Naum 4 and bishop pair ... long post only for chess freaks!

[M ANSARI]

After looking at a lot of Naum 4 games I can't help but notice how highly it evaluates the bishop pair. Although primitively high in many positions, it is extremely dangerous in positions where the bishop pair has a lot of range and scope. This evaluation inflation for the bishop pair is one of the reasons N4 is able to get a lot of wins from Rybka 3. Here are a few games that I have chosen that shows some of N4's play with bishop pair.


(3) Naum 4 - Rybka 3 _0_cont [B90]
N4_30_1_new–1 (8.2), 24.12.2008
[0.37;0.19]
Intel(R) Core(TM)2 Extreme CPU X9775 @ 3.20GHz 3990 MHzW=18.7 plies; 6,618kN/s; 80,012 TBAs; Perfect 15.ctgB=17.0 plies; 459kN/s; 21 TBAs; Perfect 15.ctg 1.e4 B/0 0 c5 B/0 0 2.Nf3 B/0 0 d6 B/0 0 3.d4 B/0 0 cxd4 B/0 0 4.Nxd4 B/0 0 Nf6 B/0 0 5.Nc3 B/0 0 a6 B/0 0 6.Be3 B/0 0 e5 B/0 0 7.Nb3 B/0 0 Be6 B/0 0 8.f3 B/0 0 Be7 B/0 0 9.Qd2 B/0 0 0–0 B/0 0 10.0–0–0 B/0 0 Nbd7 B/0 0 11.g4 0.37/21 53 b5 0.19/19 1:18 12.g5 0.31/21 1:38 b4 (Nh5) 0.21/18 16 13.gxf6 0.27/21 40 bxc3 0.21/17 0 14.Qxc3 0.16/22 1:27 Nxf6 0.23/19 35 15.Na5 0.29/22 1:58 Rc8 (d5) 0.25/18 39 16.Nc6 0.26/22 57 Qd7 (Qc7) 0.25/19 41 17.Nxe7+ 0.19/21 2:09 Qxe7 0.25/18 0 18.Qa5 0.25/21 42 Rc6 0.24/18 45 19.Rg1 0.25/21 38 Rfc8 (g6) 0.18/17 25 20.Kb1 (c3) 0.21/19 40 20...d5 (g6) 0.14/17 46 21.b3 0.02/17 32 g6 (d4) 0.20/17 56 22.exd5 0.74/18 19 Nxd5 0.11/18 21 23.Rxd5 0.41/19 13 Bxd5 0.11/17 0 24.Qxd5 0.41/21 25 Rxc2 0.11/17 0 25.Rg2 0.33/21 1:35 Qc7 (Rxg2) 0.15/18 31 26.Qd1 0.59/20 18 Rxg2 0.17/19 21 27.Bxg2 0.59/6 0 a5 (e4) 0.20/19 25 28.Bf1 0.84/20 1:01 Rd8 (Rb8) 0.21/18 24 29.Qe1 (Qc1) 0.88/21 1:08 29...a4 (Qb7) 0.29/17 58 30.Bc4 0.97/20 1:10 axb3 0.30/17 51 31.axb3 0.99/20 23 Re8 (Qd6) 0.34/18 44 32.Bd2 1.07/18 13 Kg7 0.36/17 31 33.Qe2 (Qh4) 1.07/20 31 33...Rc8 (f6) 0.31/17 46 34.Bc3 (f4) 1.07/19 17 34...f6 0.38/18 17 35.Kc2 1.07/20 16 Qc5 0.38/17 18 36.Qd2 (Qd3) 1.03/21 54 36...Qa7 (Qc7) 0.50/17 42 37.Kb2 (Qd6) 1.04/19 11 37...Qb7 0.38/17 31 38.Qd3 1.04/18 9 Qc6 0.39/18 14 39.Ba5 (h4) 1.04/18 12 39...Ra8 (Qc5) 0.38/17 16 40.Bd2 1.03/20 14 Rc8 (Qc7) 0.48/18 57 41.h4 (f4) 1.08/19 24 41...Qc7 0.44/18 13 42.Bc3 (h5) 1.07/19 6 42...Rd8 (Qc6) 0.47/18 1:14 43.Qe2 (Qe4) 1.07/19 16 43...Rc8 0.41/17 14 44.Qd1 (h5) 1.07/20 16 44...Qb7 (Qc6) 0.38/16 15 45.Qf1 (Qd3) 1.07/17 10 45...Qc7 0.31/17 14 46.Kc2 1.07/19 8 Qc6 (Qc5) 0.40/16 15 47.Qh3 1.14/20 22 Ra8 (h5) 0.51/16 36 48.h5 1.45/17 12 Ra2+ 0.58/16 12 49.Bb2 1.42/18 5 Ra8 0.60/17 14 50.h6+ 1.54/19 10 Kf8 0.74/18 43 51.Qf1 1.53/19 10 Ke8 (Ke7) 0.78/17 11 52.Bc3 1.67/18 16 Ke7 0.81/17 13 53.Kb2 1.67/18 9 Qd6 (Rb8) 0.81/17 8 54.Qe2 (Qe1) 1.70/18 8 54...Kd7 (Kd8) 0.81/16 11 55.b4 (Qe4) 1.70/17 4 55...Rc8 (Ra7) 0.62/15 9 56.b5 (Bb5+) 1.69/18 14 56...Kc7 (Ra8) 0.75/16 5 57.Ba5+ (Qe4) 1.81/17 6 57...Kb8 (Kd7) 0.80/17 10 58.Kb3 (Qe4) 1.70/20 8 58...Qd4 1.04/18 27 59.Bc3 1.62/19 36 Qd6 1.24/18 35 60.Bb4 1.73/20 26 Qd7 (Qb6) 1.26/18 45 61.Qe3 1.81/19 17 Kb7 (Qd1+) 1.26/18 20 62.Bc5 1.98/18 12 Rd8 (f5) 1.26/18 13 63.Qe4+ 2.18/18 5 Kb8 2.55/19 1:38 64.b6 2.35/19 7 Rc8 2.55/19 36 65.Qe3 2.35/19 4 Rd8 2.55/19 54 66.Kb4 2.37/19 4 Qd1 2.55/18 1:20 67.Qe4 2.37/19 7 Qd7 2.55/17 56 68.Bb3 (f4) 2.57/19 4 68...Qe8 (Rc8) 3.12/14 3 69.f4 (Ba4) 4.69/15 4 69...Qd7 (Rd2) 3.34/12 39 70.fxe5 5.73/14 4 Qd2+ 3.34/11 27 71.Ka4 6.68/15 8 f5 7.01/11 19 72.Qc4 6.68/15 3 Kb7 8.61/10 13 73.Qf7+ 6.68/14 23 Rd7 10.72/11 10 74.e6 7.71/14 6 1–0


1.e4 B/0 0 c5 B/0 0 2.Nf3 B/0 0 d6 B/0 0 3.d4 B/0 0 cxd4 B/0 0 4.Nxd4 B/0 0 Nf6 B/0 0 5.Nc3 B/0 0 a6 B/0 0 6.Be3 B/0 0 e5 B/0 0 7.Nb3 B/0 0 Be6 B/0 0 8.f3 B/0 0 Be7 B/0 0 9.Qd2 B/0 0 0–0 B/0 0 10.0–0–0 B/0 0 Nbd7 B/0 0 11.g4 0.37/21 53 b5 0.19/19 1:18 12.g5 0.31/21 1:38 b4 (Nh5) 0.21/18 16 13.gxf6 0.27/21 40 bxc3 0.21/17 0 14.Qxc3 0.16/22 1:27 Nxf6 0.23/19 35 15.Na5 0.29/22 1:58 Rc8 (d5) 0.25/18 39 16.Nc6 0.26/22 57

[d]2rq1rk1/4bppp/p1Npbn2/4p3/4P3/2Q1BP2/PPP4P/2KR1B1R b - - 0 16

N4 goes for a forced bishop removal thus keeping a bishop pair advantage

Qd7 (Qc7) 0.25/19 41 17.Nxe7+ 0.19/21 2:09 Qxe7 0.25/18 0 18.Qa5 0.25/21 42 Rc6 0.24/18 45 19.Rg1 0.25/21 38 Rfc8 (g6) 0.18/17 25 20.Kb1 (c3) 0.21/19 40 20...d5 (g6) 0.14/17 46 21.b3 0.02/17 32 g6 (d4) 0.20/17 56 22.exd5 0.74/18 19 Nxd5 0.11/18 21

[d]2r3k1/4qp1p/p1r1b1p1/Q2np3/8/1P2BP2/P1P4P/1K1R1BR1 w - - 0 23

Very complicated position with many chances of material imbalances ... evaluation in such positions is where R3 gets a huge number of wins as it have by far the most superior evaluation. But ...

