click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (32-ply):

1. = (0.00): 22...Nc4 23.Rxe8+ Qxe8 24.Qc1 Qd7 25.Re1 Qb7 26.b3 Nxa3 27.Qc5 b4 28.Bf2 Nd7 29.Qc1 Nb5 30.Re8+ Nf8 31.Nxb5 Qxb5 32.Rxf8+ Kxf8 33.Qc8+ Ke7 34.Bc5+ Kf6 35.Bd4+ Ke7 36.Bc5+

2. = (0.12): 22...Nc6 23.Rxe8+ Nxe8 24.Nf5 Qd7 25.Qd3 Rb6 26.Re1 b4 27.a4 b3 28.a5 Rb7 29.Rxe8+ Qxe8 30.Nd6 Qd7 31.Nxb7 Qxb7 32.Qxd5 Qb4 33.h3 Nxa5 34.Qa8+ Kh7 35.Qxa7 f6 36.Kh2 Qb5 37.Qe3 Qc4 38.Be1

3. ² (0.60): 22...Ng6 23.Nxb5 Rae6 24.Rxe6 fxe6 25.Qe3 Qd7 26.Nc7 Rb8 27.Re1 Rxb2 28.Nxe6 Qb7 29.Bf2 Rb1 30.Nd4 Rxe1+ 31.Bxe1 Qc7 32.Bf2 Kh7 33.Bg3 Qb6 34.Be1 Qb8 35.Bb4 Qe5 36.Qd3 Qc7 37.Nb3 Qe5 38.g3

click for larger view

The unanimous engine choice is 23...Qd8. As <tbentley> indicated, after the only other reasonable queen move, 23...Qd6, loses the knight after 24.Nb4 Ra5 25.f4. 23...Qd8 is necessary to relieve the pin on the Nd5.

<Critter 1.2, PV # 1, [0.00], d=27>: 23...Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.f4 Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rd8+ Kh7 31.Rf2 Rc1+ 32.Kh2 Rc7 33.Rfd2 Rce7 34.f5 g5 35.fxg6+ fxg6 36.R8d7 Kg7 37.<Rxe7+>

click for larger view

<37.Rxe7+> I can't conceive of anything but a draw after 37...Rxe7. And I don't have anything to say about the moves by either side leading up to this point.

<Komodo 3.0, PV # 1, [-0.02], d=28>: 23...Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.<Ne3> Qxd4 27.Rxd4 Nc6 28.Rd3 Ne5 29.Rd2 Rcc8 30.Kf2 g6 31.Rfd1 Kg7 32.Kg3 h5 33.Nd5 Nc4 34.Rc2 <b5> 35.<Kh3> Rcd8 36.Nc7 Rxd1 37.Nxe8+ Kh7 38.Nc7 Nd6 39.b3 <Kg7>

click for larger view

<26.Ne3> A deviation from the "normal" 26.f4 but it doesn't seem to be an improvement.

<34...b5> This weakening of the q-side pawns doesn't seem like a good idea. I don't know if Black gets into trouble by 34...Nd6, inviting more exchanges.

<35.Kh3> Moving further from the center doesn't seem like a good idea either. But hard to think of a plan for White other than trying to place a rook on the 7th rank.

<39...Kg7> Doesn't seem to be anything other than a draw here either.

<Spark 1.0, PV #1, [+0.06], d=26>: 23...Qd8 24.Qxd4 Ra5 25.Rd1 <Qd6> 26.Rf2 Nc6 27.Qd3 Re5 28.Nc3 Re1+ 29.Rf1 Qxd3 30.Rxd3 Rxf1+ 31.Kxf1 Ne5 32.Rd4 Rc5 33.<Kg1> f6 34.Kf2 Rc8 35.Kg3 Kh7 36.Kf4 <Kg6>

click for larger view

<25...Qd6> This Qd8-d6 seems to lose a tempo. Since White's 24th and 25th moves look forced, if Black intended ...Qd6, why not 23...Q6 directly?

<33.Kg1> Again, this decentralization and waste of time. Why not 33.Kf2 directly?

<36...Kg6> No surprises here either. Pretty even, although Black should have tried 35...Kf7-e6.

<Spike 1.4, PV #1, [+0.09], d=33>: 23...Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.f4 Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 Re1+ 31.Kh2 Rc7 32.f5 Re5 33.<g4> Kh7 34.Rfd2 Rce7 35.Rd7 Rxd7 36.Rxd7 Re2+ 37.Kg3 a5 38.Rxf7 Rxb2 39.Kh4 Ra2 40.Rb7 Rxa3 41.<Rxb6>

<33.g4> Spike is being its usual aggressive self.

