Sunday, April 1, 2018

Deep Blue versus Kasparov, 1997, Game 6 - Position by position computer analysis

The following excerpt has been taken verbatim from Wikipedia.
I've retained all hyperlinks from the original Wikipedia article; however, I have also hyperlinked each individual game move (e.g. 1.e4 c6) to the position-by-position computer analysis at the Open Encyclopedia of Chess Openings.

The game

White: Deep Blue   Black: Garry Kasparov   Opening: Caro–Kann Defense, Steinitz Variation (ECO B17)
1.e4 c6
Somewhat atypically, Kasparov plays the solid Caro–Kann Defense. In later matches against computers he opted for 1...e5 or 1...c5, the sharp Sicilian Defence, Kasparov's usual choice against human opponents.
Somewhat atypically, Kasparov plays the solid Caro–Kann Defense. In later matches against computers he opted for 1...e5 or 1...c5, the sharp Sicilian Defence, Kasparov's usual choice against human opponents.
2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7 5.Ng5 (see diagram)
[After 5.Ng5]
This relatively recent innovation breaks one of the classic opening principles ("don't move the same piece twice in the opening"), but puts pressure on the weak f7-square. Kasparov had played this move himself as White at least three times earlier.
Not 5...h6? 6.Ne6! fxe6?? 7.Qh5+ g6 8.Qxg6#; 6...Qb6 7.Nxf8 Nxf8 8.c3 Bf5 9.Nge2 Nf6 10.a4 N8d7 11.Ng3 Bg6 12.Bd3 and Deep Rybka 3 gives (0.13) advantage to White.
6.Bd3 e6 7.N1f3 h6? (see diagram)
[After 7...h6]
A strange choice by Kasparov, one of the most theoretically knowledgeable players in chess history. It has been suggested that it was a blunder and Kasparov got his opening moves mixed up, playing ...h6 a move too early. The normal 7...Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4 was played in Kasparov(!)–Kamsky, 1994 and Kasparov–Epishin, 1995, among other games. The upcoming sacrifice is well known to theory and Kasparov must have known about it (in fact, there are some reports that he even wrote an article supporting 8.Nxe6 as a refutation).
Feng-Hsiung Hsu, the system architect of Deep Blue, suggests that it was a deliberate 'anti-computer' move by Kasparov.[1] Objectively speaking, the move may be okay, although the resulting position is very tough for a human player to defend as black. White's response is very strong, but the computer programs Kasparov was familiar with could not play it properly. Several were specifically forbidden from playing Nxe6, because they lost too easily. So Hsu suggests that Kasparov expected that Deep Blue would either sacrifice the knight and then get into difficulties, or retreat it and lose a tempo.
The computer is aided by having this knight sacrifice programmed into its opening book. This move had been played in a number of previous high-level games, with White achieving a huge plus score. As an indication of how far computer chess has progressed in the 20 years after this match, modern programs deprived of their opening books are able to correctly evaluate Nxe6 as strongest; but at the time this was played it was considered probable based on other programs' performance that it was only the opening book that was responsible for this choice.[citation needed] The compensation White gets for the material is not obvious enough for the computer to see by itself.[2]
Instead of taking the knight immediately, Kasparov pins the knight to the king in order to give his king a square on d8. However, many annotators have criticized this move and said that Kasparov ought to have taken the knight immediately. Although the black king uses two moves to reach d8 after 8...fxe6 9.Bg6+ Ke7, the black queen can be placed on the superior c7-square.
White castles so that 9...Qxe6?? loses to 10.Re1, pinning and winning the black queen. Black must now take the knight or he will be a pawn down.
9... fxe6 10.Bg6 Kd8 11.Bf4 (see diagram)
[After 11.Bf4]
If Black's bishop were on d6 instead of f8, White would not be able to play this. For the sacrificed knight, White's bishops have a stranglehold on Black's position. Black, having moved his king, can no longer castle, his queen is blocking his own bishop, and he has trouble getting out his pieces and making use of his extra knight.
The first new move of the game and Deep Blue must now start thinking on its own. Kasparov's idea is to get some breathing room on his queenside and prevent White from playing c2–c4. However, this move has been marked as a mistake by Schwartzman,[3] Seirawan,[4] and Rajlich[5] as it weakens the queenside pawn structure and invites White to open lines.
12.a4 Bb7
Keeping lines closed with 12...b4 was mandatory according to Keene, but then 13.c4 would cramp Black's game.[6]
13.Re1 Nd5 14.Bg3 Kc8 15.axb5 cxb5 16.Qd3 Bc6 17.Bf5
White is pounding at Black's e6-pawn and is planning to invade the position with his rooks. Kasparov cannot hold onto all his extra material and must surrender his queen for a rook and a bishop.
17...exf5 18.Rxe7 Bxe7 19.c4 1–0
Black resigns because the white queen will soon invade through c4 or f5, and once Re1 is played, White will have a winning position. A sample line would be: 19...bxc4 20.Qxc4 Nb4 (20...Kb7 21.Qa6 mate!) 21.Re1 Kd8 22.Rxe7 Kxe7 23.Qxb4+.
[After 19.c4]
After the game Kasparov accused the Deep Blue team of cheating (i.e. having a team of human masters to aid the computer). Although Kasparov wanted another rematch, IBM declined and ended their Deep Blue program.
  1. Feng-Hsiung Hsu (2002). Behind Deep Blue. Princeton University Press. ISBN 0691118183.
  2. Chess Life. United States Chess Federation (Special Summer, 1997). Missing or empty |title= (help)
  3. Chess Life, Special Summer 1997
  4., see link in the "External links" section
  5. Rajlich, Vasik (2010). "Man vs Machine". New in Chess (2): 50–56.
  6. Raymond Keene (2005). Chess Terminators. Hardinge Simpole Publishing. p. 104. ISBN 1-84382-171-0.
Wikipedia contributors. (2018, March 23). Deep Blue versus Kasparov, 1997, Game 6. In Wikipedia, The Free Encyclopedia. Retrieved 04:02, April 1, 2018, from,_1997,_Game_6&oldid=831998847
Date of last revision: 23 March 2018 06:04 UTC

