33rd Kio Match Game 3

[Black "Habu Yoshiharu, Challenger"]
[White "Sato Yasumitsu, Kio"]
[Event "33rd Kio-sen, Game 3"]
[Date "March 8th 2008"]
1.P7g-7f     00:00:00  00:00:00
2.P3c-3d     00:00:00  00:00:00
3.P2g-2f     00:03:00  00:00:00
4.P5c-5d     00:03:00  00:02:00
5.P2f-2e     00:07:00  00:02:00
6.R8b-5b     00:07:00  00:02:00
7.G4i-5h     00:08:00  00:02:00
8.P5d-5e     00:08:00  00:05:00
9.P2e-2d     00:09:00  00:05:00

This leads to the sharpest variation of the Gokigen Nakabisha. Neither side takes 
the time to castle and goes for an all-out attack. According to Ouchi (9-dan), it 
is doubtful if this is the right idea for black. After all, it is the Furibisha 
side who initiates the fight and trying to trump this with an even sharper attack 
could easily be an overplay. It will take a while before a final conclusion is 
reached, but this game is definitely no evidence that 9.P2d is bad for black.

10.P2cx2d    00:09:00  00:05:00
11.R2hx2d    00:09:00  00:05:00
12.P5e-5f    00:09:00  00:05:00
13.P5gx5f    00:11:00  00:05:00
14.B2bx8h+   00:11:00  00:05:00
15.S7ix8h    00:11:00  00:05:00
16.B*3c      00:11:00  00:05:00
17.R2dx2a+   00:12:00  00:05:00
18.B3cx8h+   00:12:00  00:05:00
19.N*5e      00:12:00  00:05:00
20.K5a-6b    00:12:00  00:05:00
21.+R2ax1a   00:13:00  00:05:00
22.P*5d      00:13:00  00:21:00

Diagram 1

Here is Sato's "brilliant blunder" again (a term coined by Yonenaga). At first 
glance, this is a terrible move, because after 23.Nx6c+ Kx6c the move L*6f seems 
perfect for black. Not only has white no pawns to defend against this check, but 
the lance on 6f also blocks the attack of the promoted bishop on the rook on 1a. 
Therefore, this would normally be very good for black, but the results by black 
were actually very good. That is, until Habu introduced a brilliant new move in 
the second game of the Osho match against Kubo earlier this year.

23.N5ex6c+   00:18:00  00:21:00
24.K6bx6c    00:18:00  00:21:00
25.B*9f!     00:18:00  00:21:00

Diagram 2

Until further study shows otherwise, this move will be known as "Habu's famous 
bishop". No one ever considered dropping a bishop here and it is another example 
of Habu's ability to think out of the box.

26.K6c-5c    00:18:00  00:46:00

Kubo took almost two hours before playing 26.P7d and lost without a chance after 
27.L*6f +Bx6f Px6f S*9e +Rx1c. Sato tries 26.K5c instead, although in the post-mortem 
analysis of the Osho game, Kubo indicated that he had looked at it, but that he felt 
that there was no way to win for white. Interestingly, Sato said after the game that 
he didn't know about this comment and that "if Kubo said so, probably it is true that 
K5c is not good". A smokescreen to give the impression that he will never play his 
brilliant blunder move again?

27.L*6f      00:32:00  00:46:00
28.N*7d      00:32:00  00:56:00
29.P*2b      00:35:00  00:56:00

A strong move that both blocks the diagonal of the bishop and threatens to make a 
tokin on 2a.

30.N7dx6f    00:35:00  01:35:00
31.P6gx6f    00:36:00  01:35:00
32.P7c-7d    00:36:00  01:42:00
33.B9fx7d    00:40:00  01:42:00
34.+B8hx6f   00:40:00  02:04:00
35.N*6e      01:26:00  02:04:00

Diagram 3

36.K5c-6b    01:26:00  02:20:00

A big surprise to the professionals in the press room. This looks bad because of 
the obvious 37.P*6d, so everybody expected Sato to play 36.K4b here. However, Sato 
correctly assessed that after 37.P2a+ +Bx1a +Px1a the promoted pawn on 4b is 
dangerously close to white king. Black has given up the rook, but the black position 
at the moment is quite well-protected against a rook drop and it takes time to set 
up a strong attack against the black king. Sato therefore saw nothing else but to 
play 36.K6b. He actually had an interesting additional comment that gives an insight 
in the professional's mind in this kind of position: "In this type of position the 
winning pattern for white is to attack the king by taking the silver on 3i with the 
promoted bishop. There is no way to win if this promoted bishop is exchanged against 
the promoted rook on 1a".

