recent game of mine) 10 Na4 e5 11 Ne2 (11 Nf3!? Bb7 12 Ne2 Bb7
13 c3) Nd7 12 Ng3 Bb7 13 c3 with a black edge in
Soltis-Dzindishvilli, 1987.

10. ... b4

[B] This is consistent with black's idea of putting pressure on the e pawn
and the long diagonal. We're not necessarily planning on taking the pawn
right away but now seemed like a good time to misplace white's knight.
We considered 10 ... Bb7, but then white can play 11 e5 right awaay.
11 ... de 12 fe Nfd7 reaches by transposition a position that is known
to be good for white. After 10 ... b4 11 Na4 Bb7 the same idea doesn't
look as good for white. since after 12 e5 black can play 12 ... Nd5.

Incidently, from the way I've presented our conclusions it may seem
like they're based solely on positional principles. That's not
entirely true. We've spent a lot of time looking at representative
variations and come up with our conclusions based largely on them.
There are some wild possibilities coming up in this game. White will
probably have numerous opportunities to sacrifice a piece on e6 soon.

11. Na4 ...

[B] 11 e5 was also possible in this position. It's been tried before
without much success, but white gets some attacking chances after 11 e5
bc3 (11 ... de5 12 fe5 bc3 13 ef6 Bxf6 bc3 Bb7 Qg4! gives white a
dangerous attack) 12 ef6 Bxf6 (12 ... cb2? 13 fe7 ba1(Q) 14 ef8(Q)+
loses a piece) 13 bc3 (13 Qf3?! cb2 14 Rd1 d5 15 c4? b1(Q) 16 Rxb1 dc4!
17 Qa8 Bxd4 is good for black.)

[V] The Bono team has chosen the "book" recommended line. At
this point, black can choose between (a) 11...Bb7! (b) 11...Qc7!?
and (c) 11...Ne4?

A game of mine against Popovych (~2350) proceeded 11...Bb7 12 e5! de
13 Ne6! fe 14 Be6 Kh8 15 Qd8 Rd8 16 fe Ne4 17 Nb6 Bc5 18 Bc5 Nc5
19 Bc4 Nbd7 20 Na8 Ra8 21 e6 Nf6 22 Rad1 Bc6 23 e7 Nb7 24 Ba6! Re8
25 Rf4 Nc5 26 Rc4 Nfe4 27 Rd8 g6 28 Re4! Ne4 29 Bb7! 1-0 (1987)

11...Qc7!? is reasonable and white can try 12 f5 transposing to
10...Qc7 11 f5 b4! 12 Na4. I have played 11...Qc7 12 c3 Bb7 13 cb Be4
14 Be6!? with unclear complications.

11...Ne4? was dealt with in my last posting: 12 f5 d5 13 fe de 14 Nf5!!

11. ... Bb7

[B] We looked at taking the pawn with 11 ... Nxe4 but we decided that white
gets too many attacking chances after 12 f5! d5 (12 ... ef5 or 12 ... e5
13 Bd5! looks good for white.) 13 fe6 fe6 14 Rxf8+! Bxf8 15 Qf3 with the
idea of Rf1. Maybe black's position would be defensible but it didn't
seem like the variation was worth persuing.

After 11. ... Bb7 we believe that white is practically forced to
sacrifice a piece with 12 Nxe6!? if he doesn't want to concede an
advantage to black. If he tries 12 f5 e5 13 Nf3 Nbd7 black can
practiclly take the e pawn at his leisure. 12 e5 Nd5! makes 9 Be3 look
pretty bad. White is practically forced to part with one of his
bishops.

We've spent a lot of time looking at 12 Nxe6 and we think that black has
several promising defensive plans. 12 ... fe6 13 Bxe6+ Kh8 14 e5! is
practically forced. Now black can try 14 ... de5, 14 ... Nd7, 14 ...
Ne8, or 14 ... Ne4. The most ambitious of these is probably 14 ... de5,
but this is also the most dangerous. The safest way for black is 14 ...
Nd7 15 de6 Bf6. White then has three pawns for his piece, but he has no
immediate attack and black should have plenty of time to develop.
Ultimately we think that the piece should be stronger than the pawns.

All in all we feel pretty confident about black's chances after 12 Nxe6.
We will probably spend the next few days getting some more concrete
analysis of the alternatives for us on move 14.

12 e5

[V] We are now out of book, although I have published an article on
this line in the Atlantic Chess News a couple of years ago. I
have seen two games published (other than my own) on 12 e5!? At
this point, black has two choices: (a) 12...de! or (b) 12...Nd5.

The latter 12...Nd5 was published in New In Chess #7. The main
line runs 13 Bd5 Bd5 14 Nf5! Nd7! 15 Nd6 Bd6 16 ed. Black has
several choices here, but I think 16...Qb8 is best. I have also
seen 16...Qa5 and 16...Rc8?! White will retain a very small edge
and black will have limited counterplay chances.

The best chance for equality and more is 12...de! 13 Ne6 fe 14 Be6 Kh8.
The other game I saw published was 15 fe!? Nd7! 16 Rf8 Nf8 and although
white won, he is worse in this position. I would play 15 Qd8!, leaving
black the difficult choice between 15...Rd8 16 fe Ne4 17 Nb6 and 15...Bd8
16 fe (16 Bc5!?) Nd7.

White's whole premise behind this line is he will enter an endgame with a
rook verses two minor pieces. Normally, this favors the pieces. However,
I have looked at the ensueing pawn structures and concluded that the rook
is superior to the two pieces even if white has no pawns for compensation!

[B] We got our first real surprise in game 1. Mike's Team played a move
that we thought was just plain bad. Unfortunately we were just plain
wrong! What it means is that we don't have nearly as many options as
we thought in the position. White will probably sacrifice a piece next
move and our reply and probably our next few moves will be completely
forced. I still think we're OK in the position, but I'm a little
concerned that there may be some more surprises in our future.

In my last message I said that we thought that 12 e5 wasn't very good
because black could then play 12 ... Nd5. We figured then white would
probably have to give up one of his bishops with 13 Bxd5 Bxd5 and black
would have the edge. It wasn't until Mike actually played 12 e5 that
we noticed that we hadn't looked far enough ahead in that line before
judging it. After 13 ... Bxd5 white has the annoying follow-up 14 Nf5!
which threatens 15 Nb6 and 15 ed6 Bxd6 16 Nxd6 Qxd6 17 Bc5. Either of
which picks off the exchange. 14 ... ef5 is terrible after 15 Qxd5 so
black has to try 14 ... Nd7. But white gets the advantage after 15 Nxd6
Bxd6 16 ed6. Black can probably win the pawn back with 16 ... Qb8 or
even 16 ... Qa5, but white will get a strong kingside attack. None of
this looks very appetizing.

So after 12. e5 black has to move his knight someplace else or play 12
... de5. 12 ... Nd7 is bad because of 13 ed6 Bxd6 14 Nxe6. 12 ...
Ne8 and 12 ... Ne4 allow 13 Qg4! which looks pretty scary. By the
process of elimination we decided that the right move was probably:

12. ... de5.

[B] We don't back down! Now 13 fe5 allows 13 ... Nd5 with a very good
position for black so white has to play his piece sacrifice now. We
think that white will play 14 Nxe6. I don't know who has the advantage
in the position then. It seems to me that black has enough defensive
resources in the position to hold. White appears to have a forced
drawing line, though, so black definitely isn't winning.

13 Ne6

[V] Well, it's a done deal: white has offerred a piece.
Black could still refuse it with 13...Qd1 14 Rad1 ef!?,
however 15 Bc5! leaves white with the edge. Belle has
been busy for several moves now looking at the consequences
of lines like 13...fe 14 Be6 Kh8 15 Qd8! Bd8!? 16 Bc5!?

[B] Well, Mike finally made his piece sacrifice in game 1. We're not too
worried yet, though. We believe that white can force a draw, but that
it will be very risky for him to play for a win. He probably has to
try to play with a rook and 2 pawns against two potentially active
minor pieces. In theory this would be good for us. I have to admit,
though, that this is just the kind of position where having a computer
would be a big advantage.

So white decided to go for it. There really wasn't much choice now
because 13 fe5 Nd5 was good for black. Actually I suspect that there
wasn't really any decision made here at all. White must have planned
this several moves ago and maybe before the game even started.

13. ... fe6

[B] We assume white will play 14 Bxe6+ Kh8. He can then try 15 Qxd8 or 15
fe5. We think that the main line is 15 Qxd8 Rxd8 16 fe5 Ne4 17 Rf7 Nc6
18 Bf5! which forces Kg8 19 Be6 Kh8 20 Bf5 ... with a draw by repetition.

Of course, even if this is the best line it's highly unlikely that white
will play it since he's playing to win the game. Instead white will
probably try to win an exchange with a timely Nb6. His problem is that
his e pawn is in danger and, especially if he leaves the queens on, black
will get some dangerous threats on the a8h1 diagonal.

14 Be6 Kh8
15 Qd8

[V] As mentioned earlier, black's two choices are 15...Rd8
and 15...Bd8. After my game with Popovych, he was kicking
himself for not playing 15...Bd8. White then, however,
can choose between an immediate 16 Bc5!? or 16 fe.

15...Rd8 seems more natural and we would play 16 fe. The
best for black is then 16 fe Ne4, although 16...Ng8 is
playable.

There are many traps for both sides and even a computer
won't necessarily successfully negotiate them. Black has
a temporary material advantage, but lags behind in piece
development and harmony.

[B] There's been an unusual turn of events in game 1. Mike played into the
main line by trading queens. I don't know for sure, but I think he was
intending to play the forced drawing line I described in my last
posting. We decided to avoid it, though! We played a move that I'd
previously thought to be unsound. Whereas I was 97% sure the other
line was a forced draw, I have no idea what's going to happen now. It
should be amusing for the spectators in any case.

This was the move we expected. 15 fe5 Nfd7 16 Qh5 Qc7 threatens the e
pawn, the c pawn, and ... Qc6. White can win an exchange back but
black's pieces will get very active.

As I said in my last posting 15 ... Rxd8 16 fe5 Ne4 17 Rf7 Nc6 18 Bf5
forces 18 ... Kg8 19 Be6 Kh8 20 Bf5 ... with a draw by repetition. The
only other try I was able to find for white after 15 ... Rxd8 was 16
fe5 Ne4 17 Nb6 but after 17 ... Bc5 18 Bxc5 Nxc5 and then either 19
Nxa8 Nxe6 or 19 Bc4 Ra7 black seems to be better. I guess I won't find
out what white would have tried after ... Rxd8 until after the game
because we decided to avoid the whole variation with:

15. ... Bxd8

[B] This is an extremely risky move aimed at refuting white's whole idea.
Now white doesn't have any obvious forced drawing ideas and he's down
material. Of course black's pieces are not well placed and his back
rank is weak. So who's winning? I don't know. We haven't found any
forced wins for white, but then again we don't have a calculating
machine to help us.

Possible tries for white from this position include 16 fe5 and 16 Bc5.
We haven't been able to find anything particularly good for white in
either variation. 16 Bc5 wins an exchange back so he may play that if
he can't find anything else.