23.Rxd5 0.41/19 13

[d]2r3k1/4qp1p/p1r1b1p1/Q2Rp3/8/1P2BP2/P1P4P/1K3BR1 b - - 0 23

This move is not a surprise for R3 as it was expecting it ... this is all based on some exchanges that will leave a material imbalance in which static evaluation will be critical

Bxd5 0.11/17 0 24.Qxd5 0.41/21 25 Rxc2 0.11/17 0 25.Rg2 0.33/21 1:35 Qc7 (Rxg2) 0.15/18 31 26.Qd1 0.59/20 18 Rxg2 0.17/19 21 27.Bxg2 0.59/6 0 a5 (e4) 0.20/19 25 28.Bf1 0.84/20 1:01 Rd8 (Rb8) 0.21/18 24 29.Qe1 (Qc1) 0.88/21 1:08 29...a4 (Qb7) 0.29/17 58 30.Bc4 0.97/20 1:10 axb3 0.30/17 51 31.axb3 0.99/20 23

[d]3r2k1/2q2p1p/6p1/4p3/2B5/1P2BP2/7P/1K2Q3 b - - 0 31

Well look at the position ... N4 is very happy that it has the bishop pair, on the other hand R3 is not alarmed at all as it has material equality ... here is static evaluation of both sides after 5 minutes thought on Octa 4Ghz

3: Naum 4 - Rybka 3 _0_cont, N4_30_1_new-1 2008
3r2k1/2q2p1p/6p1/4p3/2B5/1P2BP2/7P/1K2Q3 b - - 0 1

Analysis by Rybka 3:

31...Rb8 32.Qc3 Qd6
= (0.13) Depth: 6 00:00:00 3kN
31...Rb8 32.Qc3 Qd6 33.Qa5
= (0.13) Depth: 7 00:00:00 8kN
31...Rb8 32.Qc3 Rd8 33.Kb2 Qe7
= (0.15) Depth: 8 00:00:00 17kN
31...Rb8 32.Qc3 Rd8 33.Kb2 Qe7 34.b4
= (0.16) Depth: 9 00:00:00 28kN
31...Rb8 32.Qc3 Rd8 33.Kc2 Qe7 34.Qa5
= (0.16) Depth: 10 00:00:00 59kN
31...Rb8 32.Qc3 Rd8 33.Kc2 Qe7 34.Qa5
= (0.25) Depth: 11 00:00:00 144kN
31...Kg7 32.Kb2 f6 33.h4 Rb8 34.Qc3 Rc8 35.Qd2
= (0.18) Depth: 11 00:00:01 547kN
31...Kg7 32.Qc3 f6 33.b4 Kh8 34.Kb2 Qd7 35.Qe1 Qb7
= (0.20) Depth: 12 00:00:02 811kN
31...Kg7 32.h4 Qc6 33.Qa5 Re8 34.Qb5 Qxb5 35.Bxb5 Re7 36.Kc2 e4 37.Kd2
= (0.22) Depth: 13 00:00:02 987kN
31...Kg7 32.h4 Qc6 33.Qc3 f6 34.Qb4 Re8 35.Qa3 Qd7 36.Qa5 Rc8
= (0.21) Depth: 14 00:00:03 1400kN
31...Kg7 32.Qc3 f6 33.b4 h5 34.Kb2 e4 35.f4 Rd1 36.h4 Qb7 37.Bd4 Qc6
= (0.24) Depth: 15 00:00:09 3762kN
31...Kg7 32.Qc3 f6 33.b4 h5 34.Kb2 e4 35.f4 Rd1 36.h4 Qb7 37.Bd4 Qc6
+/= (0.33) Depth: 16 00:00:14 5802kN
31...Qd6 32.Qa5 Kg7 33.h4 Qf6 34.Kc2 Qf5+ 35.Kc1 Rd7 36.b4 h6 37.f4 Re7 38.Qc5
+/= (0.32) Depth: 16 00:00:33 13999kN
31...Qd6 32.Qa5
+/= (0.33) Depth: 17 00:00:58 24262kN
31...Re8 32.Bc1 Qc6 33.Qe3 Qd7 34.Qe4 Kg7
+/= (0.32) Depth: 17 00:01:20 31943kN
31...Re8 32.Bd2 Kg7 33.Qh4 Qd6 34.Bh6+ Kg8 35.Qe4
+/= (0.36) Depth: 18 00:01:46 43459kN
31...Re8 32.Bd2 Kg7 33.Qh4 Qd6 34.Bh6+ Kg8 35.Qe4 Rd8 36.Be3
+/= (0.40) Depth: 19 00:03:06 78000kN

(, 23.01.2009)

3: Naum 4 - Rybka 3 _0_cont, N4_30_1_new-1 2008
3r2k1/2q2p1p/6p1/4p3/2B5/1P2BP2/7P/1K2Q3 b - - 0 1

Analysis by Naum 4:

31...Qd6 32.Qa5 Rc8 33.Kb2 Qf6 34.Qc3
+/- (0.95) Depth: 6/16 00:00:00 19kN
31...Qd6 32.Qa5 Rc8 33.Kb2 Qf6 34.Qc3 Qd6
+/- (0.96) Depth: 7/15 00:00:00 42kN
31...Qd6 32.Qg3 Qf6 33.b4 Qd6 34.Bc5 Qd1+ 35.Kb2
+/- (0.94) Depth: 8/19 00:00:00 93kN
31...Qb7 32.Qa5 Rc8 33.Bd5 Qc7 34.Qd2 Qc3 35.Qe2
+/- (0.92) Depth: 8/19 00:00:00 104kN
31...Qb7 32.Qg3 e4 33.fxe4 Qxe4+ 34.Kb2 Re8 35.Qf2 Qf5
+/- (0.94) Depth: 9/24 00:00:00 200kN
31...Qb7 32.Qg3 e4 33.fxe4 Qxe4+ 34.Kb2 Re8 35.Qf2 Qf5 36.Qd2
+/- (0.98) Depth: 10/26 00:00:00 317kN
31...Qd6 32.Qa5 Rc8 33.Kb2 e4 34.fxe4 Qxh2+ 35.Kc3 Qg3 36.Qa7
+/- (0.90) Depth: 10/26 00:00:00 397kN
31...Qd6 32.Qa5 Re8 33.Bd5 Rd8 34.Bg5 Rd7 35.Qa8+ Kg7 36.Bc4 Qc5
+/- (0.87) Depth: 11/26 00:00:00 570kN
31...Qd6 32.Qa5 Re8 33.Bd2 Rc8 34.Kc2 Qd4 35.h3 Qc5 36.Qc3 Rc6 37.Be3
+/- (0.87) Depth: 12/30 00:00:00 1024kN
31...Qd6 32.Qa5 Re8 33.Kc2 e4 34.fxe4 Qxh2+ 35.Kd3 Qh1 36.Qd5 Qf1+ 37.Kd2 Qg2+
+/- (0.84) Depth: 13/29 00:00:00 1758kN
31...Qd6 32.Qa5 Re8 33.Kc2 e4 34.fxe4 Qxh2+ 35.Kc3 Qe5+ 36.Qxe5 Rxe5 37.Kd4 Re8 38.b4
+/- (0.80) Depth: 14/46 00:00:01 5183kN
31...Qd6 32.Qa5 Re8 33.Kc2 e4 34.fxe4 Qxh2+ 35.Kc3 Qg3 36.Kd4 Qf3 37.Bd5 Qd1+ 38.Kc5 Qd3
+/- (0.81) Depth: 15/35 00:00:01 6965kN
31...Qd6 32.Qa5 Re8 33.Kc2 e4 34.fxe4 Qxh2+ 35.Kc3 h5 36.b4 h4 37.Qa7 Qe5+ 38.Bd4 Qe7 39.Qa2
+/- (0.84) Depth: 16/39 00:00:02 10637kN
31...Qd6 32.Kc2 e4 33.fxe4 Qxh2+ 34.Bd2 Qe5 35.Bd5 Rc8+ 36.Bc3 Qh2+ 37.Kd3 Qh3+ 38.Kd4 h5 39.b4 Qf3
+/- (0.88) Depth: 17/44 00:00:06 35826kN, tb=28
31...Qd6 32.Kc2 e4 33.fxe4 Qxh2+ 34.Bd2 Qe5 35.Bd5 Rc8+ 36.Bc3 Qh2+ 37.Kd3 Qh3+ 38.Kd4 h5 39.b4 Qf3 40.b5
+/- (0.91) Depth: 18/43 00:00:10 59496kN, tb=97
31...Qd6 32.Kc2 e4 33.fxe4 Qxh2+ 34.Bd2 Qe5 35.Bd5 Rc8+ 36.Bc3 Qh2+ 37.Kd3 Qh3+ 38.Kd4 h5 39.b4 Qf3 40.b5 h4
+/- (0.91) Depth: 19/46 00:00:19 118mN, tb=289
31...Qd6 32.Kc2 e4 33.fxe4 Qxh2+ 34.Bd2 Qe5 35.Qe3 h5 36.Bd5 h4 37.b4 Rxd5 38.exd5 Qxd5 39.Qd3 Qa2+ 40.Kd1 Qa1+ 41.Ke2
+/- (1.01) Depth: 20/51 00:00:50 336mN, tb=1149
31...Qd6 32.Kc2 Qd7 33.Kb2 Qd1 34.Qa5 Qd6 35.h4 Re8 36.Bd5 Rd8 37.Bc6 Kg7 38.Be4 f5 39.Bc2 Qd5 40.Qc3 Qd6 41.b4 Ra8
+/- (1.07) Depth: 21/62 00:04:29 2030mN, tb=14168
31...Qe7 32.Bd2 Kg7 33.Bc3 f6 34.Qe2 Qc5 35.Kc2 Rc8 36.Qd3 Qc6 37.Bd2 Qc7 38.Be3 e4 39.fxe4 Qxh2+ 40.Bd2 Qc7 41.Qd4 Rd8
+/- (1.01) Depth: 21/62 00:05:39 2549mN, tb=17345

