# Shogi Game 28 (professional, full annotation).
# 
# Please find below Game 4 of the 48th Meijin Title Match (1990) with full
# comment by Tanigawa Koji (Meijin).
# 
#    John Kenney <kenney@embl-heidelberg.de>
#    Submitted 5-th May 1992.
#    Pieter Stouten <stouten@embl-heidelberg.de>
#    Archived 11-th May 1992.
# 
# -------------------------------------------------------------------------
# 
# 48th Meijin Title Match - Game 4 by Tanigawa Koji, Meijin
# 
# Written by Tanigawa Koji. Translated by David Murphy.
# Typed by John Kenney. Checked by John Kenney and Pieter Stouten.
# 
# In this match so far the play has been passable.  Even the game which I
# lost from a won position.  From the opening I expected to use a lot of
# time and effort in building up an advantagous position so that during
# the game I had no time for disappointment.  But, when I got home and once
# again was watching satellite TV, etc., the quiet anger of "gyakuten"
# (turn-around) sunk in.  Still, I came to this game with the resolve to
# play it as comfortably as possible, feeling that I'd been putting too
# much into it so far.  Yet, not too easy going, because losing this game
# would be quite serious, but...
# 
# 48th Meijin, Game 4, 21 and 22 May 1990 (players have 9 hours each)
# Black:  Tanigawa Koji, Meijin.  White:  Nakahara Makoto, Kisei.
1.  P7f        P8d(1)
2.  S6h(2)     P3d
3.  P6f        S6b(1)
4.  P5f        P5d
5.  S4h        S4b(2)
6.  G7h        G3b(1)
7.  K6i(2)     K4a(2)
8.  G5h(3)     P6d
9.  P2f(22)    P8e(5)

[ Since I used the Bishop exchange as Black in Game 2, I thought to play
  a Fortress this time.  However, if I did that, it was 100% certain that
  Nakahara would adopt the system he played in this game, so I had the
  problem of finding a counter to it.  I'd already lost as Black against
  this system to both Nakahara in a quick play game and Habu (Ryu-o) in
  Game 3 of the All-Japan Pro, so it would have been poor to play this
  without anything prepared.  Therefore, though it is slightly
  contradictory, I was in fact undecided as to whether to play a Fortress
  or a Bishop exchange opening right up to directly before the game. 

  The result of my preparation effort was to hit upon P6f.  Though it is
  not that unusual a move in itself, I thought it was better than 3 S7g
  against the "Nakahara style" system.  The explanation is really for
  experts, but one counter measure is to retreat the silver at 7g to 6h, and
  play B7g. So, if one is going for this line, it must be better to leave the
  silver on 6h.  Because of this Nakahara played 9 P8e to force my silver
  to 7g early on.  For my part, by keeping S7g back I forced him to play
  P8e.  The timing of this move is delicate.  I feel that it could have
  been held back just a little. ]

10. S7g(2)     P5e
11. Px5e(6)    Bx5e
12. P2e        B2b(1)
13. G5h-6g(15) S6c
14. B7i(2)     P7d(7)
15. P2d(9)     Px2d
16. Bx2d       P'2c
17. B6h

[ Playing 3 P6f was one idea I had, a second was 14 B7i.  When the
  "Nakahara style" system began to be played, I felt that White would
  have great trouble making his attack a success in view of his weak King
  position. Though he has seized the strategic initiative a severe counter
  attack would follow in the wake of his attack, I thought.  But, after
  having faced it in practice I found the going tough, with no counter
  play, and was forced to revise my opinion of its validity.  So, in the
  Kisei tournament on April 20th I played the "Nakahara style" for the
  first time with White against Takahshi 9-Dan, thinking I ought to
  consider playing White's side, since Black has no counter to it.  In
  that game, Takahashi countered by exchanging off his rook pawn with his
  pawn and then playing B7i/B6h.  So I thought that if I retreated B7i
  first and exchanged off my 2-file pawn with the Bishop I'd gain a
  strategic victory.  For clearly it is a one tempo gain over
  Takahashi's method. ]