[V] This (15...Bd8) is a reasonable alternative to 15...Rd8
although the latter is much more logical with the continuation
16 fe Ne4 17 Nb6. I then think black's best is 17...Ra7 and
the computer calculates that white has nothing more than a
draw with 18 Na4. We had two other intended lines at that
point: (a) 18 Rad1 Rd1 19 Rd1 Nc6! 20 Bd5 Ne5 21 Nc8 Bc8
22 Ba7 Nf6 23 Ba8!? with ideas of pushing the c-pawn. (b)
18 Rf7!? Nc6 19 Raf1 and a complex game is in the offing.
Black could also delay ...Ra7 with 17...Bc5 18 Bc5 Nc5
19 Bc4 Ra7 (Popovych played 19...Nbd7!?; 19...Nc6 is another
choice). I have looked at this lightly (20 Rf7) and feel
white is doing well, but haven't reached any conclusion.

16 fe

[V] After 15...Bd8 there are two main lines: 16 Bc5 and 16 fe.
16 fe is more thematic, but 16 Bc5 recovers immediate
material. You can see how this latter consideration interferes
with the computer's ability to see clearly in this kind of
position.

The computer evaluates 16 Bc5 Re8 as .25 pawns in Black's
favor after 11 ply. Despite the fact the computer considers
this better than 16 fe, we forced it to play down the following
line: 16 fe Nfd7 17 Rad1! (17 Nc5? Bb6!; 17 Nb6!? Bb6 18 Bb6
Re8; 17 Rfd1!?) Rf1 18 Rf1 Bc6 19 Nb6 Bb6 20 Bb6 followed by
21 Bc7 with a terrible bind. If 20...g6, both 21 Rf7 with ideas
of 22 Bd4! and 21 Bd8!? (with ideas like 22 Rf8 Nf8 23 Bf6 mate!)
are dangerous. 18...h6 or 18...g6 are also possible, but white
seems to obtain a terrible attack in all lines.

The computer initially sees 16 fe as .6 pawns in blacks favor.
After 16 fe Nfd7 17 Rad1 Rf1 and ten more ply, that evaluation
improved to .4 in black's favor. I personally have been concerned
that since the computer doesn't see real material coming back
within its beam of vision that it neglects situations where
black remains all tied up. Ken assures me, though, that it
understands.

This is best. 16 Bc5 Re8 17 Bf7 would have won back an exchange but
after 17 ... Bc6! white isn't doing very well. He doesn't have time to
win the e-pawn and the black pieces will get active.

[B] Mike spent 4 days on this move (I'm assuming it was this game that he
was working on) after playing the rest of the the game quickly (usually
replying within 2 days). I'm hoping that this means we surprised him
with 15 ... Bxd8 and that he hasn't found a forced win yet.

16 ... Nfd7

[B] For a while we thought about returning the piece in this position with
16 ... Re8 17 ef6 Bxf6 18 Nc5 Bc6. Black would be a pawn down, but we
thought we'd have some pressure against the white queen side and could
get developed with ... a5 and ... Na6 or ... Ra7-e7. It dawned on me
after a while, though, that white could exploit black's lack of
development immediately with 19 Rxf6! ef6 20 Bd4 Kg7 21 Rf1 Rf8 22 Bc4!
and white is killing black.

17 Rad1

[V] This is a very strong move that may not have been anticipated
because it appears to lose a move after 17...Rf1 18 Rf1. I feel
it is in fact necessary because Black threatens 17...Re8 breaking
White's hold on the postion.

Black must now react to White's threat to his d7-Knight. I don't
think Black has any choice but to play 17...Rf1. We plan to
simply reply 18 Rf1! (18 Kf1!? was Belle's choice for a long time,
but Black gets a key tempo with an eventual ...Bb5 check in some
lines: 18 Kf1 Bc6 19 Bb3 Bc7 20 e6 Nf6 21 Bb6 Ba4 22 Bc7 Bb5+!).

I personally think White is winning and that 15...Bd8 was a
mistake. Popovych thought it was good, but I don't think he ever
considered 17 Rad1 carefully. Based on general considerations,
15...Rd8 looked better: more harmonous piece placement and
cooperation.

[B] Again, this is the best move. White has to keep the pressure on. Now
he threatens 18 Bxd7 as well as 18 Rxf8+ Nxf8 19 Rxd8. Our reply is
again forced.

17 ... Rf1

[B] Mike has played what we believe is his best continuation in game 1. Our
last two moves have been forced. We're on a roller coaster ride that
won't end until we finish our development or lose some material. We're
still feeling optimistic that we'll be able to survive the attack.

We also considered returning the piece now with 17 ... Re8 18 Bxd7 Nxd7
19 Rxd7 Bc6 20 Rf7 h6. Now black seems to win the e-pawn since he is
attacking white's knight and 21 Rf8+? Rxf8 22 Rxf8+ Kh7 wins for black
because he sill threatens the knight and also ... Bg5 winning material.
Unfortunately, though, black is still losing after he gets the e-pawn.
21 Nb6! Bxb6 22 Bxb6 Rxe5 23 Rxg7! Kxg7 24 Bd4 Re8 25 Rf5 leaves white
two pawns ahead.

We believe white's best move now is 18 Rxf1 which leaves black terribly
jumbled up on the queen side. 18 Kxf1 Bc6 19 Nc5 Bb6 appears to be OK
for black.

I have to confess that my gut feeling about the position after 18 Rxf1
is that white should be winning, but after looking at the position for
over two weeks we still haven't even found any drawing lines for him.
This kind of position really doesn't lend itself to feeling or
intuition. Either white has a winning plan or he doesn't. We're
probably not going to blunder at this point and neither is he.

18 Rf1

[V] This is a choice point for Black: (a) 18...Bc6 (b) 18...h6
(c) 18...g6 (d) 18...a5.

18...Bc6 is probably the main line. After 19 Nb6 Bb6 20 Bb6,
Black will again be at a juncture point (20...a5, 20...h6,
20...g6). White's plan will be direct: Rf7, Be6-somewhere,
and e6. The other bishop will probably go to d4 in order to
pressure the g7 square.

18...h6 and 19...Kh7 is also possible and white will follow
the same blueprint as above.

[B] White's plan is Rf7, Bd4, B at e6 moves, and then e6 winning.

18 ... h6

[B] Along with many threats to win material white was also threatening mate
in a number of variations. The move ... h6 at least puts a stop to
that. We also considered 18 ... Bc6 but after 19 Nb6 Bxb6 20 Bxb6 h6 21
Bd4 white's plan of Rf7 ... seems unstoppable.

White's next move initiates his plan. He can't play 19 Bd4 first since 19
... Nc6 gets black out of the bind.

19 Rf7

[V] I consider 18...h6 Black's number two defense. He plans to
both eliminate the back rank mate threat and, by playing
...Kh7, to neutralize to fork on his d8 Bishop.

The game's direction depends on Black's next. If 19...Bc6,
then White will play 20 Bb3. If 19...Kh7, White will
continue with 20 Bf5.

The key to White's placement of his e6 Bishop is Black's
ability to unravel his Queen-side. Once he plays ...Bc6,
the b8 Knight cannot occuy c6 (hence Bb3). Otherwise, White
will keep that Bishop on the d7-h3 diagonal.

In either case, White's subsequent plan is Bd4 and e6.

19 ... Be4

[B] If our analysis is correct this move deserves an exclamation point or
two! The idea is to retreat back to g6 to force the white rook off the
seventh. White can't win a piece with 20 Bxd7 Nxd7 21 Rxd7 because 21
... Bc6 wins material. We think that white's best now is to win two
pieces for his rook now or on the next move with Rxd7. This leads to an
endgame where white has a pawn or two for the exchange.

The alternative to 19 ... Be4 was 19 ... Bc6. White would have gotten
good winning chances in that line with just 20 b3 and continuing with
his plan of Bd4, etc. Our analysis indicated that the best that black
could have done then was to get into an ending a pawn down.

Game one is proceeding along the lines we were expecting. We've
reached a critical position and we believe our last move was an elegant
saving resource for black. We assume it wasn't a surprise for Mike,
though, since Belle must have seen it coming. We've spent several
weeks looking at variations and we believe the best white can do is get
an endgame where where he has two pawns for the exchange. We'll be
very disappointed if Mike or Belle manages to refute all our analysis.

20 Nc5

[V] We didn't consider 19...Be4, but, in any case, it does not
help Black much. Belle considers the text to lead to a edge
of .6 Pawns to White after 13 ply.

19...Be4 is based on two ideas. First, White cannot play
20 Bd7 Nd7 21 Rd7 because of 21...Bc6 winning. Also, Black
wants to play 20...Bg6 to drive the White Rook from its f7
post. These seem like very powerful ideas, but the placement
of the Bishop on e4 permits the text because Black no longer
has 20...Bb6 as a defense like before: 20 Nc5 Bb6 21 Bd7 Nd7
22 Rd7 Rc8 23 Ne4! with a big plus (23...Be3 24 Kf1 Re8!?
25 Nd6! Re6 26 Nc4).

Belle sees two possible defenses for Black, but both are
inadequate:

(1) 20...Bg5 21 Ne4! Be3 22 Kf1 Ne5 23 Rf8 Kh7 24 Bg8 Kg6
25 Bd5.

(2) 20...Nc5!? 21 Rf8 Kh7 22 Bg8 Kg6 24 Rd8 and once Black
secures his c5 Knight 25 e6! is decisive because 26 Bf4
is threatened and the b8 Knight cannot go to c6 without
disrupting the Bishop's protection of the a8 Rook.

Also interesting was 20 c4!? Black cannot play 20...Bg6 because
21 Rf2 and 22 Bd5 is threatened. Best is 20...bc! 21 Bd7 (21 Nc3?!
allows Black to unravel [alas!]) Nd7 22 Rd7 Bc6 23 Rd4 Ba4!
24 Ra4 cb 25 Rb4 Bc7 with a dead draw.

[B] I have to confess that I spent quite a bit of time looking at this move.
Somehow I convinced myself that 20 ... Bg6 21 Rf3 Bh5 was OK for black.
If white moved his rook off the third rank we'd play Bb6 and if 22 Rh3
we could play 22 ... Be8. Somehow I completely missed 22 g4 which wins
easily. In fact, I showed this variation to my teammates and they
didn't notice either!

20 ... Bg5

[B] Before I thought of 21 ... Bh5?? I decided that this would be the best
move. As it turns out it's the only move. 20 ... Bg6 21 Rf3 Nxc5 22
Rf8+ Kh7 23 Bg8+ wins easily. 20 ... Nxc5 21 Rf8+ Kh7 22 Bg8+ Kg6 23
Rxd8 followed by e5-e6-e7 is crushing.

21 Nxe4

[V] The Bono team replied to our last moves in a matter of
hours! Nonetheless, Belle sees nearly a two Pawn advantage
to White after 12 ply. Belle sees the main line as 21 Ne4 Be3
22 Kf1 Bd4 23 Bd5! Ra7 24 e6. We sent an if move of 22 Kf1
if in fact they play 21...Be3 (What else?).

I suspect they have overlooked something relative to our
21st move.

Belle also thought 21 Bf2 was worth considering, but that
21 Ne4! was simply better.

[B] The good news in game 1 is that things are finally clarifying themselves
a bit. The bad news is that I made an analysis error in a critical
variation. I discovered this unfortunate fact after I sent Mike our
move 19 ... Be4. Fortunately for us things don't look too terrible. I
think we will get some drawing chances in the ending that will result
from all this.

I still like this move (19... Be4)! The concept is so elegant! Too
bad you can't impress a computer. If our current analysis is sound then
it was still the best move in the position.

21 ... Bxe3+
22 Kf1 ...

[B] Now 22 ... Nxe5 isn't very good because of 23 Rf8+ Kh7 24 Bd5.