(, 23.01.2009)

You can see that R3 starts to warm to white's position and giving it eventually a .40 evaluation ... but is still not very alarmed. On the other hand N4 evaluation already sees a +1.01 evaluation. Obviously the white bishop pair are going to be extremely strong as with the queen they can create all sort of tactical hits. Yet it seems that if white wants to win, it has to somehow find a way to queen the b pawn and that looks like a very tall order since with queens on board the white King will be very exposed to harassing checks. Also white has disconnected f and h pawns which could be weak while black has all his pawn connected. Anyway let us see what happens

Re8 (Qd6) 0.34/18 44 32.Bd2 1.07/18 13 Kg7 0.36/17 31 33.Qe2 (Qh4) 1.07/20 31 33...Rc8 (f6) 0.31/17 46 34.Bc3 (f4) 1.07/19 17 34...f6 0.38/18 17 35.Kc2 1.07/20 16 Qc5 0.38/17 18 36.Qd2 (Qd3) 1.03/21 54 36...Qa7 (Qc7) 0.50/17 42 37.Kb2 (Qd6) 1.04/19 11 37...Qb7 0.38/17 31 38.Qd3 1.04/18 9 Qc6 0.39/18 14 39.Ba5 (h4) 1.04/18 12 39...Ra8 (Qc5) 0.38/17 16 40.Bd2 1.03/20 14 Rc8 (Qc7) 0.48/18 57 41.h4 (f4) 1.08/19 24 41...Qc7 0.44/18 13 42.Bc3 (h5) 1.07/19 6 42...Rd8 (Qc6) 0.47/18 1:14 43.Qe2 (Qe4) 1.07/19 16 43...Rc8 0.41/17 14 44.Qd1 (h5) 1.07/20 16 44...Qb7 (Qc6) 0.38/16 15 45.Qf1 (Qd3) 1.07/17 10 45...Qc7 0.31/17 14 46.Kc2 1.07/19 8 Qc6 (Qc5) 0.40/16 15 47.Qh3 1.14/20 22 Ra8 (h5) 0.51/16 36

[d]r7/6kp/2q2pp1/4p3/2B4P/1PB2P1Q/2K5/8 w - - 0 48

N4 is having problems using the bishop pair to get tactical advantage ... the black queen and black rook look very active and black King seems snug behind his pawn chain. It seems impossible to make progress without allowing King harassment

48.h5 1.45/17 12

[d]r7/6kp/2q2pp1/4p2P/2B5/1PB2P1Q/2K5/8 b - - 0 48

This is a very uncomforable move for black ... a pawn exchange will allow much more scope for black's bishop pair and queen creating untenable tactical threats, this is also an opportunity for N4 to rid itself from a nagging responsibility to protect one of its weak pawns ... also for R3 not taking on the pawn seems to only make black's position worse

Ra2+ 0.58/16 12 49.Bb2 1.42/18 5 Ra8 0.60/17 14 50.h6+ 1.54/19 10

[d]r7/6kp/2q2ppP/4p3/2B5/1P3P1Q/1BK5/8 b - - 0 50

Well R3 still keeps its pawn chain alive ... but at a huge concession. The pawn protection is gone and black King loses an important hiding square. White has definetely made a lot of progress

Kf8 0.74/18 43 51.Qf1 1.53/19 10 Ke8 (Ke7) 0.78/17 11 52.Bc3 1.67/18 16 Ke7 0.81/17 13 53.Kb2 1.67/18 9 Qd6 (Rb8) 0.81/17 8 54.Qe2 (Qe1) 1.70/18 8 54...Kd7 (Kd8) 0.81/16 11 55.b4 (Qe4) 1.70/17 4

[d]r7/3k3p/3q1ppP/4p3/1PB5/2B2P2/1K2Q3/8 b - - 0 55

N4 has done all the groundwork ... double checked all possible counterplay ... and has decided it is time to push the pawns under supervision from bishop pair and queen and amazingly King as well

55...Rc8 (Ra7) 0.62/15 9 56.b5 (Bb5+) 1.69/18 14

[d]2r5/3k3p/3q1ppP/1P2p3/2B5/2B2P2/1K2Q3/8 b - - 0 56

The white pawn advance is alarmingly quick. R3 quickly moves its King to help but with a bishop pair and a queen that leaves too many tactical opportunities for white

56...Kc7 (Ra8) 0.75/16 5 57.Ba5+ (Qe4) 1.81/17 6 57...Kb8 (Kd7) 0.80/17 10 58.Kb3 (Qe4) 1.70/20 8 58...Qd4 1.04/18 27 59.Bc3 1.62/19 36 Qd6 1.24/18 35 60.Bb4 1.73/20 26 Qd7 (Qb6) 1.26/18 45 61.Qe3 1.81/19 17 Kb7 (Qd1+) 1.26/18 20 62.Bc5 1.98/18 12 Rd8 (f5) 1.26/18 13 63.Qe4+ 2.18/18 5 Kb8 2.55/19 1:38 64.b6 2.35/19 7 Rc8 2.55/19 36 65.Qe3 2.35/19 4 Rd8 2.55/19 54 66.Kb4 2.37/19 4 Qd1 2.55/18 1:20 67.Qe4 2.37/19 7 Qd7 2.55/17 56 68.Bb3 (f4) 2.57/19 4 68...Qe8 (Rc8) 3.12/14 3

[d]1k1rq3/7p/1P3ppP/2B1p3/1K2Q3/1B3P2/8/8 w - - 0 69

There is simply no way to stop the pawn advance and even with queen and rook, black has absolutely no counterplay

69.f4 (Ba4) 4.69/15 4 69...Qd7 (Rd2) 3.34/12 39 70.fxe5 5.73/14 4 Qd2+ 3.34/11 27 71.Ka4 6.68/15 8 f5 7.01/11 19 72.Qc4 6.68/15 3 Kb7 8.61/10 13 73.Qf7+ 6.68/14 23 Rd7 10.72/11 10 74.e6 7.71/14 6 1–0