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (34-ply):

1. = (0.00): 25...Rc5 26.f4 Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 Rc1+ 31.Kh2 Rc7 32.Rfd2 g6 33.Kg1 Kg7 34.Kf2 Rce7 35.Rd6 Rb3 36.R6d4 Rbe3

2. = (0.00): 25...Nc6 26.Qd3 Ne5 27.Qe4 Qc8 28.Rfe1 Re6 29.f4 Ng6 30.Qd4 Rxe1+ 31.Rxe1 Rc5 32.Ne3 Ne7 33.Qe4 Rc1 34.Kf2 Rxe1 35.Kxe1 Qc1+ 36.Nd1 Kf8 37.Qa8+ Nc8 38.Qd5 Ke7 39.Qb7+ Ke6 40.Qe4+ Kd7 41.Qf5+

3. = (0.00): 25...Kf8 26.f4 Ng4 27.h3 Qxd5 28.Qxd5 Rxd5 29.Rxd5 Ne3 30.Rd7 Nxf1 31.Kxf1 Re3 32.Rxa7 Rb3 33.Kf2 g6 34.a4 Rxb2+ 35.Kf3 h5 36.f5 gxf5 37.a5 bxa5 38.Rxa5 h4 39.Rxf5 Rb3+ 40.Kf4 Rb2 41.Kf3

4. = (0.08): 25...Qd6 26.Ne3 Qxd4 27.Rxd4 Nc6 28.Rd3 Ne5 29.Rd2 Rc5 30.Kf2 Nc4 31.Nxc4 Rxc4 32.Rfd1 g6 33.g4 Kg7 34.Kg3 Rc7 35.h4 Kf6 36.Rd7 Rxd7 37.Rxd7 a6 38.Kf2 Re6 39.Rc7 Rd6 40.b3 Ke6 41.Ke3

5. = (0.20): 25...Re6 26.f4 Ng4 27.Nc3 Qxd4+ 28.Rxd4 Ne3 29.Rf3 g6 30.h3 Kg7 31.g4 Rc5 32.Kf2 Nc4 33.f5 gxf5 34.gxf5 Ree5 35.b4 Rc6 36.Ne4 Kf8 37.Nc5 bxc5 38.Rxc4 Ke7 39.Re3 Rxe3

click for larger view

Sorry for the lack of diagrams and commentary but the time for our next vote is getting close and I will be busy the rest of the day so I won't have a chance to look at these lines (at least the top line) until tomorrow morning. If we decide on 22...d4 then I'll go back and post the results of my review.

1. [0.00], d=29: 23.Nd5 Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.h3 Qd6 27.Qa4 Nc6 28.Nc3 Qe6 29.Qg4 Qxg4 30.hxg4 Ne5 31.Rf2 Kf8 32.Re2 Rcc8 33.Kf2 Nc4 34.Ne4

2. [-0.16], d=29: 23.Ne4 Qd8 24.f4 Nc4 25.Qe2 Ne3 26.Qxa6 Rxe4 27.Rf2 Qd5 28.Qd3 f5 29.g3 Qe6 30.Rc1 Kh7 31.Re2 h5 32.h4 Qd5 33.Rce1 Qc6

3. [-0.51], d=29: 23.f4 dxc3 24.Qxc3 Ra5 25.fxe5 Qe6 26.Qf3 Rd5 27.Re2 a5 28.h3 Re7 29.Rfe1 Red7 30.Qe3 Rd8 31.Rf1 a4 32.Rc1

4. [-1.16], d=29: 23.Nd1 Rd8 24.Nf2 Ra5 25.Re4 Nc4 26.Qe2 Rc5 27.Rd1 Rcc8 28.Ng4 Qc6 29.Ne5 Nxe5 30.Rxe5 Qd6 31.Kf2 Rc2 32.Qxc2 Qxe5 33.Qe4 Qd6 34.Rd3 a5 35.Kg1

click for larger view

No review, diagrams, or commentary. See my comment in my Critter 1.2 lines post above.