37.P*6d      01:45:00  02:20:00
38.S7a-7b    01:45:00  02:22:00
39.+R1ax3a   01:46:00  02:22:00
40.G4ax3a    01:46:00  02:22:00
41.S*6c      01:46:00  02:22:00

Diagram 4

42.K6b-7a?   01:46:00  03:12:00

Another surprise, but this is probably based on a miscalculation. The natural move 
is of course 42.K5a. After 43.Sx5b+ Gx5b Bx5b+ Kx5b G*5c K4a G4c black is threatening 
mate after N5c=. White can defend with S*5b, but then R*7a S*6a N5c= K5a Nx6a+ S7bx6a 
S*5g and both players agreed that this looks good for black. However, this variation 
is not so clear as what happened in the actual game, so Sato should have tried 42.K5a 
instead of 42.K7a. The fact that he took 50 minutes for this move shows that he was 
not sure what was best here.

43.S6cx5b=   01:46:00  03:12:00
44.L*6g      01:46:00  03:19:00
45.G5hx6g    02:12:00  03:19:00
46.+B6fx6g   02:12:00  03:19:00
47.S5bx6a+   02:12:00  03:19:00
48.S7bx6a    02:12:00  03:19:00
49.G*5h      02:12:00  03:19:00
50.+B6gx5h   02:12:00  03:32:00

A move that comes to mind is 50.R*5g, but in this position it is not even a mating 
threat, so black easily wins after 51.N5c=.

51.K5ix5h    02:13:00  03:32:00
52.R*8h      02:13:00  03:57:00
53.L*6h      02:36:00  03:57:00
54.P*6g      02:36:00  03:59:00
55.B*5c      02:43:00  03:59:00

Diagram 5

56.K7a-8b    02:43:00  03:59:00

Sato didn't explain after the game why he selected 42.K7a over 42.K5a, but Ouchi 
expressed an interesting opinion as to what went wrong. Sato probably thought that 
he could defend with 56.S*6b here and that 54.P*6g is therefore winning. However, 
after 56.S*6b, black has a rather long and complicated mate: 57.Bx6b+ Sx6b S*7b 
Kx7b B6c+ Sx6c Px6c+ Kx6c R*5c K6b S*6c K7a R5a+ and whatever white drops in between 
leads to mate. The point here is that if white could drop a pawn on 6a, there would 
be no mate, but because of the pawn on 6g, this is not possible.

57.P6d-6c+   02:44:00  03:59:00

The decisive move. The black king has a safe route to 6d, while the white king 
has no defence.

58.P6gx6h+   02:44:00  03:59:00
59.G6ix6h    02:44:00  03:59:00
60.S*6g      02:44:00  03:59:00
61.K5hx6g    02:47:00  03:59:00
62.L*6f      02:47:00  03:59:00
63.K6gx6f    02:49:00  03:59:00
64.R8hx6h+   02:49:00  03:59:00
65.P*6g      02:49:00  03:59:00
66.S*5g      02:49:00  03:59:00
67.K6f-7e    02:49:00  03:59:00
68.G*8d      02:49:00  03:59:00
69.K7e-6d    02:49:00  03:59:00
70.G8dx7d    02:49:00  03:59:00
71.K6dx7d    02:49:00  03:59:00
Resigns      02:49:00  03:59:00	

Final Diagram

The only check left is 72.B*8e, but after K6d there is no continuation. On the 
other hand, white has no defence, so Sato resigned here. This seems like a straight 
victory for black, confirming the power of Habu's new move B*9f, but there are still 
a lot of close variations and it is probably too early to throw Sato's "brilliant 
blunder" P*5d out of the window. Habu takes a 2-1 lead in this best-of-five match, 
so the pressure is on Sato to deliver with the black pieces in game 4.