22 ... Bd4

[B] This seems to be the only move that gives black a tenable ending.
Taking the knight now doesn't seem to lead to a win for white, so his
best seems to be 23 Bd5. Now 23 ... Ra7 loses, but I think black can
hang in there with 23 ... Nc6! and it seems like white can win his piece
back only by giving up the e pawn, so black gets drawing chances because
of the opposite colored bishops.

23 Bd5

[V] It now appears that Black will give back a piece with 23...Nc6! and
hope to hold the bishops of opposite color endgame. White's King is
already in the center and that, among other things, will make it more
difficult for Black.

We expect the main line to be 23 Bd5 Nc6 24 e6!! Nde5 25 Rc7! and the
c6 Knight cannot move without allowing e7 and Nd6 in some effective
fashion: 25...Rf8 26 Ke2 Nb8 27 e7 winning. A likely continuation is
25...Bb2 26 Bc6 Nc6 27 Rc6 and White is winning the R+N vs. R+B
endgame due to the e6 Pawn.

[B] As I said in my last note 23 Bxd7 didn't seem to lead to a promising
ending for white.

23 ... Nc6

[B] We finally get our knight developed! White can take either knight now
or he can play 24 e6. I don't think white can win the ending after 24
Bxc6 Nxe5 25 Bxa8 Nxf7 26 c3 bc3 27 bc3 Be5 28 c4 Nd6. After 29 c5
Nxe4 30 Bxe4 ... we get to a typical drawn opposite colored bishop
endgame. White can't push his pawn to the eighth because black can set
up an impenetrable blockade on c7. White can't win any of black's
pawns because he can only attack them with his king and black can
defend with his bishop.

For white to try to win an opposite colored bishop endgame, he has to
either keep more pieces on the board, or try to improve his position
first, say by getting his king more active. His two possibilities are
24 Rxd7 and 24 e6.

24 e6 is certainly scary, but black seems to get drawing chances if he
keeps his wits about him. 24 e6 Nde5 25 e7 Nxe7! or 25 c3 bc3 26 bc3
and now 26 ... Nxf7 loses, but 26 ... Bb6! looks OK for black. Then
27 Bxc6 Rf8+ 28 Rf7 Nxf7 29 e7 Rb8 20 e8Q Rxe8 21 Bxe8 Ne5 results in
an endgame that black can probably draw. The knight and bishop
cooperate very well to keep the king out and the queen side pawns back.
If white tries to trade knights he doesn't seem to get any winning
chances at all.

The best winning try I've been able to find for white in this variation
is to sacrifice the exchange with 24 e6 Nde5 25 Rc7 Bb6 26 Rxc6 Nxc6
27 Bxc6 Ra7 and white definitely has some winning chances, although the
position is somewhat double-edged.

24 Rxd7 Rf8+ 25 Ke2 Nxe5 26 Rb7 is another reasonable winning try. By
keeping the pieces on and trading b pawns white gets some real winning
chances. Black does get some counterplay, though, by making threats
against white's king and queen side pawns.

One reason I'm optimistic about all these endings is that the computer
probably won't be much help to Mike for a while. I sincerely doubt
that Belle knows anything at all about opposite colored bishop
endings. It's probably recommending variations that trade as many
pieces as possible.

24 e6

[V] White's problem is he can obtain various Pawn up endgames, but all
feature Bishops of opposite colors and likely drawing chances for
Black:

a) 24 Bc6 Nd5 25 Ba8 Nf7 26 c3 bc 27 bc Be5 28 h3 Nd6 29 Nd6 Bd6

b) 24 Rd7 Rf8 25 Ke2 Ne5 26 Rd6 Bb2 27 Ra6

The played move can also lead to another such endgame: 24 e6 Nde5 25 e7
Nf7 26 Bc6 Rb8 27 e8Q Re8 28 Be8

We tried winning each of these, but none seemed promising.

We next considered 24 e6 Nde5 25 Nd6!? This works except for 25...Rd8!
26 Be4!? g6! and White loses!

So we are sticking with our original idea of 24 e6 Nde5 25 Rc7.

[B] Geez! These guys just won't take presents! We've offered to return
the piece twice in the last two moves, but they keep ignoring us! It's
still not clear to me whether this or 24 Rxd7 was the best move.

24 ... Nde5

[B] Forced. Now white has to be a little careful since black is finally
getting his pieces coordinated. The e pawn is very dangerous now,
though, so white still has the edge.

25 Rc7

[V] Ken left for Japan today and won't return for two weeks. On August
3rd, I leave for London and the Computer Olympiad. I will return on the
17th. We won't be making moves in this period, except:

1) The period when Ken is back and before I leave for London
(~7/29-8/2).

2) When in London (8/7-8/16), Ken and I will try to make moves over the
phone.

White threatens to recover his piece and avoid a Bishops of opposite
color endgame with 26 Bc6. Black has two defenses:

a) 25...b3!? 26 ab Rf8 27 Ke2 Nb4 28 e7 Re8 29 Bb7 (29 Nd6!?) Ng6 30 c3
Be5 31 Rd7 Ne7 32 cb Bh2 unclear.

b) 25...Bb6 26 Bc6 Rf8 27 Rf7 Nf7 28 e7 followed by e8 and winning the
Rook back. This is probably drawn.

We intend 25...Bb6 26 Rc6! Nc6 27 Bc6 Rf8 28 Ke2 Bd8 29 c4! and though
material is roughly equal, White appears to have a winning endgame.

[B] Threatening to take the knight and hold the e pawn. We can't hold the
piece and stop the pawn so we have to offer to return the knight again.
White will have to take it now, though.

25 ... Bb6

[B] Mike's team played the most tactical continuation in game 1. We expect
them to sacrifice the exchange next move, and they will get some winning
chances in an unbalanced position. We still don't known whether the
position is objectively drawn or not.

As I said in my last note, we're counting on this move to save us. Now
26 Bxc6 Rf8+ 27 Ke2 (27 Rf7 Nxf7 28 e7 is similar) Bxc7 28 e7 Rb8 29 e8Q
Rxe8 30 Bxe8 Bb6 leads to another opposite colored bishop ending that we
can probably draw, so we expect white to try 26 Rxc6 Nxc6 27 Bxc6 Ra7.
We still aren't sure whether that endgame is a draw or or a win for
white, but we have found a number of possible drawing ideas for black.

26 Rc6

[V] If White doesn't make this attempt to win, he may as well agree to
a draw here. Though he has several ways to enter a pawn up end games,
none offer much in the way of winning chances. We expect that the next
series of moves are pretty much forced:

26 Rc6 Nc6 27 Bc6 Rf8 (27...Ra7 28 Bd7 Bd8 29 Ke2) 28 Ke2 Bd8 29 c4 and
White's advanced Pawns will give Black fits.

The computer thinks it is a Pawn up in this endgame. I have tried
unsuccessfully to defend the Black position against Belle, but the
Black pieces become passive and White can always inch forward.

[B] We're moving again! Ken Thompson is back from vacation so the Valvo
team has finally replied to our last moves. I understand that Mike is
leaving for England shortly so we may get another break in the action
soon.

It's been a while since we've started these games so here's a quick
summary of what's going on:

Last April I challenged IM Mike Valvo to an EMAIL consultation
chess match. I selected a consultation team and offered Mike
the opportunity to do likewise. In addition to myself, my team
included Bill Kelleher, Larry Tapper, and Adrian Casillas -- all
masters in the Cambridge area. Mike accepted with a team
consisting of himself, Ken Thompson, and the Belle Chess
Computer. We agreed to play two games so both teams could get
to be white and black.

Since April we've been playing a move in each game every two or
three days, but we've had to suspend play from time to time
while people have been on vacation or away. Periodically Mike
and I have been posting messages to rec.games.chess with
comments on how we think the games are going. We've agreed not
to read each other's postings until after the games are done.
Mike's postings all have "Valvo - Bono Match" in the subject
line and mine all have "Bono - Valvo Match".

In game 1 as black we accidently played into one of Mike's pet
lines in the Najdorf Sicilian defense. Mike sacrificed a piece
on move 13 and we've been defending ever since. Larry Tapper
spent some time talking to Mike at the World Open and found out
that Mike has known about this piece sacrifice for years and that
we did well to get any kind of playable position out of that
opening. We've spent most of the game so far playing for a draw
and, from that point of view, things are looking very well right
now.

All in all the games have been very interesting so far. It's
too bad that we didn't play the openings better, but we've been
hanging in there since then. In fact, I think we're doing well
in both games right now.

This (26 Rxc6) is white's only chance to try to win the game. 24 Bxc6
Rf8+ 25 Ke2 Bxc7 26 e7 Rb8 27 e8Q Rxe8 28 Bxe8 Bb6! is pretty clearly
drawn. White's king can't get in unless he can trade off the black
super knight, but then the opposite colored bishops will insure a draw
for black.

26 ... Nxc6
27 Bxc6 Rf8+

[B] In previous notes I indicated that we were planning on playing 27 ... Ra7
here and that we thought that black got good drawing chances. I still
believe that to be true, but after looking at a lot of possibilities it
occured to us that if black isn't losing this ending then he might also be
able to get some winning chances of his own. 27 ... Rf8+ activates the
rook and threatens in some variations to move to the 7th or 8th rank to
start attacking white's pawns. So if white gets a little overambitious,
black might start to think about winning this game.

28 Ke2 Bd8

[B] Forced. We have to stop the e pawn.

29 c4 ...

[B] 29 Kd3 was dangerous since it allowed 29 ... Rf1 (30 Nd6?? Rd1+).

29 ... bc3

[B] Now white has to decide whether to take with the knight and play for
Nd5, b4, a4, b5 or with the pawn to play for c3-c4-c5-c6. Either way
black will reposition his rook and move his king to the center to help
blockade the pawns.

30 bc

[B] We expected 30 Nxc3. We thought the main line was 30 Nxc3 Rf5 31 Kd3
(31 Nd5 Re5+ wins the e pawn) g5! which stops Nd5-f4-g6. Then after 32
b4 Kg7 black isn't doing badly. He'll blockade the e pawn with his king
and the eventual passed b pawn with his bishop and use the rook to
harrass white from the rear.

30 ... Rf5

[B] This move does several good things. It sets up a temporary blockade on
the fourth rank and it vacates the f8 square so the king can come to the
center. Black needs his king to help blockade the e pawn.

31 c4

[V] Although we considered this idea earlier (Rf5), we didn't consider
it in this position. It is clearly best. The point is the f8 square is
cleared so a path is available for the Black King to get to e7.

We had only considered 30...Be7 in order the clear the f8 square by Rb8
or Rc8. The Black Bishop is best off where it is so it can control the
advance of both White passed pawns.

While I now wonder if in fact the game is now drawn, we still have a
couple of good ideas: (1) Bd7 to force the Rook off of f5 (threat e7)
and (2) the direct g4.

A possible sequence is 30...Rf5 31 c4 Kg8 32 Bd7 Be7 (32...Ra5 33 a4)
33 Ke3 Rf1! intending to sacrifice back the exchange after 34 c5 by
34...Re1, 35...Re4 and 36...Bc5. 33 Kd3! Rf1 34 c5 doesn't appear to
help because the rook can attack the knight: 34...Rd1 35 Kc3! Re1, but
36 c6! saves the day as the c-Pawn threatens to queen.

[B] In game 1 the Valvo Team surprised us a little by making a passed c pawn
instead of aiming for an eventual passed b pawn. It seems to us that
this makes our task of blockading his pawns somewhat easier, although we
thought we were OK in the other line too.