1. [+0.03], d=29: 23.Nd5 Qd8 24.Qxd4 Ra5 25.Rd1 Qd6 26.Rd2 Nc6 27.Qf4 Qxf4 28.Nxf4 Rc5 29.Nd5 Ne5 30.b3 Rcc8 31.Rfd1 f6 32.a4 Kf7 33.f4 Ng4 34.h3 Ne3 35.Nxe3 Rxe3 36.Rd7+ Re7 37.f5 Ke8 38.R7d2 Rc3

2. [-0.16], d=29: 23.Ne4 Qd8 24.f4 Ng4 25.Qe2 Ra5 26.Qxg4 Rxe4 27.Rxe4 f5 28.Qf3 fxe4 29.Qxe4 d3 30.f5 Rc5 31.f6 Rc7 32.Rd1 d2 33.Qe3 Rd7 34.Qb3+ Kh8 35.f7 Qf8 36.Rf1 Qc5+ 37.Kh1 Rd8 38.Qf3 Rf8 39.Qd3

3. [-0.70], d=29: 23.f4 dxc3 24.Qxc3 Ra5 25.fxe5 Qe7 26.Qf3 Qc7 27.e6 fxe6 28.b4 Rf5 29.Qxf5 exf5 30.Rxe8+ Kh7 31.Rfe1 Qc4 32.R8e3 Qd4 33.g3 Kg6 34.Kg2 Kg5 35.R1e2 g6 36.h4+ Kg4 37.Re6 Qd5+ 38.Kh2 Kh5 39.Re7 a6 40.R2e6

4. [-1.41], d=29: 23.Nb1 Rd8 24.Qe2 Ra5 25.f4 Ng6 26.g3 Qf5 27.b4 Rad5 28.Qe4 Qxe4 29.Rxe4 Rc8 30.Rd1 f5 31.Ree1 Rc2 32.Nd2 Rc3 33.a4 Kf7 34.Kg2 Rc2 35.h4 h5 36.Kf2 d3 37.Re3 Rd4

click for larger view

No review, diagrams, or commentary. See my comment in my Critter 1.2 lines post above.

1. [+0.05], d=35: 23.Nd5 Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.f4 Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 Rc1+ 31.Kh2 Rc7 32.Rfd2 Rce7 33.f5 f6 34.Kg1 h5 35.Kf2 R3e5 36.g4 Kf7 37.Rd7 hxg4 38.hxg4 Re4 39.Rxe7+ Kxe7 40.Rc2 Kd6 41.Kf3 Re1 42.b4 b5

2. [-0.12], d=35: 23.Ne4 Qd8 24.f4 Ng4 25.Qe2 f5 26.Nf6+ Nxf6 27.Qxa6 Rxe1 28.Rxe1 d3 29.Qc4+ Kh7 30.g3 d2 31.Rd1 Qd7 32.Kf1 Ne4 33.Ke2 Qe8 34.Kf1 a6 35.Qd3 a5 36.Rxd2 Nxd2+ 37.Qxd2 Qc6 38.Kf2 Qh1 39.Ke3 Qe4+ 40.Kf2 a4 41.Qd1 Qc4 42.Qd2 Qc6 43.Qe2 Qh1 44.Ke3 Qe4+ 45.Kd2 Qc6 46.Qe5 Kg6 47.Kd3 Qf3+ 48.Qe3 Qd5+ 49.Qd4 Qb3+ 50.Ke2

3. [-0.92], d=35: 23.f4 dxc3 24.Qxc3 Ra5 25.b4 Rd5 26.fxe5 Qe6 27.Re4 Rxe5 28.Qxe5 Qxe5 29.Rxe5 Rxe5 30.Rf3 a5 31.Rb3 f5 32.Kf2 axb4 33.Rxb4 b5 34.Rd4 Rc5 35.Rd7 Rc2+ 36.Kg1 Rb2 37.Rb7 Kh7 38.a4 b4 39.a5 Kg6 40.a6 Ra2 41.Rxb4 Rxa6 42.Kf2 h5 43.Kg3 Ra3+ 44.Kf4

4. [-1.58], d=34: 23.Nd1 Rd8 24.Re4 b5 25.Rfe1 Re6 26.b3 g5 27.h4 gxh4 28.Nf2 Kh7 29.f4 Ng6 30.Rxe6 fxe6 31.Re4 Rd7 32.Qd3 Rd5 33.a4 bxa4 34.bxa4 a5 35.Qc4 Qf7 36.Qc2 Kg7 37.Nd3 Qf5 38.Qe2 Rd6 39.Qf3 Rd7 40.Ne5 Rd5 41.Nc6

click for larger view

No review, diagrams, or commentary. See my comment in my Critter 1.2 lines post above.