(23) Rybka 3_no_cont - Naum 4 [C88]
N4_60_1_gaunt (28.1), 03.01.2009
[0.16;0.00]
Intel(R) Core(TM)2 Extreme CPU Q6850 @ 3.00GHz 3000 MHzW=16.0 ply; 226kN/s; 1,688 TBAs; HS-7moves.ctgB=17.6 ply; 3,823kN/s; 140,214 TBAs; HS-7moves.ctg 1.e4 B/0 0 e5 B/0 0 2.Nf3 B/0 0 Nc6 B/0 0 3.Bb5 B/0 0 a6 B/0 0 4.Ba4 B/0 0 Nf6 B/0 0 5.0–0 B/0 0 Be7 B/0 0 6.Re1 B/0 0 b5 B/0 0 7.Bb3 B/0 0 d6 B/0 0 8.d3 0.16/18 1:07 0–0 0.00/22 1:05 9.a4 (Bd2) 0.16/18 27 9...Bd7 (Be6) 0.00/23 4:00 10.h3 (Nc3) 0.14/18 1:21 10...Na5 -0.01/22 1:18 11.Ba2 0.15/19 50 bxa4 -0.01/22 32 12.Nc3 0.19/18 34 c5 (Rb8) 0.09/22 5:23 13.Bb1 (Bd5) 0.14/18 1:57 13...Qe8 (Nc6) 0.04/21 1:08 14.b3 0.24/19 1:23 Nc6 0.07/21 1:21 15.Nxa4 0.24/19 1:51 Qb8 0.05/22 2:46 16.c3 0.19/18 29 Qc7 0.17/21 4:19 17.Nb2 (d4) 0.20/18 59 17...d5 (Na5) 0.00/19 51 18.exd5 (Bd2) 0.24/17 54 18...Nxd5 0.00/20 25 19.Qc2 0.19/17 1:24 Bf5 -0.04/20 1:17 20.Nc4 0.15/17 1:54 Rfe8 0.00/20 1:38 21.Qb2 0.25/16 1:16 f6 0.00/20 49 22.Qc2 (Bc2) 0.18/16 50 22...g6 (h6) -0.04/19 45 23.Bd2 0.08/16 3:42 Red8 (a5) 0.00/19 43 24.Qa2 0.11/17 1:38 Kg7 -0.15/20 2:06 25.Bc2 (Nh4) 0.03/17 2:13 25...Nb6 -0.14/20 1:13 26.Nb2 0.04/16 55 a5 -0.15/20 1:01 27.Qb1 0.03/17 1:09 Rab8 (Qd7) -0.14/19 1:08 28.Qd1 0.02/16 47 Qd7 -0.16/19 38 29.Be3 (Nh4) 0.01/16 25 29...g5 (Nd5) -0.17/18 34 30.Nd2 (Qe2) 0.02/16 2:06 30...Nd5 -0.13/18 35 31.Na4 0.02/14 0 Nd4 -0.22/18 1:04 32.cxd4 0.00/15 21 cxd4 (exd4) -0.19/19 31 33.Bxg5 0.03/16 37 fxg5 -0.19/20 17 34.Rxe5 (Nc4) 0.00/16 28 34...Rb5 -0.26/19 32 35.Qf3 0.00/17 43 Nb4 -0.23/19 14 36.Rxb5 0.00/17 35 Qxb5 -0.23/6 0 37.Bb1 (Qe2) 0.00/17 3:35 37...Qe5 (Bxd3) -0.32/18 32 38.Qd1 -0.09/15 19 Re8 -0.39/19 24 39.Nf3 (Nc4) -0.33/17 2:01 39...Qf6 -0.56/19 20 40.Nd2 -0.43/17 1:55 Bd6 -0.42/19 2:09 41.Nb2 -0.68/17 5:51 Bc7 -0.42/18 1:56 42.Nbc4 -0.68/16 4:08 h5 -0.73/17 40 43.Nxa5 (Qf1) -1.01/16 2:54 43...g4 -1.08/17 27 44.Nac4 -1.05/16 2:00 gxh3 -1.42/16 12 45.Ra7 -1.05/16 25 Nd5 (Re7) -1.26/17 17 46.Ne4 -0.62/12 8 Qg6 -1.40/18 17 47.Qf3 (g3) -0.67/13 25 47...Re7 -1.77/17 26 48.Ng3 (Rxc7) -0.68/13 4 48...Re1+ -2.04/17 19 49.Kh2 -1.66/15 49 Be6 (hxg2) -2.04/18 22 50.Nb6 -1.66/13 6 hxg2 -2.50/16 30 51.Kxg2 -1.66/16 26 Re5 -2.50/18 22 52.Rxc7+ (Kh2) -1.66/15 43 52...Nxc7 -2.50/6 0 53.Qf4 (Kh2) -2.17/16 31 53...Qg5 (Qf6) -2.96/18 13 54.Qxd4 (Qxg5+) -1.31/9 3 54...Nd5 (Nb5) -3.17/18 13 55.Nxd5 -2.72/13 21 Bxd5+ -3.37/18 11 56.Kh3 (Kh2) -3.07/13 4 56...h4 (Kf7) -3.37/19 20 57.Ne4 -2.30/14 13 Qf5+ (Be6+) -3.97/18 29 58.Kg2 (Kh2) -1.34/12 10 58...h3+ -5.21/14 30 59.Kh2 (Kg3) -2.30/11 0 59...Qf4+ -7.43/15 1:11 60.Kxh3 -2.30/10 0 Kg8 -7.43/16 1:04 61.Qe3 -5.08/12 8 Rh5+ -7.43/15 57 62.Kg2 -5.08/11 0 Bxe4+ -#27/12 0 63.dxe4 (Kf1) -#25/13 0 63...Rh2+ -#25/3 0 64.Kg1 -#25/13 0 Qxe3 -#24/3 0 65.fxe3 -5.08/11 0 Rd2 -#23/3 0 66.Kf1 (b4) -#23/12 0 66...Rd1+ -#22/3 0 67.Kg2 (Ke2) -#21/11 0 67...Rxb1 -#21/6 0 68.e5 (Kf3) -#20/10 0 68...Rxb3 -#19/3 0 69.Kf2 (Kf3) -#18/0 0 69...Kf7 -#18/0 0 70.Kf3 -#17/0 0 Ke6 -#17/0 0 71.Ke4 -#16/0 0 Rb4+ -#16/0 0 72.Kd3 -#15/0 0 Kxe5 -#15/0 0 73.Kc3 -#14/0 0 Ra4 -#14/0 0 74.Kb3 -#13/0 0 Re4 -#13/0 0 75.Kc3 -#12/0 0 Rxe3+ -#12/0 0 76.Kc4 -#11/0 0 Re4+ -#11/0 0 77.Kc3 -#10/0 0 Rd4 -#10/0 0 78.Kb3 -#9/0 0 Ke4 -#9/0 0 79.Kc2 -#8/0 0 Rd3 -#8/0 0 80.Kb2 -#7/0 0 Kd4 -#7/0 0 81.Kc2 -#6/0 0 Kc4 -#6/0 0 82.Kb2 -#5/0 0 Rc3 -#5/0 0 83.Ka2 -#4/0 0 Rc2+ -#4/0 0 84.Ka1 -#3/0 0 Kb3 -#3/0 0 85.Kb1 -#2/0 0 Rc3 -#2/0 0 86.Ka1 -#1/0 0 Rc1# -#1/0 0 0–1

In this game you get to see how R3 sometimes does not evaluate the bishop pair enough ... and also R3 has difficulty evaluating the uselessness of a blocked bishop. N4 on the other hand will always evaluate the bishop pair highly and seems to have code that understands blocked bishops. I have seen quite a few games where R3 will play with almost no evaluation drop in a position where its bishop will be out of action for a long long time. N4 seems to be able to get some wins due to that.

1.e4 B/0 0 e5 B/0 0 2.Nf3 B/0 0 Nc6 B/0 0 3.Bb5 B/0 0 a6 B/0 0 4.Ba4 B/0 0 Nf6 B/0 0 5.0–0 B/0 0 Be7 B/0 0 6.Re1 B/0 0 b5 B/0 0 7.Bb3 B/0 0 d6 B/0 0 8.d3 0.16/18 1:07 0–0 0.00/22 1:05 9.a4 (Bd2) 0.16/18 27 9...Bd7 (Be6) 0.00/23 4:00 10.h3 (Nc3) 0.14/18 1:21 10...Na5 -0.01/22 1:18 11.Ba2 0.15/19 50 bxa4 -0.01/22 32 12.Nc3 0.19/18 34 c5 (Rb8) 0.09/22 5:23 13.Bb1 (Bd5) 0.14/18 1:57 13...Qe8 (Nc6) 0.04/21 1:08 14.b3 0.24/19 1:23 Nc6 0.07/21 1:21 15.Nxa4 0.24/19 1:51 Qb8 0.05/22 2:46 16.c3 0.19/18 29 Qc7 0.17/21 4:19 17.Nb2 (d4) 0.20/18 59 17...d5 (Na5) 0.00/19 51 18.exd5 (Bd2) 0.24/17 54 18...Nxd5 0.00/20 25 19.Qc2 0.19/17 1:24 Bf5 -0.04/20 1:17 20.Nc4 0.15/17 1:54 Rfe8 0.00/20 1:38 21.Qb2 0.25/16 1:16 f6 0.00/20 49 22.Qc2 (Bc2) 0.18/16 50 22...g6 (h6) -0.04/19 45 23.Bd2 0.08/16 3:42 Red8 (a5) 0.00/19 43 24.Qa2 0.11/17 1:38 Kg7 -0.15/20 2:06 25.Bc2 (Nh4) 0.03/17 2:13 25...Nb6 -0.14/20 1:13 26.Nb2 0.04/16 55 a5 -0.15/20 1:01 27.Qb1 0.03/17 1:09 Rab8 (Qd7) -0.14/19 1:08 28.Qd1 0.02/16 47 Qd7 -0.16/19 38 29.Be3 (Nh4) 0.01/16 25 29...g5 (Nd5) -0.17/18 34 30.Nd2 (Qe2) 0.02/16 2:06 30...Nd5 -0.13/18 35 31.Na4 0.02/14 0

Interesting position arises with what looks to be equality ... but N4 uncorks ...