31 ... Kg8

[B] White naturally starts to push his passed pawn and black promptly brings
his king to the center. Black had to play Kg8 right away. After 31 c4
white was threatening 32 Bd5 which would prevent ... Kg8 because of
e6-e7+

We expect white to play 32 Kd3 now and then we'll continue moving our
king towards the center with 32 ... Kf8.

32 Bd7

[V] Belle and I have practically (individually!) lost hope of winning
this one. The f5 Rook move was the back breaker. Black can now
play 32...Be7 (looks forced although 32...Rf4 is possible) and
White cannot force c5 without allowing a simplification (32 Bd7
Be7 33 Ke3 Kf8 34 Kd4 Rf1 35 c5 Rd1 36 Ke5 Rc1= or even 33...Rf1
34 c5 Re1 35 Kd4 Re4= or 33 Kd3 Rf1 34 c5 Rd1=). The try 32 g4
Rf4 33 h3 h5! is too risky.

White may try to safeguard all his loose Pawns before trying to
advance his King in the center: 32 Bd7 Be7 33 a4 Kf8 34 g3 Ba3
35 h3 Ke7 36 Kd3 Rf1 37 c5 Rd1 38 Kc2 Re1 39 Kb3 Re3 40 Ka2 Bb4
41 c6!? Re4 42 c7, but this, too, is clearly drawn.

We offered a draw with our last move.

[B] Well, the Valvo team has finally realized that they're not winning game
1. They offered us a draw with their 32nd move. We're not going to let
them off that easy, though. We now believe we have good winning
chances, so we're playing on.

Sometime around move 22 we noticed the possibility of Mike's exchange
sacrifice on move 27. Our first impression (like Mike's I suppose) was
that this endgame was probably winning for white. We originally played
into it because it seemed like our best chance to draw, but after a
while we began to get a completely different feeling about the
endgame. We decided that, in fact, it was white and not black who was
in danger!

Rather than give a lot of concrete analysis about the position, I'll try
to explain the steps that led us from our original theory to our current
one.

Firstly, why did we think our position was bad? White's e pawn was very
far advanced and there were immediate threats to push it through.
White's king was more active and could be used as a weapon. Black's
b-pawn was weak. His rook wasn't active.

Direct analysis showed that there were solutions to all these problems.
The opposite colored bishops helped black stop the e-pawn. Black
immediately set up a blockade on e7. Also, black was able to quickly
activate his rook via the Rf8-f5 maneuver that we played on moves 27 and
30. Although white's king started out more active, it was too dangerous to
advance it immediately, since the black rook could have gotten to the
eighth rank where it could harass white's pawns. And the only way for
white to get a second passed pawn was to trade off the weak black
b-pawn.

OK. So black can draw. Why does he have winning chances? Basically
because there are too many white pawn islands. It is very difficult
for white to set up a blockade position and defend all his pawns. If
black can blockade the passed pawns with his king and bishop the rook
can attack pawns on both sides of the board at once. White's minor
pieces can't easily defend them.

I don't know if black is really winning, but there's no reason to accept
a draw here. White has a lot of defending to do. In a tournament
situation black might have accepted white's draw offer because he would
have been so relieved that he'd survived white's attack, but in a postal
game black had plenty of time to relax and look at the position
objectively.

Back to the current move:

We didn't expect the move 32 Bd7 since it appears to lose an important
tempo. White's going to have to waste a move now to safety his h pawn.
We thought white's best chance was 32 Kd3 Kf8 33 c5. Black then would
have numerous winning tries: 33 ... Rf1, 33 ... Bc7, 33 ... Ke7, 33 ...
Re5, and even 33 ... Bf6. I think the best of these is 33 ... Re5.
Here are a few variations: 33 ... Re5 34 Bd7 Rd5+ 35 Kc4 Rd1. Now if 36
Nc3 Rc1 37 Kd3 Bf6 38 Ne4 Be5 Black is ready to attack the a-pawn. 36
Nd6 allows 36 ... Rd2 and white's threats don't look strong enough to
compensate for the pawns he's going to lose. 36 Bc8 Ra1 37 Kb3 a5 38
Nc3 Rc1 also looks pretty good for black.

32 ... Rh5

[B] This move threatens the h pawn and repositions the rook to an safe
square. Now black has time to improve his king and bishop position. We
don't think white can afford to give black the rook pawn now. 33 c5
Rxh2 34 c6 Kf8 blockades the pawns. So 33 h3 appears to be forced.
Black will then play 33 ... Kf8 threatening Bb6 to stop the c-pawn on
the 4th rank. If then 34 c5 Rd5! cuts off the white king from
penetrating on the queen side.

33 h3

[V] When we were considering our 24th move, we then thought the
exchange sac that we actually played was our best chance to win. We
never considered 30...Rf5 or appreciated its subtle points: with Rf5,
Black is both able to vacate the f8 square for his King and restrain
the Pawn c5 advance. In addition, he doesn't have to waste a tempo with
the d8 Bishop.

Now we have an interesting development: they have declined our draw
offer! Odd. Lein once refused a draw offer from me in a position where
he had previously offered a draw and was a pawn down because I had
said, "OK, draw." "Draw?", he said in a raising voice as though I had
insulted him by offerring it in this manner. I apologized and said, "I
offer a draw" which was promptly accepted. There is a way of thinking
that if your opponent offers a draw, he must think he is worse and
therefore you should refuse.

We are sorry we screwed up this game and didn't go for a pawn up
endgame. We saw we no longer can win and therefore offered a draw. We
really didn't consider we could be losing! In fact Belle thinks we
still are better! I must admit that I feel personally insulted by this
refusal and have half a mind to resign on the spot.

In retrospect, Black must play this if he is trying to win because
White can force a draw after 32...Be7 by playing 33 g4 Rf4 34 Ke3 and
now (a) 34...Rf1 35 c5 Rc1 36 Kd2! or (b) 34...Rg4 35 c5 Rg1 36 c6 Rc1
37 Kd2!

Black must hope to run us out of moves. 32...Rh5 is designed to avoid
blocking the e7 square for the King.

[B] We thought this move was forced, but I was relieved to see it anyway.
With 33 h3 white is implicitly agreeing that he can't push his pawns
in, so he has to start defending.

33 ... Kf8

[B] Finally getting our king to a useful square. We are now ready to
activate our bishop.

34 Kd3

[V] I really didn't want to continue to play this game, but Ken/Belle saw no
reason to stop and I could see Ken was confused by the thought of giving
up. Apparently he wondered if this was some device I was trying to use to
get back at them for continuing. At first I thought of just letting Belle
play by herself, but I have moved off of that position. I will continue
to participate.

Belle, since she thinks she is still winning, wants to play 34 c5. This
would not be good because Black could force the Pawn to advance again with
34...Be7 and White would be overextended. Ken changed a parameter so she
now thinks a draw is very big. Hence, 34 Kd3. She expects 34...Bb6! and
appears to like 35 Nc3 as a reply. Black's intent is to eventually run
White out of moves, but at this point it doesn't look possible. Interesting
is the idea 34 Kd3 Bb6 35 Bc6!? intending 35...Ra5!? 36 Nc3!? Ra3 37 Ba4! and
White looks forward to Bb3 and Na4, but 36...Re5! destroys the White position.
The direct idea 34 Bc6 is also interesting intending 34 Bc6 Ra5 35 a4 Ke7
36 c5! and the Black Rook is trapped. This works well against 35...Be7
36 Kd4 as well, but 35...Bb6! stops White cold (key idea=control of d4).
The game would be over (i. e. drawn) right now if we had played 32 Bd5.
The played 32 Bd7 (with the draw offer) gave them a chance to play for a win.
Objectively, they do have winning chances if the Rook can penetrate to the
eighth rank and harass the White King side Pawns.

For what it is worth, I had prepared a resignation speech. We are not
resigning, but it expresses my feelings about the matter. I know I am setting
myself up for being judged by net readers and expect to be flamed. What the
hell, it is tiring watching Shogi dominate this group! I could pretend these
are not my feelings and act 'proper', but I won't. My feelings and
reasons may not be in harmony with those of others or what they may think
is correct or right. They are simply how I feel. The wonder, the enthusiam,
the interest of this whole approach is very flat for me right now. The spirit
has gone out. It is time to take my ball and go home.

Without any more comments, here it is as I prepared it two days ago. This
was done without consulting Ken and he has nothing to do with it:

"I am resigning this game and I suppose I owe an explanation. First of all,
I am more of an artist or poet than a competitor at this game. I don't make
a living from playing and, therefore, I don't have to be a prisoner to it.
In my last posting, I indicated indignation that our draw offer has been
refused. I have nearly gotten over that now that I see they have objective
justification for playing on. But the game is over for me. I have no more
heart for it. It was over after Rf5 when I saw that a win was not possible.
At that point the only issue was whether or not a win was possible. To
continue to play now would only mean we would be trying to determine if a 1,
0, or .5 would be assigned after a series of more moves. It would be like
starting another game from this point as this position no longer has anything
to do with the game we were playing.

This probably makes no sense to those for whom winning is the only thing.
Anyone who has seen my article 'Can Chess Help Adapt to Life?' already knows
my opinion about that.

The style of the Bono team has always been an irritation to me when I have
had to face such play in the past. The second game is a perfect example of
what I mean: They play a boring opening, continue by misplaying it by
combining h3, Bg5 and Qd2, don't play the most challenging line (Bh4!),
then retreat and proclaim, "Let's see you beat me!" This complements the
feeling I get from the first game at this point: "Now we are going to
punish you for playing for a win... " I have no interest in being tortured
for thirty moves: they want to be able to put a 1 after this game, so let
them!

This kind of street smart, live to play another day, scrappy attitude has
never been my approach to Chess. It is an important hobby for me and if
that were my attitude towards it, I would either give it up or find some
way to make significant money from it.

The Bono team deserves praise, though, for their conduct of the defense
in this game. I thought their Bd8 was a positional mistake (Rd8!), but,
in the end, may have been tenable. I had the advantage of having looked
at this line for many years, though I never had to face Bd8!? before."

[B] Allowing us to blockade the c-pawn on the fourth rank. This is a major
concession, but after 34 c5 Rd5 the white king is cut off from the queen
side and black has plenty of time to play ... Be7 forcing c5-c6 then ...
Bd8, ... Ke7 followed by ... Ra5 winning the a-pawn. The Pc6, Bd7, Pe6
configuration is very bad for white since his bishop is immobile and the
black king and bishop can easily blockade the e and c-pawns. In fact,
if black gets his king to d8 or d6 he can temporarily blockade both
pawns with the king alone. This theme comes up frequently in this
endgame.

34 ... Bb6

[B] Stopping the c-pawn and keeping the king from advancing to d4.
White doesn't seem to be able to improve his position anymore.

35 Nc3

[V] We basically are going into a mode of moving back and forth. This move
prevents an immediate Ke7 and postures White for playing Nd5 in some lines.
Black can never block access to Ra5 or White can play Bc6-d5 insuring the
draw (e. g. 35...Bc5? 36 Bc6! Ke7 37 Bd5=). Black's best seems to be Re5
when White has nothing better than returning his Knight to e4.

We expect Black will try some kind of zugzwang position by moving his King
to d8 (to eliminate Bd7-c8 repetitions). By then both sides will have
posted their a-Pawns on a4 and a5 respectfully (Black cannot allow a d5
Knight access to b4 and c6; White doesn't want a loose a-Pawn hanging
around). At that point White will play Nd6 to protect his g2-Pawn:
1...Re3 2 Kd2 Rg3? 3 e7+! Ke7 4 Nf5+.