17. ...        S5d(17)
18. S5g(6)     S5c(1)
19. S4f(3)     S4d
20. P'5f(4)    R5b(2)
21. K7i(17)    G6b

[ Exchanging off the 2nd file pawn with the rook is safe, but doing it with
  the Bishop is slightly dangerous.  The reason is that after 17 B6h White
  has the move 17 ... N7c.  However, after 18 S5g P6e 19 S5f White can't
  launch a successful attack, as was established in the post-mortem.
  Naturally, Nakahara, too, had read out this sequence.  In any event he
  gave up the idea and played the quiet 17 S5d.  To tell the truth, I was
  a little relieved. For now I have definitely gained a clear tempo.  That
  the attack with 17 ... N7c 18 ... P6e does not work owes a lot to the
  move 9 ... P8e.  Had he played another move in place of this it would
  have made a big difference.  20 ... G5h instead of 20 ... R5b would be
  usual, but since Black is a tempo up compared to the normal position,
  he judged that is would be poor to follow along the same lines. ]

22. R2e(24)    P'5e(52)
23. Px5e(13)   N3c(1)
24. R2h(4)     S5dx5e

[ On the last move White had a difficult choice.  21 ... G7b was the
  alternative, but it has a drawback.  For then the attack P3f and P3e
  works.  Against P3f White can play opposing silvers by S5d-4e.  However,
  with the Gold on 7b, a nasty silver stab at 6a is created after the
  silver exchange.  But, on the other hand, after 21 ... G6b as in the
  game, 22 R2e! works.  The threat is simply 23 Rx8e, but it is
  suprisingly hard to meet.  22 ... N7c would be normal, but then 23 P7e S6c
  24 Px7d Sx7d 25 P'7e S6c 26 K8h with the outpost on the 7th file Black
  has a clear advantage.  So Nakahara in reply played 22 ... P'5e. For my
  part I wanted to force this pawn drop to bring about a "semeai" (mutual
  fight).  The following sequence was forced.  If he recaptures on 5e
  without playing 23 ... N3c, then I exchange Silvers and play P'5f, to
  be followed by Rx8e. ]

25. Sx5e(75)   Sx5e(50)
26. S'2a(7)    P1d(1)

[ I spent an hour and a quarter over 25 Sx5e.  I could have played 25 P'5f
  Sx4f 26 Bx4f, but I was going deeply into all the ramifications of the
  game sequence rather than that during this time.

  White's reply became the sealed move.  I had been expecting 25 ... Sx5e.  I,
  therefore, planned to think about the 25 ... Rx5e possiblity.  In that case,
  26 P'5f R5a 27 K8h leads to another game entirely.  However, Nakahara chose
  25 ... Sx5e because he was worried about 26 S'8b in reply to 25 ... Rx5e.
  26 S'8b isn't my style and I don't think I'd have played it, but going for
  some material gain like this may be the best move.

  Returning to the game, 26 S'2a was my idea!  If in reply 26 ... G3a
  27 P'2d Px2d 28 P'2c Gx2a 29 Px2b+ Gx2b, gaining a Bishop for a Silver,
  then not unnaturally I would be better.  However, White has S5f, an
  unpleasant move for Black, so this was a variation I was a little
  worried about.  26 ... P1d on the other hand does prepare to meet 27
  Sx3b+ Kx3b 28 P'2d Px2d 29 P'2c with 29 ... B1c!  But, I was not expecting
  it. ]

27. P'5f(17)   Sx5f(53)
28. Gx5f(24)

[ After finding 27 P'5f I thought I was clearly winning!  If White
  retreats 27 ... S4d I have no quick win, but after 28 Sx3b+ Kx3b 29 K8h
  White's position is in disarray.  Having said that I was expecting 27
  ... S4d.  However, Nakahara took the pawn with 27 ... Sx5f.  This bad
  move was a surprise.  At this stage it was still only about 10:20am on
  the second day of the game.  To be honest I thought the game would be over
  in the morning.  I spent 17 minutes on 27 P'5f, analysing the mating
  lines, and 24 more minutes checking them one more time before playing
  Gx5f.  However, this was a serious error.  If move order was 28 Sx3b+ Kx3b
  then 29 Gx5f was correct.  This would have been clear cut.  I didn't think
  about which to play here, Sx3b+ or Gx5f.  I thought not unnaturally that
  28 ... Rx5f was his only reply to 28 Gx5f ... ]