Nd4! -0.22/18 1:04

[d]1r1r4/3qb1kp/5p2/p1pnpbp1/N2n4/1PPPB2P/2BN1PP1/R2QR1K1 w - - 0 32

Fantastic move by N4 which R3 also likes ... N4 sacs a pawn and in return will get the bishop pair ... and also it gets to inactivate white's white squared bishop for many moves.

32.cxd4 0.00/15 21 cxd4 (exd4) -0.19/19 31 33.Bxg5 0.03/16 37 fxg5 -0.19/20 17 34.Rxe5 (Nc4) 0.00/16 28

[d]1r1r4/3qb1kp/8/p2nRbp1/N2p4/1P1P3P/2BN1PP1/R2Q2K1 b - - 0 34

Rybka is quite happy in this position and views it as equal. But notice how open the position is and N4 believes it has the edge due to the having the bishop pair. Also look at white's c2 bishop ... seems like it will be out of action for a long time. In such open positions where the bishops have a lot of scope the bishop pair can be very dangerous.

34...Rb5 -0.26/19 32 35.Qf3 0.00/17 43 Nb4 -0.23/19 14 36.Rxb5 0.00/17 35 Qxb5 -0.23/6 0

[d]3r4/4b1kp/8/pq3bp1/Nn1p4/1P1P1Q1P/2BN1PP1/R5K1 w - - 0 37

Now things are really dangerous and already black has a huge advantage. N4 evaluates it as better for black ... mostly due to the bishop pair, and due to the fact white bishop will need a few tempi to get in the action. Note that even on big hardware R3 does not feel in danger with a 0.00 evaluation even after 3 minutes on Octa 4ghz

37.Bb1 (Qe2) 0.00/17 3:35 37...Qe5 (Bxd3) -0.32/18 32 38.Qd1 -0.09/15 19 Re8 -0.39/19 24 39.Nf3 (Nc4) -0.33/17 2:01 39...Qf6 -0.56/19 20 40.Nd2 -0.43/17 1:55 Bd6 -0.42/19 2:09

[d]4r3/6kp/3b1q2/p4bp1/Nn1p4/1P1P3P/3N1PP1/RB1Q2K1 w - - 0 41

OK ... things are seriously deteriorating now and black has a huge if not totally winning advantage. R3 starts to feel uncomfortable but still underevaluates the position.

41.Nb2 -0.68/17 5:51 Bc7 -0.42/18 1:56 42.Nbc4 -0.68/16 4:08 h5 -0.73/17 40

[d]4r3/2b3k1/5q2/p4bpp/1nNp4/1P1P3P/3N1PP1/RB1Q2K1 w - h6 0 43

N4 can obviously go for the d3 pawn ... but it is in attack mode and sacs an extra pawn to accelerate the attack

43.Nxa5 (Qf1) -1.01/16 2:54

[d]4r3/2b3k1/5q2/N4bpp/1n1p4/1P1P3P/3N1PP1/RB1Q2K1 b - - 0 43

R3 tries to get out of the vise it is in by taking out a pawn, but as usually happens when the opponent has a superior position ... tactics usually favor the stronger side ... and with 2 very powerful bishops and a queen still on there will be plenty of tactical opportunities

43...g4 -1.08/17 27

[d]4r3/2b3k1/5q2/N4b1p/1n1p2p1/1P1P3P/3N1PP1/RB1Q2K1 w - - 0 44

Very strong ... black opens up lines against white King.

44.Nac4 -1.05/16 2:00 gxh3 -1.42/16 12 45.Ra7 -1.05/16 25

[d]4r3/R1b3k1/5q2/5b1p/1nNp4/1P1P3p/3N1PP1/1B1Q2K1 b - - 0 45

R3 is trying to give back material to blunt black's initiative ... R3 is expecting Re7 where R3 plans to sac the rook for bishop and pick up the hanging a pawn with queen thus getting some counterplay ... but

Nd5! (Re7) -1.26/17 17

[d]4r3/R1b3k1/5q2/3n1b1p/2Np4/1P1P3p/3N1PP1/1B1Q2K1 w - - 0 46

This takes R3 totally by surprise ... N4 values greatly the bishop pair and thus will not give them up without some serious compensation. N4 anticipates that R3 will try to get rid of a bishop with a rook sac but decides that if that is going to happen the rook wills still hold a firm grip on the e file. The e file is critical in many tactical lines ... now the rook sac does not work as well and so the bishop pair stays on

46.Ne4 -0.62/12 8 Qg6 -1.40/18 17 47.Qf3 (g3) -0.67/13 25 47...Re7 -1.77/17 26 48.Ng3 (Rxc7) -0.68/13 4

[d]8/R1b1r1k1/6q1/3n1b1p/2Np4/1P1P1QNp/5PP1/1B4K1 b - - 0 48

R3 does feel a little uncomfortable and is struggling to equalize ... if only it can get that h5 pawn with check

48...Re1+ -2.04/17 19

[d]8/R1b3k1/6q1/3n1b1p/2Np4/1P1P1QNp/5PP1/1B2r1K1 w - - 0 49

Now the rook on the e file shows its usefullness

49.Kh2 -1.66/15 49

[d]8/R1b3k1/6q1/3n1b1p/2Np4/1P1P1QNp/5PPK/1B2r3 b - - 0 49

Here in analysis R3 with its superior search starts to find all sort of winning lines with hxg2! but N4 plays a funny move ... it will hold off the carnage for one more move ... no way in hell will it give up the bishop pair

Be6 (hxg2) -2.04/18 22

[d]8/R1b3k1/4b1q1/3n3p/2Np4/1P1P1QNp/5PPK/1B2r3 w - - 0 50

This is also very strong

50.Nb6 -1.66/13 6 hxg2 -2.50/16 30 51.Kxg2 -1.66/16 26 Re5 -2.50/18 22 52.Rxc7+ (Kh2) -1.66/15 43

[d]8/2R3k1/1N2b1q1/3nr2p/3p4/1P1P1QN1/5PK1/1B6 b - - 0 52

Finally R3 gets to sac its rook for one of the bishops ... but it is already way too late and positional battle has already ended ... now tactics take over

52...Nxc7 -2.50/6 0 53.Qf4 (Kh2) -2.17/16 31 53...Qg5 (Qf6) -2.96/18 13 54.Qxd4 (Qxg5+) -1.31/9 3 54...Nd5 (Nb5) -3.17/18 13 55.Nxd5 -2.72/13 21 Bxd5+ -3.37/18 11 56.Kh3 (Kh2) -3.07/13 4 56...h4 (Kf7) -3.37/19 20 57.Ne4 -2.30/14 13 Qf5+ (Be6+) -3.97/18 29 58.Kg2 (Kh2) -1.34/12 10 58...h3+ -5.21/14 30 59.Kh2 (Kg3) -2.30/11 0 59...Qf4+ -7.43/15 1:11 60.Kxh3 -2.30/10 0 Kg8 -7.43/16 1:04 61.Qe3 -5.08/12 8 Rh5+ -7.43/15 57 62.Kg2 -5.08/11 0 Bxe4+ -#27/12 0