[B] In game 1 the Valvo Team has set up a blockade and the issue is now
whether we can break it. We believe we can, but the position is still
extremely complicated. They may be able to escape by threatening some
tricky knight forks.

This move doesn't really threaten Nd5 because ... Bc5 followed by ...
Re5-Re1 is a good for black, but it does stop black from playing Ke7
right away and it threatens Bc8 followed by e7+ and Nd5 in some
variations. Now black would like to try to maneuver his rook in right
away, but 35 ... Rh4 36 Ne4 Rf4 c5 Bc7 Kd4 is good for white.

35 ... Re5

[B] White's best chance seems to be to try to hold the blockade with 36 Ne4.
Then black can't make any immediate progress, but white is almost in
zugzwang. Knight moves allow ... Re3+ followed by ... Rg3. The king
can't move since it protects the knight and all the bishop can do is
shuttle between d7 and c8. So black will aim for a complete zugzwang by
blockading white's pawns and playing ... Bg1 and ... Kd8 or ... Kc7. I
think this plan will work, but it's very tricky. The white knight has
many ways of harassing black by threatening knight forks and white can
use the threat of e6-e7+ to indirectly guard his king side pawns with
the bishop in some variations.

36 Ne4

[V] We wait while Black sets up his zugzwang position: K on d8, B on g1.
We, meanwhile will play our a-Pawn to a4. Black can optionally play
his Pawn to a5. Belle doesn't see Black winning anything in a 12 ply
search of that position, although it does see its value dropping
slightly each ply.

From a human perspective, I see (from that position) 1 Nd6 Re1 2 c5
with counterplay. In any case, we wait for Black to take his best shot.

36 Ne4

[B] White's just repeating moves now. Considering the fact that black can
indeed set up a zugzwang, I suspect that white should have tried 36 Nd5
Bc5 37 a4. Black will win a pawn, but his king is stuck on f8 so white
will get some counterplay.

36 ... Ke7

[B] Heading for our zugzwang.

37 Nc3

[V] We have nothing to do but wait. I suppose the game will continue
37 Nc3 Kd8 (or 37...Bg1 transposing) 38 Ne4 Bg1 39 a4 a5 40 Nd6. Black must
then make his play as 40...Ke7 41 Ne4 repeats and 40...Re3 41 Kd2 Rg3 42 e7+!
Ke7 43 Nf5+ fails. 37...Re3+ would be a mistake because Black would have
three targets available to a Knight on d5.

I still don't see anything for Black, and if they try too hard...

[B] No news is good news in game 1. The Valvo Team hasn't come up with
anything clever and we're slowly working towards our zugzwang
position. I haven't quite convinced my teammates yet, but I believe
we're winning now.

Incidently, various friends who read the network have told me that
Valvo has said some unflattering things about us in his comments about
the games. I'm sorry to hear that, but I suppose I'd be angry too if
I'd just blown a game that I thought I was winning. I can't really
comment more until I see what Mike wrote firsthand, but my feeling
about these games is that the Valvo Team played very well in both
openings and then made a number of errors afterwards. We, on the other
hand, made some errors in our openings, but have played very tough
since then.

I don't think I need to apologize for our play in this match. We've
played about as well as we could have and the games have been
entertaining and instructive. I hope the network readers have been
enjoying them.

[B] I think this (37 Nc3) is a small inaccuracy. White could have forced
our a-pawn to a5 with 37 Bc8. It will probably go there eventually anyway,
though.

37 ... Kd8

[B] We expect 38 Ne4 now and we'll play 38 ... Kc7. This is an important
move to prevent the white knight from getting to d6. After that we'll
put our bishop in g1 and white will be reduced to pawn moves. We'll
eventually force him to move his knight from e4 with our bishop on g1
and we'll penetrate with our rook with Re3+. I think that at least one
of white's pawns will fall then.

38 Ne4

[V] Both sides plod towards the semi-Zugzwang position.

38 ... Kc7

[B] This is the best square for the king. It stops the knight from getting
to d6 and the bishop from going to c8 while staying close to the
e-pawn.

39 Nc3

[V] Black has nearly achieved his maximum potential. He wants to have White's
Knight on e4 and his Bishop on a7 or g1. When White plays Nc3 at that point,
then he will have Re3+ and Rg3. We will probably place a Pawn on a4, but
that is not relevant.

[B] We still can't play 39 ... Re3+ 40 Kd2 Rg3 because 41 Nd5+ Kb7 42 e7
wins.

39 ... Bc5

[B] Now if 40 Nd5 then we'll play 40 ... Kd8 followed by Re1-g1. We expect
40 Ne4 and we'll play 40 ... Bg1 (or maybe 40 ... Ba7). Then white
can't play Nc3 again because we can play Re3-g3. So white will have to
start using up his pawn moves, but he only has a limited supply of
those...

40 Ne4

[V] We wait for Black to make his move. An interesting continuation is
40 Ne4 Bg1 41 a4 Ba7 42 Nc3 Re3 43 Kd2 Rg3 44 Nd5 Kd8 45 Nb4! Rg2
46 Kd3 Bc5 47 Na6 Be7 48 c5.

If 40 Ne4 Bg1 41 a4 a5, then 42 Nc3 Re3+ 43 Kd2 Rg3 44 Nb5 Kd8 45 Bc6.

40 ... Ba7

[B] We're finally ready to play Re3+ if they move their knight again. Their
bishop and king have been immobile for some time so they're forced to
start making pawn moves.

41 a3

[V] Belle prefers this to an immediate a4. I don't see the difference
myself. I suppose she is pushing trouble over the horizon. She evaluates
the position as dead even.

[B] Game 1 continues to grind along. They've started using up their pawn
moves to avoid zugzwang and, from the looks of things, they're going to
try to go as slowly as possible. I don't think anything will happen for
a while, but they may get impatient and force the issue.

I guess they want to draw things out as long as possible. I don't think
the pawn can stop at a3. It's too vulnerable there.

41 ... Bb6

[B] Now 42 Nc3 Bc5! threatens the a-pawn and Re3+ so white won't have time
for Ne4 again. We expect white to play 42 a4 now and we'll reply with
42 ... Bg1. White will then have to start moving his kingside pawns or
give one (or both of them) up by playing 43 Nc3.

42 a4 Bg1
43 Nc3

[V] I guess this is the position we have all been waiting for since 32...Rh5.
It must be admitted that things doen't look too good for White and we have
to play the hand we were dealt. Black will play 43...Re3+ 44 Kd2 Rg3 45 Nd5
Kd8 46 Nb4 (46 Bc6!?) Rg2+ 47 Kd3 (47 Kd1!?).

[B] Just the move that we've spent the last six moves trying to prevent!
White allows the rook to penetrate. This was inevitable in the long
run, but white might have been better off pushing his kingside pawns as
far as they could go first. In that case we would have waited for the
pawns to come forward by shuttling the bishop back and forth.

We thought the main line was 43 h4 Ba7 (43 ... h5 44 Be8! is unclear,
and 44 ... g5 allows Nf6 with counterplay) 44 g4 Bg1 45 g5 hg5 46 hg5
Ba7 47 g6 Bg1 and black finally has achieved his zugzwang! Actually
it's more than zugzwang, because the black rook now has several
different avenues to enter white's position. It can infiltrate vi Re3+
or on the kingside via f5, g5, or h5.

In that variation black has numerous winning tries. He will win one
pawn almost immediately and then the rest of white's pawns will also
become weak as black advances his king. The Valvo Team must have decided
that they would get more counterchances by giving up a pawn immediately.

43 ... Re3+

[B] This wins a pawn right away after 44 Kd2 Rg3. The tricky 45 Bc6 Bc5!
46 Ke1 Rg6 47 Nf4 Rf6 doesn't quite work. White can win the a-pawn
for his g & h pawns, but then it will be very difficult for him push his
own pawns in time to counter black's connected passed kingside pawns.

The Valvo Team surprised us in the first game. They apparently have
given up hope of holding all their material and played a move which
allows us to win a pawn immediately. We considered this possibility,
but decided that it wouldn't give them enough counterplay to save the
game.

Ken Thompson is in Japan right now so the games are on hold. He'll be
back in about a week.

44 Kd2

[V] With 43 Nc3 we have decided to give up a Pawn, but not the one they must
think! We plan 44 Kd2 Rg3 45 Nd5+ Kd8 46 Bc6!? and the g2 Pawn is immune:
46...Rg2+? 47 Ke1! Bc5 48 e7+ winning. Best is 46 Bc6 Bc5! 47 Nc3 Bb4
48 Bf3 Ke7 49 Kd3. In return for the Pawn, the White pieces get very
active.

44 ... Rg3

[B] Attacking the g-pawn and moving the rook to a safe spot.

45 Nd5+

[V] We follow the steps of destiny. Black will play 45...Kd8 and we will
sacrifice the e-Pawn with 46 Bc6!? as mentioned in our previous posting.

[B] We expected this. White forces the king back and improves the knight's
position. Now 45 ... Kd6? allows 46 Bc8! and black can't stop the
e-pawn.

45 ... Kd8

[B] Things are moving right along in game 1. We're in some complications
now, but we expect things to resolve themselves in our favor.

White can win the a-pawn after 46 Nb4, but black will get the g & h
pawns. If 46 Bc6 (threatening e7+) then 46 ... Bc5 and black will pick
off one of white's pawns although white gets to decide which one.

46 Bc6

[B] This is a tricky defensive idea that saves the g-pawn. White is
threatening e6-e7 now so black can't take the g-pawn right away. 46 ...
Rxg2+ 47. Ke1 Bc5 48 e7+! Bxe7 49 Nxe7 discovering an attack on the
rook.

46 ... Bc5

[B] Threatening the g-pawn again. Now white can save it with 47 Ke1 then if
47 ... Rxg2? 48 e7+ as above, but black can win the e-pawn with 47 ...
Rg6 48 Nf4 Rf6.

47 Nc3

[V] There have been two complete moves since our last posting, but since
everything has been forced, and since this has all been predicted, there
seemed to be no need for a serarate posting. As mentioned previously,
we expect 47...Bb4 48 Bf3 Ke7 49 Kd3 Ke6 50 Ne4 Rg6 51 Kd4 with 52 c5 to
follow. White gets free play and activity in return for that yoke of a
e6-Pawn.

[B] Now white has saved his g-pawn, but at the cost of leaving the e-pawn
unguarded.

47 ... Bb4

[B] For a while we were considering 47 ... Re3. The idea was to hold off
from winning the pawn for while to keep our pieces active. We looked at
a lot of variations that looked pretty good for black such as 48 Ne4 Bd4
followed by ... Ra3. We noticed at the last minute, though, that white
can equalize with 48 Be4! which threatens 49 e7! Kxe7 50 Nd5+ forking
the king and rook.

47 ... Ke7 48 Bf3 Kxe6 49 Ne4 Bb4+ 50 Kd3 Rg6 transposes to the game.

48 Bf3 ...

[B] This is a great resource that saves the knight and deactivates black's
rook. Suddenly black's pieces aren't at all well placed. It would be
nice if we could win the e-pawn with 48 ... Rg6 but then white can play
49 Bd5 defending everything.

48 ... Ke7
49 Kd3

[V] Nothing new to add; everything was played as forecast. It is not likely
Black will play 49...Bc3:, so 49...Ke6 50 Ne4 Rg6 51 Kd4 is still on
the schedule. White's pieces cooperate very well, while Black's all
seem to be misplaced.