Sunday, April 30, 2017


by U.S. Champion Bobby Fischer
International Grandmaster

The King's Gambit has lost popularity, but not sympathy. Analysts treat it with kid gloves and seem reluctant to demonstrate an outright refuatation. "The Chessplayers Manual" by Gossip and Lipschutz, published in 1874, devotes 237 pages to this gambit without arriving at a conclusion. To this day the opening has been analyzed romantically - not scientifically. Moderns seem to share the same unconscious attitude that caused the old-timers to curse stubborn Steinitz: "He took the beauty out of chess."
To the public, the player of the King's Gambit exhibits courage and derring-do. The gambit has been making a comeback with the younger Soviet masters, notably Spassky (who defeated Bronstein, Averbach and myself with it). His victories rarely reflected the merits of the opening since his opponents went wrong in the mid-game. It is often the case, also, as with Santasiere and Bronstein, that the King's Gambit is played with a view to a favorable endgame. Spassky told me himself the gambit doesn't give White much, but he plays it because neither does the Ruy Lopez nor the Giuoco Piano.
The refuatation of any gambit begins with accepting it. In my opinion the King's Gambit is busted. It loses by force.
1 P-K4 P-K4 2 P-KB4 PxP 3 N-KB3 P-Q3!
This is the key to a troublesome position, a high-class "waiting move." At Mar Del Plata, 1959, I played 3...P-KN4 against Spassky, but this is inexact because it gives White drawing chances in the ensuing ending: e.g., 4 P-KR4 P-N5 5 N-K5 N-KB3 6 P-Q4 P-Q3 7 N-Q3 NxP 8 BxP B-N2 and now 9 P-B3! (replacing Spassky's 9 N-B3) 9...Q-K2 10 Q-K2 B-B4 11 N-Q2 leads to an ending where Black's extra Pawn is neutralized by White's stranglehold on the dark squares, especially KB4.
Another good try, but also inexact, is the Berlin Defense: 3...P-KR3 4 P-Q4 P-KN4 5 P-KR4 B-N2 6 P-KN3 P-N5 (also playable is 6...P-Q3 7 PxBP P-N5) 7 N-R2 PxP 8 NxP (8 QxP loses to 8...PxN 9 QxB QxP+ 10 K-Q1 Q-B3) 8...P-Q4 9 P-K5 B-B4 10 B-KB4, where Black cannot demonstrate any advantage.
Of course 3...P-Q4 equalizes easily, but that's all.
4 B-B4
4 P-Q4 transposes, the only difference if White tries to force matters after 4...P-KN4 5 P-KR4 P-N5 6 N-N5 (White also gets no compensation after 6 BxP PxN 7 QxP N-QB3 or 6 N-N1 B-R3) 6...P-KB3! 7 N-KR3 PxN 8 Q-R5+ K-Q2 9 BxP Q-K1! 10 Q-B3 K-Q1 and with his King and Queen reversed, Black wins easily.
This in conjunction with Black's previous move I would like to call the Berlin Defense Deferred. By this subtle transposition Black knocks out the possibility open to White in the last note (to move 3).
5 P-Q4 P-KN4 6 0-0 B-N2 7 P-B3
Necessary to protect the QP. 7 P-KN3 is always met by P-N5.
Here there is disagreement as to Black's best move. Puc and Rabar, Euwe, Keres, and most analysts give the text as the main line and mention 7...N-K2(!) in passing. I think 7...N-K2 is best because there is no reason why Black should not strive to castle K-side: e.g., 8 P-KN3 P-Q4! 9 PxQP PxNP 10 PxP (if 10 N-K5 PxP+! 11 K-R1 0-0 12 P-Q6 QxP wins) 10...0-0 11 Q-N3 Q-Q3 12 K-N2 N-B4 wins. There is little practical experience with this sub-variation.
8 Q-N3
If 8 P-KN3 P-N5 9 N-R4 P-B6 10 N-Q2, Euwe and other analysts betray their soft-mindedness toward this opening by giving the inferior 10...B-B3(?) 11 N(2)xP PxN 12 QxP - "unclear"!! This is yet another example of sentimental evaluation - after 12...Q-K2 followed by B-R6 and 0-0-0 Black wins easily. The Pawn on KB6 is a bone in White's throat so why force him to sacrifice when he must anyway? 10...Q-K2 is the strongest move.
In this last variation (instead of 10 N-Q2) White can vary with 10 Q-N3 but then comes Nimzovitch's beautiful winning line: 10...Q-K2 11 N-B5 BxN 12 PxB (if 12 QxP R-N1 13 QxN+ Q-Q2 14 QxQ+ BxQ and Black has a winning endgame) 12...0-0-0 13 BxP Q-K7 14 Q-K6+ (if 14 R-B2 NxQP! 15 RxQ PxR wins) 14...R-Q2! 15 R-B2 Q-Q8+ 16 R-B1 Q-B7 17 N-Q2 N-B3 (threatening N-Q1) 18 B-N6 (if 18 Q-N3 QxQ 19 BxQ P-Q4 with a winning endgame) 18...P-Q4 followed by N-K2 with a winning game for Black.
8...Q-K2 9 P-KR4 N-B3
Again theoretical disagreement. Perfectly good is 9...P-N5! 10 BxP (forced, not 10 KN-Q2 NxQP! 11 PxN BxP+ etc.) 10...PxN 11 RxP - given by analysts again as "unclear," but after N-B3 followed by 0-0, White has nothing for the piece.
10 PxP PxP 11 NxP NxKP
A wild position, but Black is still master.
12 BxP+
The game is rife with possibilities. If 12 NxN QxN 13 RxP Q-K8+ 14 R-B1 Q-R5 15 BxP+ K-Q1 16 Q-Q5 N-K4! 17 PxN BxP (threatening B-R7 and mate) 18 R-Q1 Q-N6 wins, owing to the threat of R-R8+.
12...K-Q1 13 NxN
Not 13 N-K6+ BxN 14 QxB QxQ 15 BxQ NxQP!
13...QxN 14 BxP
14 RxP also loses to 14...Q-K8+ 15 R-B1 R-R8+ 16 KxR QxR+ 17 K-R2 QxQB etc.
And Black wins...
Of course White can always play differently, in which case he merely loses differently.
  1. Bobby Fischer, "A Bust to the King's Gambit", American Chess Quarterly, Vol. 1, No. 1 (Summer 1961), pp. 3-9.