1. [-0.03], d=30: 23.Nd5 Qd8 24.Qxd4 Ra5 25.Rd1 Rc5 26.h3 Rc4 27.Qf2 Rc5 28.Ne3 Nd3 29.Qd2 Qg5 30.Qxd3 Rxe3 31.Qd8+ Kh7 32.Qxg5 hxg5 33.Rf2 Re7 34.Rd6 f6 35.Rfd2 Kg6 36.Kf2 Rec7 37.Re2 Rc2 38.Rdd2 Rxd2 39.Rxd2 Kf5 40.Kg3 Kg6 41.Re2

2. [-0.12], d=30: 23.Ne4 Qd8 24.f4 Ng4 25.Qe2 Ne3 26.Qxa6 Rxe4 27.Qd3 Qa8 28.Rf3 Qc6 29.g3 h5 30.b4 b5 31.Rexe3 dxe3 32.Rxe3 Rc4 33.Re1 a6 34.Qd8+ Kh7 35.Qd3+ Qg6 36.Qf3 h4 37.Re5 Qb6+ 38.Kg2 g6 39.gxh4 Qd4 40.h5 Rc2+ 41.Re2 Rxe2+ 42.Qxe2 Qxf4 43.hxg6+ fxg6 44.Qe7+ Kg8 45.Qd8+ Kg7 46.Qe7+ Kh6

3. [-0.64], d=30: 23.f4 dxc3 24.Qxc3 Ra5 25.fxe5 Qe6 26.Qf3 Rd5 27.Re2 a5 28.b4 a4 29.Re3 b5 30.Qf2 Qd7 31.Rfe1 Rd2 32.R3e2 Rd4 33.Re3 Rd5 34.R3e2 Qe6 35.Qa7 Rc8 36.Rf1 Rd7 37.Qe3 Re8 38.Qf2 Rd3 39.Qc5 Rd5

4. [-1.15], d=30: 23.Nd1 Rd8 24.Re4 b5 25.Nf2 Nc4 26.Qe2 Re6 27.Rxe6 Qxe6 28.Qxe6 fxe6 29.Nd3 a5 30.Kf2 Kf7 31.Re1 Rd5 32.h4 g6 33.Re4 Ne3 34.g4 g5 35.hxg5 hxg5 36.Ke2 Kf6 37.b4 a4 38.Nc5 e5 39.Kd3 Nc4 40.Re1 Nxa3 41.Ne4+ Kg6

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (32-ply):

1. = (0.00): 26.Qe4 Qc8 27.Rfe1 Rc4 28.Qe3 Re6 29.b4 Rc2 30.f4 Ng6 31.Qd3 Rxe1+ 32.Rxe1 Rc1 33.Rxc1 Qxc1+ 34.Kf2 Kh8 35.Qe3 Qc4 36.Qe8+ Kh7 37.Qxf7 Qa2+ 38.Ke1 Qa1+ 39.Ke2 Qa2+ 40.Kd1 Qb3+ 41.Kd2 Qb2+

2. = (0.00): 26.Ne3 Qc8 27.Rfe1 Nc4 28.Nf1 Re6 29.Rxe6 Qxe6 30.Rc1 Rc6 31.Rc3 Qe2 32.Ng3 Qe1+ 33.Nf1 Qe2

3. = (0.00): 26.f4 Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 Rc1+ 31.Kh2 Rc7 32.Rfd2 Rce7 33.Kg1 Re2 34.Rd8+ Kh7 35.Kf1 Rxd2 36.Rxd2 Kg6 37.Kf2 Rc7 38.Rd6+ f6 39.Ke3 Rc2 40.Rd2 Rc1 41.g3 Rg1

4. = (0.00): 26.b4 Rc2 27.Qe4 Rb2 28.a4 Qc8 29.Rc1 Qd8 30.Rcd1

5. = (0.00): 26.h3 Qd6 27.Qe4 Rc4 28.Qe3 Rcc8 29.Rfe1 Kf8 30.Qf4 Qc5+ 31.Kh2 Ng6 32.Rxe8+ Rxe8 33.Qc1 Qd6+ 34.f4 Rd8 35.Nc3 Qf6 36.g3 Re8 37.Rd7 Ne5 38.fxe5 Qf2+ 39.Kh1 Qf3+ 40.Kh2 Qf2+

6. = (-0.04): 26.Nc3 Qxd4+ 27.Rxd4 Nc4 28.Ne4 Rcc8 29.Rxc4 Rxc4 30.Nd6 Rec8 31.g3 f5 32.Nxc8 Rxc8 33.Rf2 Kf7 34.f4 Rc6 35.Kg2 Ke6 36.Kf3 Kd7 37.Ke3 Re6+ 38.Kd4 Kd6 39.Rd2 g5 40.Kc3+ Ke7 41.Rf2 Kf6