49 ... Kxe6

[B] We thought some about 49 ... Bxc3 50 Kxc3 Kxe6 to try to simplify, but
then white can draw easily by attacking and winning black's a-pawn.
Black's king doesn't do much good on e6. He has to get his king back to
the queen side for defense before he can trade off his bishop for
white's knight.

50 Ne4

[B] Attacking the rook and improving the knight's position. We were hoping
for 50 Nd5 which looks like a good move because it threatens the black
bishop and Nc7+ followed by Nxa6. But black has a good reply in 50 ...
Ba3! which threatens ... Kd6 blockading the white pawn on c4. Then on
51 Nc7+ Kd7 52 Nxa6 Rg6 the knight gets trapped.

50 ... Rg6
51 Kd4

[V] White has it in mind to play c5 and drive the Black King away with Bg4+.
It is not clear how Black can respond to this central White advance, but
we no longer think White is in any way worse off.

[B] White keeps making aggressive moves. In most endgames playing
aggressively is not only the best way to win, it's also the best way to
draw. White's only compensation for his material deficit is his
activity. It's important to exploit it as much as possible. If white
tries to set up some kind of passive blockade black will eventually
break it.

An amusing possibility now is 51 ... Kf5? 52 Bg4+ Kf4?? 53 g3 mate!

51 ... Kd7

[B] Scurrying back to the queen side. As soon as our king gets within
defensive range of our a-pawn we will happily trade our bishop for
white's knight.

Well we finally managed to win that pesky e-pawn in game 1. So now we
have a rook for a knight and a pawn. Unfortunately for us, though, we
had to displace our pieces some to get this. Now we have to do some
serious piece redeployment if we are going to win any more of their
pawns. I don't see any immediate drawing ideas for them, so I assume
that our material advantage should eventually tell.

I have to say that the last month or so has been bad for our team. A
month ago I thought we would get 1.5 points from this match and now I
think that we'll do well to get 1 point. So what went wrong? They've
played better than us! Part of the problem is that I've gotten a little
bored with the games and as a result we've been playing a little
superficially.

I have to give the Valvo Team credit. They've managed to keep the
pressure on for a long time. We started these games last April, and
except for a brief period over the summer, they've played good moves all
the way through. Perhaps that's the biggest advantage to having a
computer as a teammate. It never gets bored or distracted.

52 c5

[V] White will basically play his King to d5 and wait. If possible, White
will try to induce Black to play a6-a5 (e. g. Be2) so that Bb5 is possible,
but this is not essential. The White pieces are maximally placed and
Black's remain disorganized. 51...Kd7! was probably the best first step
towards reorganization.

[B] Pushing the pawn restricts black's movements and keeps the black rook
from moving to b6. If we got the chance we would have played ... Rb6,
... Be1 and ... Rb4 to win the a-pawn. If black gets a passed pawn
white won't have much chance to draw. The knight and bishop aren't
very good at setting up a blockade.

52 ... h5

[V] Their 52...h5?! is really a mystery to me. What's the point? Obviously,
we would not take it and allow the Rook to penetrate our position. But
what are they trying to accomplish? Prevent a Bishop check on g4? Reroute
their rook to h6 and h8? They could accomplish the latter with Rc6-c7.

[B] This move clears a square for the rook to move to (h6) and threatens to
gradually set up another bind starting with h5-h4. Despite our material
superiority it's not at all easy to decide whether we can win this
position. White's pieces are very well placed and his c-pawn is
dangerous. Simple attempts to reposition the rook don't work at all: 52
... Re6? 53 Bg4. 52 ... Rc6 53 Kd5 Rc7 (53 ... Rc8? 54 Bg4+) 54 Bg4+
Kd8 55 c6 Re7 56 Nc5 and white is doing fine. 52 ... a5 53 Nc3!
(threatening c6 and Nb5) Bxc3 54 Kxc3 followed by Kc4-b5 is also good
for white.

Now white has several ways to try to stop black from playing h5-h4, but
each is risky: 53 Bxh5 Rxg2 and black's rook is suddenly very active.
53 g4 hg4 and then both 54 Bxg4 or 54 hg4 leave white's position on
the kingside a little loose. The Bf3 Pg2 Ph3 configuration is very safe
for white so he should be relunctant to part with it. 53 g3 threatening
Bxh5 is possible. Then black can try 53 ... Rh6 with the idea of ...
Be1 and then repositioning the rook or he can try 53 ... h4!?. Then if
54 gh4 Rg1 activates the rook. If 54 Bg4+? black can get a winning B
vs N endgame with 54 ... Rxg4! 55 hg4 h3! 56 Nf2 h2 57 Kd5 Be1 58 Nh1
g5.

53 Nf2

[V] I think they are getting desperate in their obcession to play for a win.
We could just sit on the position with 53 Kd5 or 53 Kc4. Or try the tricky
53 Nf2!? threatening 54 Nd3 hitting the Bishop followed by 55 Ne5 forking
the King and Rook or 55 Nf4 forking the Rook and Pawn. The computer (both
selective search version and brute force version) very slightly prefers
53 g3 and expects 53...Rh6. I, however, feared 53 g3 h4! 54 gh Rg1 and
Black has penetrated the White position with his Rook. Could this latter
variation and knowledge that the computer was involved been their motivation
for 52...h5?

After 53 Nf2, I think 53...Bd2 is best, but the computer prefers 53...Ba5
or 53...Ba3. Then (53 Nf2 Bd2) comes 54 Nd3 Kc7 55 Ke5 when White plans to
annoy Black on the King's side. The point of repositioning the Knight to d3
is to protect the c5 Pawn so the King is free to wander.

[B] We didn't expect this move since the knight was already on its best
square at e4. But it's not a bad move since it sets up some tricks and
white doesn't have much to do anyway.

The first point is that 53 ... h4? loses to 54 Nd3 threatening the
bishop and a fork on e5. If black tries to take advantage of the
knight displacement by playing 53 ... Rg5?! white can reply 54 Nd3 Bd2
55 h4! Rf5 56 g4!! Then black can't take the bishop because of Ne5+
and Nxf3. He also can't take the pawn with hg4 because Bxg4 pins the
rook. So black has to retreat his rook and allow white to either take
the h-pawn or play g4-g5. Either way their position would look pretty
good.

53 ... Rh6 54 Ne4 Rh8 was possible, but it seems better to leave the
rook on g6 until we find a better square for our bishop. So we decided
to go ahead with our slow repositioning plans.

53 ... Kc7

[B] The king moves away from the fork square. Now black is ready to
activate his rook if white lets him via Rg5-f5. We were hoping that
white would misplace his knight now by playing 54 Nd3 Bd2 55 Ne5 Rg5
when we threaten Bc3+ forcing a won rook verses bishop and pawn endgame.

54 Ne4

[V] It still bothers me why they played 52...h5. Recently I wondered if they
intend to play Rh6 and g5. But they could never force the Pawn to g4
because the Rook couldn't both protect the h5 Pawn and provide support
along the g-file.

We have decided to sit and see what they have in mind. If 54 Nd3 Bd2 and
White has no adequate defense to the Black idea of Rf6-f4-a4 except to
play Nf2-e4! So why bother?

[B] It's clear that white was just marking time with Nf2. The knight
restricts the black rook, king, and bishop on e4.

[B] The Bono - Valvo games have been going for more than 8 months now.
Since there are probably at least a few readers who weren't around when
these games started, this is what's been going on in the Bono - Valvo
match:

Last April I challenged IM Mike Valvo to an EMAIL consultation
chess match. I selected a consultation team and offered Mike
the opportunity to do likewise. In addition to myself, my team
included Bill Kelleher, Larry Tapper, and Adrian Casillas -- all
masters in the Cambridge area. Mike accepted with a team
consisting of himself, Ken Thompson, and the Belle Chess
Computer. We agreed to play two games so both teams could get
to be white and black.

Since April we've been playing a move in each game every two or
three days, but we've had to suspend play from time to time
while people have been on vacation or away. Periodically Mike
and I have been posting messages to rec.games.chess with
comments on how we think the games are going. We've agreed not
to read each other's postings until after the games are done.
Mike's postings all have "Valvo - Bono Match" in the subject
line and mine all have "Bono - Valvo Match".

In game 1 as black we accidently played into one of Mike's pet
lines in the Najdorf Sicilian defense. Mike sacrificed a piece
on move 13 and we barely survived the next fifteen moves.
Larry Tapper spent some time talking to Mike at the World Open
and found out that Mike has known about this piece sacrifice
for years and that we did well to get any kind of playable
position out of that opening. Around move 30 they got a little
carried away trying to win a drawn position and allowed us to get
an edge in the endgame. They offered a draw on move 32 which we
refused. Since then we've been trying to grind out a win. We
won a pawn on move 49, but they've been defending very stubbornly
so it's still not clear what will happen.

We're still hoping to get at least a point out of this match.
Both games have been extremely tough battles and it's still not
clear (at least to us anyway!) how they'll will turn out.

54 ... h4

[B] This move restrains any white expansion on the kingside and removes the
pawn from danger. The disadvantage to h5-h4 is that it gives the white
bishop the g4 square. White can now mark time with Bf3-g4-f3 as well as
Kd5-c4-d5. Our plan is to move our bishop and king to their best
squares and then activate our rook over the next few moves.

55 Kc4

[V] Just biding our time.

[B] White is apparently just waiting to see what we're up to. We wanted to
reposition the bishop anyway.

55 ... Be1

[B] We're steering the bishop toward f4 via g3. It could also have gotten
there via Ba3-c1-f4, but that would have given white a little more
flexibility since he could have tried Nc3 with the idea of Nd5+. Now
if white plays 56 Nc3 then 56 ... Bxc3 57 Kxc3 Rf6 is a winning endgame
for black.

White has to be careful here. Black is threatening 56 ... Re6
followed by ... Re8 and then by ... Rd8 or ... Rb8.

56 Kd5

[V] All our pieces are on optimum squares, so we continue to wait. It is
true that our a4 weakness is more significant than their a6 and g7
weaknesses, but it remains to be seen how they can approach it. At some
point they will have to try either Rc6 or Rh6 (the reason behind h6!?)
and allow White additional counterplay because the Bishop is no longer
tied to guarding the g2-square.

[B] We're busily repositioning our pieces in game 1. The Valvo Team is
just playing waiting moves now so it looks like the question is
whether we can really escape from the bind that they've set up. At
this point I believe that we can and that we will soon be able to start
making real threats.

White can either leave his bishop on g4 and move his king back and
forth between c4 and d5 or he can leave his king on d5 and move his
bishop back and forth between f3 and g4. I don't think it matters much
which he chooses.

56 ... Bg3

[B] We weren't quite ready to move our rook yet. 56 ... Rc6 is useless
because the king and rook get tangled up. 56 ... a5 with the idea of
... Ra6 gives white the b5 square and it weakens the a-pawn. 56 ...
Rh6 allows 57 Ng5 followed by Ne6 with some annoying threats.

We intend to play ... Rh6 as soon as we can. First we'll play ... Bf4
to keep the knight out of g5, and only then ... Rh6, ... Rh8, etc.
White doesn't seem to be able to stop this, but he may be able to make
some counterthreats by moving his knight to d5.

57 Bh5

[B] White continues to do nothing. The bishop can't afford to move too far
away from the g-pawn so white isn't really threatening anything (except
our rook that is).