Sunday, August 7, 2016

Philidor Countergambit (PCG) Variation Tree | Position-by-position Computer Analysis

Philidor Defense

1.e4 e5 2.Nf3 d6  

Lopez Countergambit

  1. 3.Bc4 f5 
    1. 4.d3
      1. 4...Be7
      2. 4...c6
        1. 5.a4
        2. 5.exf5 Bxf5
          1. 6.Bg5 Nf6
          2. 6.0-0 Be7
        3. 5.Nc3
        4. 5.0-0
      3. 4...Nc6
        1. 5.Nc3
        2. 5.Ng5 Nh6
        3. 5.0-0
    2. 4.d4
      1. 4...exd4
      2. 4...fxe4 5.Nxe5
      3. 4...Nc6
        1. 5.dxe5
          1. 5...dxe5 6.Qxd8
            1. 6...Kxd8
            2. 6...Nxd8
              1. 7.exf5 7.exf5 Bxf5
              2. 7.Nc3
              3. 7.Nxe5
              4. 7.0-0 fxe4 8.Nxe5 Nf6
          2. 5...fxe4
        2. 5.Ng5 Nh6 6.d5
          1. 6...f4 7.h4 Nb8 8.Ne6
          2. 6...Ne7
            1. 7.exf5
            2. 7.Nc3
        3. 5.0-0 fxe4 6.Nfd2 d5 7.Bb5
    3. 4.exf5 Bxf5
      1. 5.d3
      2. 5.Nc3
      3. 5.0-0 Be7
    4. 4.Nc3
      1. 4...Be7
      2. 4...Nc6 5.d4 Be7 6.dxe5 dxe5
      3. 4...Nf6
    5. 4.Ng5 Qxg5