28. ...        G3a(30)
29. Rx2c+(5)   Gx2a
30. P'5d       Rx5d(4)
31. S'5e(21)   R5a(14)
32. +Rx3d      G3b(3)
33. B3e(4)

[ If he had played 28 ... Rx5f, I intended 29 Sx3b+ Kx3b 30 P'2d Px2d 31
  S'2c! Kx2c 32 Rx2d K3b 33 G'4b Kx4b 34 Rx2b+.  If then 34 ... S'3b 35
  B'3a K4a 36 Bx6d+, threatening both 37 +R3a, etc. mating and 37 +B4f,
  wins. Or if 34 ... G'3b 35 B'3a K4a 36 +Rx3b Kx3b 37 G'4b K2a 38 P'2c
  G'1b 39 B6h-1c+ and White can not prevent mate.  In truth the game
  would probably have been over before noon.  However, Nakahara played
  ... G3a!  Had I played 28 Sx3b+ first this could not have happened.
  When he played it I not only regretted getting the move order wrong,
  but I was also a little impatient.  After 28 ... G3a! it became a bit
  dangerous.  Being now in an unstable psychological state too, there was
  danger of a "gyakuten" (turn-around).  If 31 ... R5b, intead of
  Nakahara's 31 ... R5a, then 32 +Rx3d G3b 33 P'2d S'2e 34 +Rx6d S'6c 35
  P'5c Gx5c 36 +Rx5c Rx5c 37 P'5d R5b 38 G'5c. This must still be slightly
  better for Black, we decided in the postmortem, but the truth is I was
  uneasy about it, having given him a rook in hand, and ... P'6g! in the
  air.  But because he retreated the rook all the way to 5a, 32 B3e hit
  the loose Gold.  Even so, I was still uneasy. ]

33. ...        G5b(28)
34. P'5c(24)   G5b4b(3)
35. P'2d(3)    B1c(2)
36. K8h

[ Here I was more worried about the defense 33 ... P'5c.  If in reply 34
  P'5d S'2c 35 Px5c+ Gx5c 36 Bx5c+ Sx3d 37 P'5b, then 37 ... B1c! and 38
  P4f if, 38 ... Rx5b 39 +Bx5b K3a, suprisingly a position in which White
  is hard to mate. If I can't do it, then with the opponent having rook and
  silver in hand would spell trouble.  I had been calculating this endgame
  specifically, and then along the way I noticed 37 ... B1c! Frankly I had
  lost my vigilence.  With the memory of Game 3 in mind, I was predicting
  a gloomy outcome.  However, Nakahara went 33 ... G5b.  Then after 34 ...
  G5b-4b there was no longer any danger of losing my rook so I had nothing
  to fear.  36 K8h was the decisive move.  After that all my worries were over. ]

36. ...        P8f(14)
37. Sx6d(2)    S'4d(63)
38. Bx4d(4)    Px4d
39. P2c+       S'7i
40. Gx7i       Px8g+
41. Kx8g       Bx7i+
42. S'5b(1)    Rx5b
43. Px5b+      Kx5b
44. R'8b       P'6b
45. P'5c       Resigns
# 
# (Time taken:  Black, 5:51; White, 6:29)
# 
# So I won with Black, but I haven't been able to win with White.  That is
# the crux of the matter.  Nakahara will probably play "Aigahari" in Game 5.
# Though I lost both Games 1 and 3 my strategy was overall a success regarding
# the positions reached.  So it will be interesting to see what Nakahara comes
# up with next time...
# 