With our last move we invited white to play Nxg3. The resulting
rook verses bishop endgame would have been easily won for black after
57 ... hg6. Black's idea would be to get the rook to the seventh or
eighth rank and then harass the white pawns on both sides of the
board. White only gets drawing chances in the bishop verses rook
endgame if he can somehow get rid of black's a-pawn.

57 ... Rh6

[B] This move only incidently threatens white's bishop. We want to
reposition the rook via our first rank with Rh8 and then Rd8, Re8, or
Rf8. We're still not worried about 58 Nxg3 since we can simply
recapture the knight and then white can't stop black's rook from
entering his position.

58 Bf3

[V] Again we wait. Interesting is 58 Be2!? Bf4 59 Kd4 Re6 60 Bc4 Rg6
61 Bf1 Be3. Or 58 Be2 Bf4 59 Bg4 Rh8 60 c6!? Rh6! (60...Rd8 61 Bd7)
61 Nc5 Rd6+ 62 Ke4 Kc6! with complications. We can still transpose
into this latter line.

Belle likes 58 Bf3 Bf4 59 Nc3! Rh8 60 Kc4 (60 c6!? Rd8 61 Kc4) and
White seems to threaten 61 Nd5+ followed by Nb4 winning a Pawn.

[B] We are almost done repositioning are pieces in game 1. Things are
still looking very good for us and now we are in a position to choose
which way we want to try to win. Game 2 on the other hand, is looking
at least as bad as game 1 is good. The Valvo Team seems to have many
ways to win now.

[B] 58 Be2 temporarily freezing the rook along the third rank would have
cut down our options. In that case we would probably have aimed for
the game position with 58 ... Bf4 59 Kd4 (what else?) Re6 60 Kd5 (60
Bd3 or 60 Bc4 allows 60 ... Rg6 61 Bf1 Kc6! 62 Bxa6 Rxg2 and black's
kingside pawns are much faster than white's queenside pawns.) Rg6 61
Bf3 Rh6.

58 ... Bf4

[B] 58 ... Rh8 also looked good. Then 59 Nxg3 is still bad and black is
already threatening to enter the game with his rook. White would have
to try to mix things up with 59 Ng5 Rd8+ 60 Kc4 (60 Ke6 Re8+ 61 Kf5
Re5+ 62 Kg6 Bf4! 63 Ne4 Bh6 and white's king is way out of position.)
Re8 61 Bd5 Re2! 62 Ne6+ Kb8 and now if white tries to win the g-pawn
with 63 Nxg7 then 63 ... Rc2+ wins the c-pawn in return. White can put
up some resistance with 63 c6 but then black can play to win the
a-pawn.

We decided against this line because white's knight gets very active
and it looks like we can get something similar while keeping the knight
out of e6. We now expect white to play 59 Be2. This keeps black from
moving his rook off the third rank for a move, but after 59 ... Be3
white is in a mini-zugzwang: every move concedes something to black.

59 Nc3

[V] This is basically a waiting move, but it packs venom. White intends
60 Kc4 and 61 Nd5+ followed by Nb4. This essentially prevents Black
from playing 59...Rh8 (though it it not at all clear that it would be
fatal if Black could safely play Rh8). Belle thinks 59...Be3 is best,
but this looks like Black must make some move and this is merely the
best of the lot (according to her). Generally, Belle sees no particular
danger in the position for White. She understands that a Rook
penetration could potentially be dangerous and she learned this when
considering 58 Ng3 winning the minor exchange (Bishop for Knight) and
doubling Black's Pawns.

[B] Tricky. We were expecting 59 Be2 and then after 59 ... Be3 White would
be short of moves. After their move 59 ... Rh8 doesn't quite work --
but it's close. We found a very long variation which just barely draws
for them: 60 Kc4 Bd2 61 Nd5+ Kb8 62 c6 (white has to go forward; 62 Nb4
Bxb4 loses) 62 ... Ba5 63 Kc5 Re8! (preventing Be2) 64 Kd6 Bc7+! 65 Kd7
(65 Nxc7 Rd8+ leads to another B vs R endgame that white can't hold) 65
... Rd8+ 66 Ke7 a5! (else Nb4) 67 Nc3 Rd4 68 Nb5! Now if 68 ... Rxa4 69
Kd7 and white can win the bishop for the c-pawn and then just barely
get back in time to stop black's a-pawn. If 68 ... Rd2 white can draw
by shuttling his king between e8 and e7. Black can complicate things a
little by playing Bf4 and then Pg5-g4 but white can still draw by
playing Bxg4 and then Be6.

59 ... Bd2

[B] I made a psychological blunder when I sent this move to them. I sent
along an if move with it: If 60 Ne4 then 60 ... Rh8. The idea was
that if white then played 61 Nxd2 black could play 61 ... Rd8+ and 62
... Rxd2 which would be another winning R vs B endgame for black.
Even if white didn't play Nxd2 then ... Rd8 would be very strong for
black. This was all well and good, but 60 Ne4 wasn't forced!

They must have wondered what was going through our heads. If we were
hoping that they would make the error 60 Ne4? then why point out the
trap to them? I guess the answer is impatience on my part. My
teammates were none too happy about it, but I just wanted to get the
game over with! I wonder if The Valvo Team would have fallen for the
trap if I hadn't pointed it out to them....

60 Ne2

[V] Black's 59...Bd2 was accompanied with "if 60 Ne4, then 60...Rh8." 60 Ne2
was Belle's original intention. White wants to get his Knight to e6 if
possible (61 Nd4, 62 Ne6) or d5 (60...Be3 61 Nc3 Rh8 62 Kc4). Essentially
we are still waiting while not allowing Black to get his balance.

Interesting is 60 Ne4 Rh8 (we had to look at it because of the if move)
61 c6!? Rd8 62 Kc4 Be3 63 Nc5, but 63...Bc5 will probably win because
White's pieces has lost their mobility and the Black King can lend its
influence to the position.

[B] Now white is threatening Nd4-e6+. Black can't allow this with
his bishop on d2.

60 ... Be3

[B] Keeping the knight out of d4 since as usual white can't afford a trade.

61 Nc3

[V] Black has maximized the position of his pieces except for his Rook. At this
point he can try either Rh8 or Rf6. 61...Rh8 62 Kc4! Bf2 63 Nd5+ Kb8 64 Nb4!
Ka7 65 Nc6+ is an immediate draw through perpertual non progress. If 61...Rf6,
62 Kc4 and the threat of Nd5+ will pick up something (62...Rf4+ 63 Kd3! is no
help).

Belle preferred 61 Kc4!? instead of 61 Nc3. This would prevent a third Black
alternative 61 Nc3 Rc6. However, 61 Nc3 Rc6 62 Ne4 and the Rook remains
inmobalized (eg 62...Kb7 63 Nd6+ Kc7 Ne8+ with a perpetual check because
the Black King must continue to protect the Rook).

Belle has had limited value because she sees no danger from anything Black
can do. I noticed this as a problem with the perfect endgame database Ken
Thompson has developed: When the computer is within the set of moves that
lead to a draw, it will not try to win by giving its opponent the most
difficult problem. It sees that everything leads to a draw and cannot
distinguish between two different lines (there have been attempts like
choosing the longest line to draw [it knows the complete set of possibilities]
but nothing has been effective). When holding on the problem appears even
more paradoxical because the computer will allow its opponent to achieve
the maximum position with no resistance, then it will walk a tight wire of
forced moves to draw.

That's the problem here: Belle is uninterested because she sees no danger!

[B] Forced. If 61 Kc4 Re6 62 Nc3 Bd2 63 Nd5+ Kb8 white can't stop the rook
from penetrating.

After 61 Nc3 black would like to improve the position of his rook, but
the white pieces are now on their ideal squares and black has to worry
about possible knight forks. 61 ... Rf6 fails to 62 Ke5! threatening
Nd5+ forking the king, rook, and bishop and black just can't get all of
his pieces out of the way in time. 61 ... Rh8 leads to the variation
that black could have gotten after 58 ... Rh8 in which white appears
to draw.

[B] Things are proceeding slowly in Game 1. The Valvo Team found a defense
that we didn't see so one of our ideas doesn't seem to work. We're not
too worried, though. There are still many tricks for us hiding in the
position. I'm not sure we can win given perfect defense, but it makes
sense for us to try to squeeze whatever we can out of the position.

61 ... Kd7

[B] This is a finesse. The idea is to get a position similar to the one we
have now, but with white's pieces on less then ideal squares -- say
with white's bishop on g4 and/or his knight on e4. We are threatening
... Rf6 now.

62 Kc4

[V] Believe it or not, White threatens 63 c6+ Kd6 64 c7! winning. Black can
prevent this with 62...Bf2 or 62...Bd2. Black is looking for trouble if
he tries 62...Rh8: 63 Nd5!? (63 c6+?! Kd6 64 Ne4 Bd1! 65 c7 Ba5-+) Bd1
66 Bg4+ Kc6 67 Nb4+. Perhaps even better is 62...Rh8 63 Bb7!

White basically is still waiting. 62...Bd2 63 Ne4 Be1 64 Kd5 is roughly
the same position we have been looking at.

[B] This sets a little trap for us. White threatens 63 c6+ Kd8 64 c7+! and
now black can't take the c-pawn because of Nd5+ winning the bishop and
if 64 ... Kc8 then 65 Bg4+ forces 65 ... Kxc7.

62 ... Kd8

[B] We would much prefer to activate our rook by simply moving it to f6 or
e6 rather than to h8 then d8. So we've let white make some progress in
return for allowing our rook access to these squares. A white formation
with the knight on d5 looks threatening, but if necessary we can stop
the c-pawn by blockading c7 with our king and bishop.

63 Ne4

[V] Black's last (62...Kd8) impressed me quite a bit. It falls into the
category of a "classy move." The purpose is to clear the 7th rank for
the rook. Black realizes that progress can not be made by trying Rh8,
so they have prepared for either Re6-e7 or Rf6-f7. This is basically
the remaining attempt for them to win.

We felt that (a) 63 Bb7 (b) 63 Nd5 Bd2 64 Kd3 and (c) 63 Ne4 were our
choices in the position. Belle saw little difference between them. I
felt that 63 Ne4 was the most solid and flexible.

The main attempt seems to be 63...Re6 64 Kd5 Re7 65 Kc6! when White
threatens to annoy the a6 Pawn. Probably best then is 63...Kc8 when
White can oscillate back with 64 Kd6 Rd7+ 65 Kc6 Rd4?! 66 Nd6+ Kb8
67 Nf5 Rd3 68 Ne3 Re3 69 Kb6.

[B] We can play 63 ... Re6 now, but unfortunately our king still doesn't
seem to be on the right square. White can reply 64 Kd5 Re7 65 Kc6.
This forces us to defend against white's king invasion with 65 ... Kc8.
If, however, in this variation our king was on b8 to start with we could
reply to 65 Kc6 with 65 ... Rb7 threatening to win white's a-pawn. So
we decided to move our king to b8 before moving the rook.

63 ... Kc7

[B] Despite the fact that we haven't made any real progress in the last few
moves they still have to be careful. For example, the obvious 64 Kd5
allows 64 ... Rh8 followed by ... Rd8 and we've activated our rook.