[d]6k1/8/8/7r/4bq2/1P1PQ3/5PK1/1B6 w - - 0 63

It is all over ... N4 sees mate

63.dxe4 (Kf1) -#25/13 0 63...Rh2+ -#25/3 0 64.Kg1 -#25/13 0 Qxe3 -#24/3 0 65.fxe3 -5.08/11 0 Rd2 -#23/3 0 66.Kf1 (b4) -#23/12 0 66...Rd1+ -#22/3 0 67.Kg2 (Ke2) -#21/11 0 67...Rxb1 -#21/6 0 68.e5 (Kf3) -#20/10 0 68...Rxb3 -#19/3 0 69.Kf2 (Kf3) -#18/0 0 69...Kf7 -#18/0 0 70.Kf3 -#17/0 0 Ke6 -#17/0 0 71.Ke4 -#16/0 0 Rb4+ -#16/0 0 72.Kd3 -#15/0 0 Kxe5 -#15/0 0 73.Kc3 -#14/0 0 Ra4 -#14/0 0 74.Kb3 -#13/0 0 Re4 -#13/0 0 75.Kc3 -#12/0 0 Rxe3+ -#12/0 0 76.Kc4 -#11/0 0 Re4+ -#11/0 0 77.Kc3 -#10/0 0 Rd4 -#10/0 0 78.Kb3 -#9/0 0 Ke4 -#9/0 0 79.Kc2 -#8/0 0 Rd3 -#8/0 0 80.Kb2 -#7/0 0 Kd4 -#7/0 0 81.Kc2 -#6/0 0 Kc4 -#6/0 0 82.Kb2 -#5/0 0 Rc3 -#5/0 0 83.Ka2 -#4/0 0 Rc2+ -#4/0 0 84.Ka1 -#3/0 0 Kb3 -#3/0 0 85.Kb1 -#2/0 0 Rc3 -#2/0 0 86.Ka1 -#1/0 0 Rc1# -#1/0 0 0–1


An impressive win by Naum 4 exposing some of Rybka's very rare evaluation weaknesses.




(24) Rybka 3_0_contempt - Naum 4 [D94]
N4_60_1_7move (6.1), 30.12.2008
[0.22;0.30]
Intel(R) Core(TM)2 Quad CPU @ 2.66GHz 3192 MHzW=19.8 ply; 267kN/s; 229 TBAs; HS-7moves.ctgB=23.5 ply; 4,221kN/s; 981,946 TBAs; HS-7moves.ctg 1.d4 B/0 0 Nf6 B/0 0 2.c4 B/0 0 g6 B/0 0 3.Nc3 B/0 0 d5 B/0 0 4.Nf3 B/0 0 c6 B/0 0 5.e3 B/0 0 Bg7 B/0 0 6.Be2 B/0 0 0–0 B/0 0 7.0–0 B/0 0 Nbd7 B/0 0 8.cxd5 0.22/18 57 cxd5 0.30/22 2:59 9.Qb3 (h3) 0.30/20 1:56 9...e6 (Nb6) 0.33/20 1:16 10.Bd2 (Bd3) 0.31/17 51 10...Ne4 0.23/21 1:20 11.Rad1 (Be1) 0.31/18 1:11 11...Nxd2 0.21/21 34 12.Rxd2 0.35/19 32 a6 (Qb6) 0.22/22 59 13.Rc2 0.41/19 5:15 b6 (Qa5) 0.24/23 5:21 14.Rfc1 0.31/19 1:52 Bb7 0.21/22 1:41 15.a3 (Na4) 0.29/18 1:39 15...Rc8 0.11/19 57 16.h3 (g3) 0.25/17 2:43 16...Bf6 (Re8) 0.03/21 43 17.Ne1 (Na2) 0.25/18 1:25 17...Qe7 0.00/22 1:06 18.Nd3 0.23/18 1:08 Qd6 0.00/21 1:38 19.Nb4 (a4) 0.22/18 1:28 19...b5 (Nb8) 0.00/21 25 20.a4 (Nd3) 0.22/18 48 20...Be7 (a5) -0.12/20 45 21.axb5 0.33/19 1:26 Qxb4 -0.16/20 46 22.Qxb4 0.33/18 0 Bxb4 -0.16/6 0 23.bxa6 0.33/18 0 Ba8 -0.14/22 2:22 24.Na4 (a7) 0.18/19 2:04 24...Rxc2 -0.05/22 1:11 25.Rxc2 0.18/18 0 Nf6 0.00/23 3:30 26.a7 (f3) 0.19/18 1:37 26...Ne4 -0.16/23 45 27.Rc1 (Bd3) 0.00/18 42 27...Rd8 (Nd6) -0.45/20 46 28.b3 (Ba6) -0.17/18 1:26 28...Kg7 (Kf8) -0.73/22 52 29.Ra1 -0.15/18 18 Nc3 (Rd6) -0.72/24 1:13 30.Nxc3 -0.28/20 46 Bxc3 -0.72/6 0 31.Rc1 -0.86/20 9:03 Ba5 -0.81/24 52 32.Ra1 -1.03/20 5:04 Bb4 (Bc7) -0.81/23 2:53 33.Ra4 (Bb5) -0.29/18 15 33...Be7 -1.05/22 1:10 34.b4 -0.59/18 38 Rc8 -1.19/23 2:24 35.g3 -0.67/19 34 Kf8 -1.23/23 48 36.Ra1 (Kg2) -0.80/19 45 36...Ke8 -1.29/22 21 37.Ba6 -0.80/19 8 Rc3 -1.29/22 12 38.Rb1 -0.83/19 18 Bd8 (Rc7) -1.27/22 36 39.Ra1 (Bb5+) -0.96/17 22 39...Ke7 (Bb6) -1.27/23 51 40.Be2 (Bb5) -1.15/17 39 40...Rb3 (Rc2) -1.55/22 40 41.Ra4 -1.31/19 29 Bc7 (Bb6) -1.75/21 31 42.Kg2 (h4) -1.69/16 22 42...Bd6 -2.18/20 19 43.b5 -1.94/17 26 Kd7 -2.13/22 12 44.Kf3 -2.08/17 12 Kc7 -2.23/23 15 45.Kg4 -2.04/19 13 Rb2 -2.24/22 12 46.Kf3 -2.65/20 1:59 e5 -2.44/22 24 47.Bf1 -2.85/20 1:35 Rb4 -2.67/21 18 48.Rxb4 -2.90/19 41 e4+ -2.73/24 10 49.Ke2 (Kg2) -2.90/21 4 49...Bxb4 -2.74/28 25 50.f3 -3.09/22 12 exf3+ -2.81/27 28 51.Kxf3 -3.09/23 21 Kb6 -2.81/28 15 52.Kf4 (h4) -3.10/23 9 52...Bd6+ (Kxa7) -3.04/26 15 53.Kf3 -3.39/21 8 Kxa7 -3.11/27 9 54.Kg2 (g4) -3.58/23 20 54...Kb6 (Bb7) -3.34/27 19 55.Kf2 -3.67/21 6 Bb7 -3.41/28 28 56.Be2 (g4) -3.86/23 27 56...Bc8 -3.55/27 25 57.h4 (g4) -3.92/23 19 57...h5 (Bf5) -3.93/26 13 58.Bf1 -4.18/21 5 Bf5 (Be6) -3.93/28 15 59.Bg2 (Be2) -4.18/24 5 59...Be4 -3.93/28 33 60.Bf1 -4.18/26 2 Bb1 (Bb8) -3.93/29 16 61.Be2 -4.18/23 5 f5 (Bf5) -4.13/30 41 62.Bf3 (Bf1) -5.12/22 1:05 62...Ba2 -5.46/24 13 63.Bd1 (Be2) -5.12/20 7 63...Kxb5 -5.46/27 14 64.Be2+ (Bf3) -5.12/18 5 64...Bc4 -5.46/23 27 65.Bd1 -5.67/17 5 Kb4 -5.46/22 1:07 66.Bc2 -5.43/17 4 Kc3 (Bb5) -5.47/21 54 67.Ba4 -6.39/13 7 Kd2 -6.99/21 55 68.Be8 -6.39/12 8 Bxg3+ -9.52/20 9 69.Kf3 -6.81/14 5 Bxh4 -5.47/18 15 70.Bxg6 -6.83/14 3 Bg5 -5.47/17 8 71.Bxf5 (Kg2) -6.88/14 3 71...Bxe3 -5.47/17 8 0–1

Here is another game by Naum 4 showing the power of the bishop pair


1.d4 B/0 0 Nf6 B/0 0 2.c4 B/0 0 g6 B/0 0 3.Nc3 B/0 0 d5 B/0 0 4.Nf3 B/0 0 c6 B/0 0 5.e3 B/0 0 Bg7 B/0 0 6.Be2 B/0 0 0–0 B/0 0 7.0–0 B/0 0 Nbd7 B/0 0 8.cxd5 0.22/18 57 cxd5 0.30/22 2:59 9.Qb3 (h3) 0.30/20 1:56 9...e6 (Nb6) 0.33/20 1:16 10.Bd2 (Bd3) 0.31/17 51 10...Ne4 0.23/21 1:20 11.Rad1 (Be1) 0.31/18 1:11