Philidor Countergambit

  1. 3.d4 f5

    1. 4.Bc4 (See 4.d4 above in Lopez Countergambit)
    2. 4.Bd3 fxe4 5.Bxe4 d5
      1. 6.Bg5 Be7 7.Nxe5 dxe4 8.Qh5 g6 9.Nxg6 Bg4 10.Qxg4 hxg6
        1. 11.Be3
        2. 11.Qe5
        3. 11.Qxe4
          1. 11...Kd7
          2. 11...Kf7
          3. 11...Kf8
          4. 11...Nc6 12.Qxg6 Kf8 13.Be3 Nxd4 14.Nc3 Nc6
            1. 15.h4
            2. 15.Ne2
            3. 15.Qf5 Nf6 16.Qg6
      2. 6.Nfd2
      3. 6.Nxe5 dxe4 7.Qh5 g6 8.Nxg6 hxg6 9.Qxh8 Kf7
        1. 10.0-0 Bg7 11.Qh7 Bf5 12.Na3 Qxd4 13.Qh4 Nf6 14.Qg3 Na6 15.Qb3
        2. 10.Qh7 Bg7
    3. 4.dxe5 fxe4
      1. 5.Nd4 a6
      2. 5.Nfd2
        1. 5...Bf5
        2. 5...d5
        3. 5...e3 6.fxe3
        4. 5...Nc6
      3. 5.Ng5 d5
        1. 6.c4 Bb4
          1. 7.Bd2 Qxg5 8.Bxb4 d4 9.Qxd4 Nc6
            1. 10.Bd2 Nxd4 11.Bxg5
              1. 11...h6 12.Be3 Nc2 13.Kd2 Nxa1 14.Be2 Bf5
              2. 11...Nc2
                1. 12.Be3
                2. 12.Kd2 Nxa1 13.Be2
                  1. 13...Bf5 14.Na3 h6 15.Be3 0-0-0 16.Kc3 Ne7 17.Rxa1 Nc6
                  2. 13...h6
                    1. 14.Be3
                    2. 14.Bh5+
            2. 10.Qc5 Qc1 11.Ke2
              1. 11...Bg4+
              2. 11...Nh6
            3. 10.Qd2 e3 11.fxe3 Nxb4 12.Qxb4 Ne7
          2. 7.Nc3 d4 8.a3 Bxc3 9.bxc3
            1. 9...e3 10.f4
            2. 9...Nc6
        2. 6.e6
          1. 6...Bb4 7.c3 Bc5 8.Nxe4 Be7
          2. 6...Bc5
            1. 7.Nc3
            2. 7.Nf7 Qf6
              1. 8.Be3 Bxe6 9.Nxh8 Bxe3 10.fxe3
                1. 10...Nc6 11.Nc3 0-0-0
                2. 10...Qxb2 11.Nd2 Qc3 12.Rb1 Qxe3 13.Be2 Nf6 14.Rxb7 Nd7 15.Nb3 0-0-0 16.Rb5 Bg4 17.Qd2 Qxe2 18.Qxe2 Bxe2 19.Kxe2 Rxh8 20.Ra5 Kb7 21.Rb1 Nb6 22.Nd4 Rf8
              2. 8.Qd2
            3. 7.Nxe4 Bb4 8.c3 dxe4 9.Qxd8 Kxd8 10.cxb4 Bxe6 11.Nc3 Nf6 12.Bg5 Nbd7 13.Nxe4 h6 14.Bxf6 Nxf6 15.Nxf6 gxf6 16.0-0-0 Kc8
          3. 6...Nc6
          4. 6...Nf6 7.Nf7 Qe7 8.Nxh8 Bxe6
            1. 9.Be2 Nc6 10.Bh5 g6
              1. 11.Be2