64 Nc3

[V] So, this was a triangulation of a sort after all! I had an intuition
they were doing that, but I didn't see to what purpose. Now they can
try 64...Re6!? and things get interesting: (a) 65 Bg4!? (the purpose
is to drive the rook off the sixth rank so Pawn to c6 becomes
possible) Re5 66 Nd5+ Kd8! 67 c6?! Re4+ 68 Kd3 Rd4+ or 68 Kc3 Bf4-/+
or (b) 65 Nd5+ Kd8 66 Nb4 Bd2 67 Nc6+ (67 Kd5!? Rf6 68 Nc6+ Kc7 69 Nd4)
Kc7 68 Nd4 and White can start dreaming of a win (except Black could
always give back the exchange).

[B] They continue to wait.

[B] I think we're finally making some more progress in the first game. We
are now in a position to activate our rook.

***

I met Mike Valvo for the first time a few weeks ago. He was in
Cambridge to comment on the Karpov - Deep Thought match. Bill Kelleher
and Larry Tapper were also there. We talked a little bit about the
Bono - Valvo match. I think a couple of things he said are worth
repeating.

Mike said that it was very difficult teaming up with the computer
because the computer played by concrete analysis while he played by
intuitition most of the time. His moves "looked better" than the
computer's but he couldn't always say why. Mike said that Ken Thompson
was really enjoying the exercise, though, and that Ken found it
very helpful to be getting so much feedback on how Belle was doing.

This much was very interesting to me. I would have thought that having
the computer around to verify that good looking moves weren't
tactically flawed would be a tremendous advantage. Another advantage of
having the computer as a partner is that it isn't at all biased. Our
team made several errors because we convinced ourselves that our moves
were good without any real tactical verification.

Mike also said that most of the good moves in the Bono - Valvo match
were his and most of the bad moves were the computer's. He said that
we might have noticed that sometimes their team looked aimless. That
was because he would sometimes trust the computer. Other times he
would take over the games himself and then the team would make real
progress. Mike also said that he'd wanted to resign both games a while
back because of some undisclosed faux pas that we made, but that Ken
Thompson had talked him out of it.

Wow! I don't know how much of all that is true, but it sounds to me
like Mike has quite an ego. Although I have a great deal of respect for
his play, I don't think I'd ever want to play another "friendly game"
against Mike Valvo.

***

64 ... Kb8

We're expecting 65 Ne4 Re6 Kd5 Re7 and it's not entirely clear who's
making progress, but things are happening and black has some real
threats.

65 Nd5

[V] They have removed their King from the field of action, so we decided to
get frisky. If allowed, we will play Nb4 next. Otherwise we have thoughts
of Ne7.

[B] If white continued to repeat moves with 65 Ne4 we decided that 65 ...
Rh8 was best. Then if 66 Kd5 we could reply 66 ... Kc7 threatening ...
Rd8+ followed by a rook infiltration. It appears that white would then
be helpless to prevent it: 66 Nd6 allows 66 ... Rd8 followed by ...
Bxc5; 66 Be2 Rd8+ 67 Ke5 Kc6! threatens both the c-pawn and 68 ... Re8+
69 Kf5 g6+! winning the knight.

65 Kd5 would probably have led to the same variations as above after 65
... Rd8 since 66 Kc6 fails to 66 ... Rc8+, so white would probably
have to try 66 Ne4 Kc7 transposing.

65 ... Bd2

[B] White can't retreat the knight now since the rook verses bishop endgame
is winning for black after BxN. So white has to go forward. Black is
threatening 66 ... Re6 followed by ... Re1.

66 Bg4

[V] The computer prefers this to the unclear 66 Ne7. White "threatens to
threaten" 67 Bd7 and c6-c7-c8. This enables White to force the Bishop
to a5, if desired. White could then effectively shut the Black King out
except for the weakness at g2 (Rg6!). However, the White King can move
around now that the Black Bishop is not putting pressure on the White c5
Pawn. Black's winning chances do not seem to have improved.

It must be admitted that the computer thinks we have lost 5/100's of a
Pawn during the last few moves. I don't think that is significant, but
Ken is slightly worried.

[B] Peventing an immediate rook penetration by keeping the rook off of e6.
This is white's only chance. 66 Ne7 Rf6 is very strong for black. After
67 Nc6+ Kc7 Nd4 Be3! white is in full scale retreat.

66 ... Rg6

[B] I love this move! Black freezes white's bishop on g4. Now any bishop
or knight move allows the rook to find an open file. If white tries to
hold his blockade now by shuffling his king back and forth with 66 Kd4
Be1 67 Kc4 black's rook will soon enter their position with 67 ... Rg5
threatening ... Re5. White can play 68 c6 but he reaches a dead end
after 68 ... Bg3 69 Kc5 Bc7.

67 Ne7

[B] They obviously decided that they had to try something now. At least
they get some counterplay after this move.

67 ... Rf6

[B] At last! An open file! We're threatening ... Rf4+ and ... Rf2 now.
White can try 68 Kd5 or 68 Nf5. We consider the latter of these to be
the main line since it keeps our rook out for at least another move,
although our position appears to be close to winning after 68 ... Be1!

[B] Things are finally happening again in the first game. The Valvo Team
has abandoned their blockade and they've formed a new one. This one
isn't as effective as the other one, though, since they can't keep our
rook out of their position for long. It looks like the game will soon
turn into a race of sorts in which they try to make counterplay with
their c-pawn while we try to win either their a or g-pawns.

68 Nf5

[V] I talked Ken into playing this risky line in favor of 68 Kd3 and
continuing to shuffle around. Belle is now afraid of 68...g5, but
I assured Ken humans were unlikely to play it; especially when
the dynamic looking 68...Be1 is possible. Belle fears we will be
run out of moves after 68...g5 69 Kd5 Kc7, but I feel that 70 Nd4!?
Rf2 71 Bf3 will give us enough dynamic play even if the a-Pawn goes
away. After 68...Be1 70 Ng7 Rf4+ 71 Kb3 (71 Kd5!?), things get very
unclear. But interesting.

[B] Game 1 has become a melee. The Valvo Team has temporarily won a pawn,
but their knight is now out of play. I believe the superior power of
our rook will now decide the game, but everything is still very
complicated.

The alternative was 68 Kd5 but after 68 ... Rf2 69 Bf3 (69 Kc6?! Rxg2
threatens 70 ... Rxg4! 71 hg4 h3 winning) Bg5 70 Nf5 g6 71 Nd4 Rd2 black
is probably winning.

68 ... Be1

[B] We offer white the g-pawn. This move improves the bishop's position,
threatens to dislodge the knight from f5 with ... g6, and guards the
h-pawn. The g-pawn isn't particularly important right now and white has
to misplace his knight to take it.

68 ... g5 69 Kd5 is much less clear since white's pieces are already all
on good squares. Black won't necessarily be winning even if he gets his
rook to the seventh rank. For example the straightforward attempt to win
with 69 ... Kc7 70 Nd4 Rf2 71 Ne6+ Kb8 72 Bf3 Bf4 73 Kc6! Ra2 74 Bd5
Rxa4 75 Kd7! Ra1 Nd8! gives white at least a draw.

69 Nxg7 ...

[V] Belle is happier now than she has been for a while, but that is to be
expected from a materialist computer. Black can try either 69...Rf2 or
69...Rf4+. Ken and I have another informal bet as to what they will
play: the machine expects 69...Rf2, but I feel they will play 69...Rf4+.
In either case, we will get stretched out, but it seems things hold
together.

[B] We were a little surprised that they took the pawn. We expected 69 Kd5
g6 70 Nd4. Then if 70 ... Rf2 71 Nf3! Bd2 (71 ... Re2 72 Kc6!) 72 Nxh4
g5 73 Nf5 Rxg2 74 Kc6 gives white some drawing chances. We intended to
play 70 ... Rf4 threatening ... Bf2. Black gets good winning chances
after 71 Nc6+ Kc7 72 Nd4 Rf2 73 Ne6+ (73 Nf3 Re2!) Kb8 74 Bf3 Ra2 75
Kc6 Rxa4.

69 ... Rf4+

[B] Now white can keep his king active and return the pawn immediately with
70 Kd5 Rxa4, but then black's a-pawn will be very difficult to stop, or
white can try to defend the a-pawn with Kb3, but this is very passive
and black can quickly win one of white's pawns.

70 Kd5

[V] Both Belle and I prefered to play this way, so we agreed to do it.
Black is welcome to our weak a-Pawn and we pin our hopes that Black
won't be able to queen it. We aren't working so much with concrete
variations as typical lines where Black has difficulty in advancing
while we keep the Black King and Bishop busy with our advanced Pawn.

[B] Giving up the a-pawn right away. This is very bad for white since
rooks tend to be good at escorting passed pawns in, but there really
wasn't much choice. 70 Kb3 holds temporarily but it leaves white's king
trapped in the corner and black wins a pawn soon anyway: 70 ... a5! 71
Ne6 (What else? 71 Bd7? Rf7; 71 Nf5 Rb4+ Ka3 Rc4; or 71 Bf3 Rb4+ 72
Ka3 Bd2 73 Bc6 Bc1+ 74 Ka2 Rb2+ 75 Ka1 Rc2 winning the c-pawn.) Rb4+ 72
Ka3 Bd2 73 c6 Bc1+ 74 Ka2 Rxa4+ Kb3 Rb4+ Kc3 Bf4 and white can't do
much with his c-pawn without help from his king.

70 ... Ra4

[B] We didn't need to be asked twice. This is what we've been aiming at
for the last 20 moves. Now black has a passed pawn of his own, his
h-pawn restricts white's kingside pawn majoritiy, and his king and
bishop can easily stop white's c-pawn. Black is winning.

71 Ne6

[V] Belle says the Black a-Pawn runs too fast and we are lost. That may be true,
but I'm not convinced. My plan is to put the c-Pawn on c6 and run back with
the King to harass the Black a-Pawn. I assume the Black Bishop will be kept
busy guarding the c7 square, so it can't play a complete supporting role
in ushering the Pawn. If I'm wrong about this, then I was wrong about 67 Ne7
because that's where I made this determination.

[B] The knight rushes back to get into the action. This offers a little
more hope than Nf5 since the knight on e6 can help threaten Pc7
in some variations.

71 ... Rb4

[B] Improving the position of the rook and clearing the way for the
a-pawn. The a-pawn, which has been languishing on a6 since move 5,
suddenly has a bright future ahead of it. White will now have to do
some fancy footwork to stop black from simply pushing it through to the
eighth rank. Here's a variation that illustrates how quickly the pawn
can become dangerous: 72 Kc6!? a5 73 Kd7 a4 74 c6 Bg3 75 Bf3 (75 c7+
Bxc7 76 Nc7 Rb7) a3! 76 Bd5 (else the a-pawn queens) Rb2 (or 76 ...
Ra4) 77 c7+ Bxc7 78 Nxc7 Rd2! followed by Rxd5! and the a-pawn queens.

Things are looking very good for Bono Team right now. The Valvo Team
offered us a draw in game 2 and game 1 appears to be won for us now. I
just hope we can avoid the kind of complications in game 1 that cost
them the win in game 2.

72 Resigns

[V] I can see it now. The a-Pawn is unstoppable, So we resigned.

[B] The Bono - Valvo Match is over. The Valvo Team resigned game 1. We
were a little surprised that they gave up so soon, but their position
was quite lost. The combination of our rook and passed pawn was too
much for their minor pieces.

These must be among the most annotated games ever played. Over the
last year Mike Valvo and I both regularly posted our own comments and
analysis of these games to this news group. In total the two of us
posted more than one hundred pages of commentary about the two games.
We agreed not to read each other's postings until the end of the games
so I am just now going through IM Valvo's postings. I will post some
general comments about the games when I'm done.
John Richards