[d]r1bq1rk1/pp1n1pbp/4p1p1/3p4/3Pn3/1QN1PN2/PP1BBPPP/3R1RK1 b - - 0 11

Note here that N4 was expecting Be1 preserving Rybka's bishop pair ... but to be honest I don't agree with that as Be1 looks very artificial. Still it gives a great insight on how different evaluations can affect play. Rybka plays a developing move instead

11...Nxd2 0.21/21 34 12.Rxd2 0.35/19 32

[d]r1bq1rk1/pp1n1pbp/4p1p1/3p4/3P4/1QN1PN2/PP1RBPPP/5RK1 b - - 0 12

Both sides seem content with their position. And now play continues normally

a6 (Qb6) 0.22/22 59 13.Rc2 0.41/19 5:15 b6 (Qa5) 0.24/23 5:21 14.Rfc1 0.31/19 1:52 Bb7 0.21/22 1:41 15.a3 (Na4) 0.29/18 1:39 15...Rc8 0.11/19 57 16.h3 (g3) 0.25/17 2:43 16...Bf6 (Re8) 0.03/21 43 17.Ne1 (Na2) 0.25/18 1:25 17...Qe7 0.00/22 1:06 18.Nd3 0.23/18 1:08 Qd6 0.00/21 1:38 19.Nb4 (a4) 0.22/18 1:28 19...b5 (Nb8) 0.00/21 25 20.a4 (Nd3) 0.22/18 48

[d]2r2rk1/1b1n1p1p/p2qpbp1/1p1p4/PN1P4/1QN1P2P/1PR1BPP1/2R3K1 b - - 0 20

Here R3 has just played a4 and is expecting a5 by black driving the Knight back but N4 plays differently

20...Be7! (a5) -0.12/20 45

[d]2r2rk1/1b1nbp1p/p2qp1p1/1p1p4/PN1P4/1QN1P2P/1PR1BPP1/2R3K1 w - - 0 21

This turns out to be very strong, but R3 actually doesn't mind at all giving the Knight up for what looks like very strong compensation of 2 passed pawns

21.axb5 0.33/19 1:26 Qxb4 -0.16/20 46 22.Qxb4 0.33/18 0 Bxb4 -0.16/6 0 23.bxa6 0.33/18 0

[d]2r2rk1/1b1n1p1p/P3p1p1/3p4/1b1P4/2N1P2P/1PR1BPP1/2R3K1 b - - 0 23

OK look at the position for a minute or two. Both engines are very happy with their position ... R3 highly values passed pawns and N4 highly values the bishop pair. Both have evaluations that tell them they are ahead ... play continues

Ba8 -0.14/22 2:22 24.Na4 (a7) 0.18/19 2:04 24...Rxc2 -0.05/22 1:11 25.Rxc2 0.18/18 0 Nf6 0.00/23 3:30 26.a7 (f3) 0.19/18 1:37 26...Ne4 -0.16/23 45 27.Rc1 (Bd3) 0.00/18 42 27...Rd8 (Nd6) -0.45/20 46 28.b3 (Ba6) -0.17/18 1:26 28...Kg7 (Kf8) -0.73/22 52 29.Ra1 -0.15/18 18

[d]b2r4/P4pkp/4p1p1/3p4/Nb1Pn3/1P2P2P/4BPP1/R5K1 b - - 0 29

When you see such a move by R3 you know things are not going well. R3 gives up the c file ... apparently in search it sees problems in pawn vulnerability ... in the meantime N4 bishop pair are extremely active and if pieces are now exchanged the a and b pawns will be no match against the bishop pair

Nc3 (Rd6) -0.72/24 1:13 30.Nxc3 -0.28/20 46 Bxc3 -0.72/6 0

[d]b2r4/P4pkp/4p1p1/3p4/3P4/1Pb1P2P/4BPP1/R5K1 w - - 0 31

Knights are off and thus black's initiative accelerates

31.Rc1 -0.86/20 9:03 Ba5 -0.81/24 52 32.Ra1 -1.03/20 5:04 Bb4 (Bc7) -0.81/23 2:53 33.Ra4 (Bb5) -0.29/18 15

[d]b2r4/P4pkp/4p1p1/3p4/Rb1P4/1P2P2P/4BPP1/6K1 b - - 0 33

R3 desperately wants to show that it has compensation by moving the b pawn to link up with the a pawn. In the meantime N4 keeps its cool and simply improves its position by placing the black King closer to the a and b pawns. If rooks go off then this is quickly game over

33...Be7 -1.05/22 1:10 34.b4 -0.59/18 38 Rc8 -1.19/23 2:24 35.g3 -0.67/19 34 Kf8 -1.23/23 48 36.Ra1 (Kg2) -0.80/19 45 36...Ke8 -1.29/22 21 37.Ba6 -0.80/19 8 Rc3 -1.29/22 12 38.Rb1 -0.83/19 18 Bd8 (Rc7) -1.27/22 36 39.Ra1 (Bb5+) -0.96/17 22 39...Ke7 (Bb6) -1.27/23 51 40.Be2 (Bb5) -1.15/17 39 40...Rb3 (Rc2) -1.55/22 40 41.Ra4 -1.31/19 29

[d]b2b4/P3kp1p/4p1p1/3p4/RP1P4/1r2P1PP/4BP2/6K1 b - - 0 41

Position is going from bad to worse ... white rook ... whites strongest piece is now relegated to try to protect the pawns. White is simply helpless and has absolutely no counter play

Bc7 (Bb6) -1.75/21 31 42.Kg2 (h4) -1.69/16 22 42...Bd6 -2.18/20 19 43.b5 -1.94/17 26 Kd7 -2.13/22 12 44.Kf3 -2.08/17 12 Kc7 -2.23/23 15 45.Kg4 -2.04/19 13 Rb2 -2.24/22 12 46.Kf3 -2.65/20 1:59 e5 -2.44/22 24 47.Bf1 -2.85/20 1:35 Rb4 -2.67/21 18 48.Rxb4 -2.90/19 41

[d]b7/P1k2p1p/3b2p1/1P1pp3/1R1P4/4PKPP/5P2/5B2 b - - 0 48

N4 forces the exchange of rooks and black bishop pair is still on ... so now it is game over

e4+ -2.73/24 10 49.Ke2 (Kg2) -2.90/21 4 49...Bxb4 -2.74/28 25 50.f3 -3.09/22 12 exf3+ -2.81/27 28 51.Kxf3 -3.09/23 21 Kb6 -2.81/28 15 52.Kf4 (h4) -3.10/23 9 52...Bd6+ (Kxa7) -3.04/26 15 53.Kf3 -3.39/21 8 Kxa7 -3.11/27 9 54.Kg2 (g4) -3.58/23 20 54...Kb6 (Bb7) -3.34/27 19 55.Kf2 -3.67/21 6 Bb7 -3.41/28 28 56.Be2 (g4) -3.86/23 27 56...Bc8 -3.55/27 25 57.h4 (g4) -3.92/23 19 57...h5 (Bf5) -3.93/26 13 58.Bf1 -4.18/21 5 Bf5 (Be6) -3.93/28 15 59.Bg2 (Be2) -4.18/24 5 59...Be4 -3.93/28 33 60.Bf1 -4.18/26 2 Bb1 (Bb8) -3.93/29 16 61.Be2 -4.18/23 5 f5 (Bf5) -4.13/30 41 62.Bf3 (Bf1) -5.12/22 1:05 62...Ba2 -5.46/24 13 63.Bd1 (Be2) -5.12/20 7 63...Kxb5 -5.46/27 14 64.Be2+ (Bf3) -5.12/18 5 64...Bc4 -5.46/23 27 65.Bd1 -5.67/17 5 Kb4 -5.46/22 1:07 66.Bc2 -5.43/17 4 Kc3 (Bb5) -5.47/21 54 67.Ba4 -6.39/13 7 Kd2 -6.99/21 55 68.Be8 -6.39/12 8 Bxg3+ -9.52/20 9 69.Kf3 -6.81/14 5 Bxh4 -5.47/18 15 70.Bxg6 -6.83/14 3 Bg5 -5.47/17 8 71.Bxf5 (Kg2) -6.88/14 3 71...Bxe3 -5.47/17 8 0–1

A simple technical win for N4 showing the power of the bishop pair.