              2. 11.Nxg6 hxg6 12.Bxg6 Kd7 13.h4 Qg7 14.h5 Bc5
                1. 15.Be3 Bxe3 16.fxe3 Qh6
                2. 15.Rh4
            2. 9.Nc3 Nc6 10.Bb5
          5. 6...Nh6 7.Nc3 c6
            1. 8.Be3 Nf5 9.Nf7 Nxe3 10.fxe3
            2. 8.g3
              1. 8...g6
              2. 8...Qe7 9.Bh3 g6
            3. 8.Ngxe4
              1. 8...dxe4 9.Qh5 g6 10.Qe5 Rg8 11.Bg5 Bg7 12.e7 Qd7 13.Qf4 Qf5
              2. 8...Nf5
                1. 9.Ng3
                  1. 9...Bxe6 10.Bd3
                  2. 9...Nxg3 10.hxg3 Bxe6
                  3. 9...Qf6
                2. 9.Ng5 Qf6 10.Bd3
        3. 6.f4
          1. 6...Bb4+
          2. 6...Bc5
          3. 6...Be7
        4. 6.g3
          1. 6...Bb4
            1. 7.c3
            2. 7.Nc3
            3. 7.Bd2
          2. 6...Be7
          3. 6...Nc6
        5. 6.Nc3 Bb4
          1. 7.a3 Bxc3 8.bxc3 Ne7
    4. 4.exf5
      1. 4...Bxf5 5.Nc3 e4 6.Ng5 d5 7.f3 e3 8.Bxe3 h6
      2. 4...e4
          1. 5.Ne5
            1. 5...Bxf5
            2. 5...dxe5 6.Qh5
          2. 5.Nfd2
          3. 5.Ng1 Bxf5
            1. 6.c4
            2. 6.d5
            3. 6.Nc3 d5
            4. 6.Ne2
          4. 5.Ng5
            1. 5...Bxf5 6.Nc3 d5 7.f3 e3 8.Bxe3
            2. 5...Nf6
          5. 5.Qe2
    5. 4.Nc3 fxe4 5.Nxe4
      1. 5...d5
        1. 6.Nc3 e4 7.Ne5
        2. 6.Ned2
        3. 6.Neg5
          1. 6...e4 7.Ne5 Nh6 8.Nxe4 dxe4 9.Bxh6 Qf6
          2. 6...exd4
          3. 6...h6
            1. 7.Bb5
            2. 7.Nf7
        4. 6.Ng3
          1. 6...e4 7.Ne5 Nf6
            1. 8.Be2
            2. 8.Bg5 Bd6 9.Nh5 0-0 10.Qd2 Qe8 11.g4 Nxg4 12.Nxg4 Qxh5 13.Ne5 Nc6 14.Be2 Qh3 15.Nxc6 bxc6 16.Be3 Rb8 17.0-0-0 Rxf2 18.Bxf2 Qa3 19.c3 Qxa2 20.b4 Qa1 21.Kc2 Qa4 22.Kb2 Bxb4 23.cxb4 Rxb4 24.Qxb4 Qxb4 25.Kc2 e3 26.Bxe3 Bf5 27.Rd3 Qc4 28.Kd2 Qa2 29.Kd1 Qb1
            3. 8.f3
          2. 6...exd4
        5. 6.Nxe5 dxe4
      2. 5...Nf6
        1. 6.Bb5
        2. 6.Bg5 Be7
        3. 6.Nxf6 gxf6 

        Visit for an up-to-date variation tree.