Here is another similar position from a game by N4 and R3 again showing power of bishop pair in an open position. In this case material is equal yet the long term effectiveness of the bishop pair is obvious

[d]2b5/1p3k2/p1pB1p2/P1P5/4P2r/3B3p/6p1/R5K1 w - - 0 84

Take some time to look at this position. R3 looks like it has some very dangerous passed pawns but N4 has the bishop pair. Which is stronger and which side would you prefer to play ... here play continued

84.Bc2 0.26/19 10 Rh8 (Rh5) -0.17/20 18 85.Re1 0.26/21 13 Rh6 (Rh7) -0.17/20 10 86.Rb1 (Re3) 0.27/20 6 86...Kg6 (Rh4) -0.17/21 15 87.Rb3 0.27/21 7 Rh5 0.00/20 26 88.Bd1 (Rg3+) 0.27/20 5 88...Rh4 0.00/19 14 89.Kh2 (Rg3+) 0.40/20 19 89...Kf7 (Rh7) 0.00/19 10 90.Re3 0.46/20 6 Bd7 (Rh6) 0.07/19 32 91.Bg3 0.46/20 4 Rh7 0.31/19 30 92.Rd3 (e5) 0.46/20 18 92...Ke8 (Ke7) 0.14/18 20 93.Bf2 (e5) 0.61/21 5 93...Rg7 0.43/20 21 94.Rg3 (Bh5+) 0.61/21 3 94...Kf8 (Rg5) 1.39/21 1:57 95.Bh5 1.08/21 7 Be6 (Rg5) 1.53/22 42 96.Bg6 1.56/21 10 Rd7 1.73/22 1:06 97.Bf5 1.57/22 5 Rd2 (Bxf5) 1.73/22 47 98.Bxe6 1.75/20 7 Rxf2 1.73/23 21 99.Rg8+ (Rxh3) 1.71/21 11 99...Ke7 1.73/21 0 100.Bxh3 1.83/23 12 Rb2 (Ra2) 1.62/17 5 101.Rg7+ 1.83/22 4 Kf8 1.81/20 8 102.Rh7 1.83/22 5 Rc2 (Rb5) 1.88/19 10 103.Rxb7 1.98/20 5 Rxc5 2.05/19 5 104.Kxg2 2.01/20 4 Rxa5 2.21/19 11 105.Be6 (Rb6) 2.01/21 3 105...Ra3 (Re5) 2.21/16 8 106.Rf7+ (Kf2) 2.30/19 9 106...Ke8 2.22/17 5 107.Kf2 2.31/20 5 Rc3 2.22/19 8 108.Rxf6 (Ra7) 2.36/20 4 108...Ke7 2.54/16 6 109.Rh6 2.56/20 6 Kd6 2.74/16 7 110.Bg4+ 2.95/19 8 Ke5 3.29/18 28 111.Bf3 2.96/19 4 Rc4 (Kf4) 3.29/17 10 112.Rh5+ 3.08/19 3 Kd6 3.36/18 5 113.Ra5 3.16/19 3 Rb4 (Rc5) 4.40/18 53 114.e5+ (Rxa6) 3.48/18 3 114...Ke6 (Kd7) 4.06/16 2 115.Bxc6 3.98/18 18 Rb6 4.32/17 2 116.Rc5 3.98/20 5 a5 4.52/17 3 117.Ba4 (Bb5) 3.98/20 6 117...Ra6 (Rb8) 4.63/18 4 118.Ke3 3.99/20 3 Rb6 (Ra8) 5.12/18 17 119.Rxa5 (Bb5) 3.99/20 4 119...Rb2 5.12/17 3 120.Bc6 3.99/20 5 Rb4 (Rb6) 5.12/16 5 121.Be4 (Bg2) 3.99/21 4 121...Rb8 5.12/13 5 122.Rc5 (Bg2) 3.99/19 4 122...Rd8 5.12/14 4 123.Bd3 (Bg2) 3.99/19 4 123...Rb8 (Rh8) 5.12/13 2 124.Bc4+ 6.67/16 2 Kf5 (Kd7) 6.22/15 18 125.e6+ 17.93/16 5 Kf6 (Kg6) 6.03/12 2 126.Kf4 (Ke4) 12.70/18 4 126...Kg6 (Rh8) 9.44/13 13 127.Rc7 (Ke5) #23/15 1 127...Rf8+ #23/12 8 128.Rf7 (Ke5) #22/6 0 128...Rd8 #22/8 0 129.Bd3+ #21/4 0 Rxd3 #20/3 0 130.Rf8 #20/0 0 Rd4+ #19/0 0 131.Ke5 #19/0 0 Rd1 #18/0 0 132.Rg8+ #18/0 0 Kh6 #17/0 0 133.Rg4 #17/0 0 Kh5 #16/0 0 134.Rd4 #16/0 0 Re1+ #15/0 0 135.Re4 #15/0 0 Rg1 #14/0 0 136.e7 #14/0 0 Rg8 #13/0 0 137.Kf5 #13/0 0 Rg5+ #12/0 0 138.Kf6 #12/0 0 Rg8 #11/0 0 139.e8Q+ #11/0 0 Rxe8 #10/0 0 140.Rxe8 #10/0 0 Kg4 #9/0 0 141.Ke5 #9/0 0 Kf3 #8/0 0 142.Kf5 #8/0 0 Kg3 #7/0 0 143.Re3+ #7/0 0 Kf2 #6/0 0 144.Kf4 #6/0 0 Kg2 #5/0 0 145.Rf3 #5/0 0 Kh2 #4/0 0 146.Rf2+ #4/0 0 Kh1 #3/0 0 147.Kf3 #3/0 0 Kg1 #2/0 0 148.Kg3 #2/0 0 Kh1 #1/0 0 149.Rf1# #1/0 0 1–0

While still no match for R3 ... I think N4 has shown that R3 can benefit from a slight modification of its bishop pair evaluation in open positions.

[jdart]

That's interesting. There are a lot of positional factors and the material imbalance (bishop pair) is only one of them. It's possible that that alone is not a very large factor but instead bishop mobility is valued highly (in several of these games the winning side not only had the two bishops but also greater mobility).

In fact Rybka is famous for evaluating imbalanced material accurately. So it may or may not benefit from tweaking that factor. Also, 2 bishops can be better than a Rook in the middlegame but Dvoretsky says the advantage if any is much smaller in the endgame.

--Jon

[mclane]

"accurately"
is IMO the wrong word.

material imbalance situations are not static.

different positions have different dynamics.
if you evaluate it statically you make it wrong.
you have to do it dynamically. IMO.

[Albert Silver]

"accurately"
is IMO the wrong word.

material imbalance situations are not static.

different positions have different dynamics.
if you evaluate it statically you make it wrong.
you have to do it dynamically. IMO.

Jon didn't say anything about the evaluation being static.

[Albert Silver]

Must have taken you ages to put that post together. Very interesting, and much appreciated. Thanks!

In the last position, I think black is in big trouble. The passed pawns are blocked, the white-squared bishop is a prisoner protecting the q-side pawns, and he is a piece down. I cannot imagine a single master or GM thinking this was remotely equal.

A fantastic book dealing with precisely this type of queenless middlegame with imbalanced material was published last year by Glenn Flear, "Practical Endgame Play - beyond the basics". It is a masterpiece and I cannot recommend it enough. Basically, he takes positions from modern GM play, down to 2 pieces each or so, plenty of pawns, and no queens, and analyzes the goals of each side. Some are balanced, such as Rook plus Bishop vs. Rook and knight with pawns, and some not, such as Rook vs. two minor pieces. The book, with superb production values, is organized according to the piece configurations, and Flear does a wonderful job explaining the plans.

Albert

[M ANSARI]

It is difficult to evaluate material imbalance accurately ... because something that works well in one position will not work well in another. Rybka 3 is legendary in how good it is in handling such positions, with incredibly accurate static evaluations in very unbalanced positions. However there seems to be room for improvement in some areas which has become more obvious now that a strong program such as N4 is available. Rybka 3's incredible search advantage against other engines probably still masks some areas of vulnerability ... where it can draw or even reverse a lost game simply by searching better than its opponent. The bishop pair in open positions where bishops have a lot of scope is one obvious area where Rybka 3 can improve ... other less obvious area is when Rybka 3 should or shouldn't accept 3 pawns for a piece. I expect R4 to improve on some of those issues.