7. = (-0.04): 26.b3 Qd6 27.b4 Rc2 28.Rf2 Qg6 29.Qe3 Qc6 30.Rxc2 Qxc2 31.Qd2 Qxd2 32.Rxd2 Nc4 33.Rd3 Re1+ 34.Kf2 Ra1 35.Ne3 Ra2+ 36.Kg3 Nxe3 37.Rxe3 Kf8 38.h4 g6 39.Rc3 Kg7 40.b5 h5 41.Kh3 Rb2

8. = (-0.08): 26.Nb4 Qxd4+ 27.Rxd4 Kf8 28.Rfd1 g5 29.Kf2 Nc4 30.Nd3 Rcc8 31.Rd7 Kg7 32.a4 a6 33.b4 Red8 34.Rxd8 Rxd8 35.Nb2 Rxd1 36.Nxd1 Kf6 37.Ke2 Ke5 38.Kd3 Nd6 39.Ne3 f5 40.g3 h5 41.Ng2 f4

9. = (-0.12): 26.Qa4 Re6 27.Nc3 Qb8 28.Qf4 Qc7 29.Rd4 Rc4 30.Rfd1 Qc5 31.b4 Rxd4 32.Qxd4 Qxd4+ 33.Rxd4 Rc6 34.Nd5 Rc2 35.b5 Rc4 36.Rxc4 Nxc4 37.a4 f6 38.Nc3 Kf7 39.Kf2 Ke6

click for larger view

1. [0.00], d=29: 26.Ne3 Qc8 27.Rfe1 Nc4 28.h4 Nxe3 29.Rxe3 Rxe3 30.Qxe3 Rc2 31.Rd2 <Rxd2> 32.Qxd2 Qc4 33.g3 Qe6 34.Kg2 g6 35.g4 Qe5 36.f4 <Qe4+>

click for larger view

<31...Rxd2> Here, of course, Black has the choice to exchange queens instead of rooks by <31...Qc5>. I don’t know if it makes any difference after 31...Qc5 32.Qxc5 Rxc5 33.Rd7 a5

<36...Qe4+> About the only question I have is whether in this type of ending it's better to advance the pawns in front of your king or not.

2. [0.00], d=29: 26.h3 <Qd6> 27.b4 Rc2 28.Rf2 Qg6 29.Ne7+ Rxe7 30.Qd8+ Kh7 31.Qxe7 Rxf2 32.Kxf2 Qc2+ 33.Ke1 Nd3+ 34.Rxd3 Qxd3 35.Qxf7 <Qxa3> 36.Qf5+ Kh8 37.Qc8+ Kh7 38.<Qc4> a5 39.Qc2+ Kh8 40.bxa5 bxa5 41.Kf2 <a4> 42.Qc8+ Kh7 43.Qc2+ Kh8

<26...Qd6> Given that White has just provided us with a tempo, the question is whether there's a more active move for Black. 26...Rc2? 26....Rc4? 26...Ng6? 26...Rc6? Even 26...Kf8 or 26...Kh7 look like possibilities.

<35...Qxa3> Well, that was an interesting forced sequence.

<38.Qc4> White has a draw by repetition by 38.Qf5+ if he wants it.

<41...a4> Given that a draw by repetition is coming, and since White can't afford to swap queens, is there anything that Black can do about it like 41...Qf8? Critter doesn't think so, 42.h4 Qe8 43.Qc4 a4 44.Qa6 h5 45.Qa5 Kh7 46.Qd5 g6 47.Qb7+ Kh6 48.Qd5 a3 49.Qd6 Qa8 50.Qe7 a2 51.Qg5+ Kh7 52.Qe7+ Kg8 53.Qe6+ Kh7 54.Qe7+ and draw by repetition.

3. [0.00], d=29: 26.b4 Rc2 27.Qe4 Ra2 28.a4 Qg5 29.Rf2 Ra3 30.f4 Qg4 31.Qb1 <Nf3+> 32.Kh1 Qh4 33.g3 Qh5 34.Ne7+ Kh8 35.Nf5 Re1+ 36.Rxe1 Nxe1 37.Qb2 Qxf5 38.Qxa3 Qe4+ 39.Kg1 Nd3 40.Ra2 Qd4+ 41.Kg2 Qd5+ 42.<Kg1>

<31...Nf3+> This is just a cute tactical trick that doesn't accomplish anything. Black should be looking at more solid moves like 31...Ng6

<42.Kg1> And a draw by repetition.

4. [-0.04], d=29: 26.Nc3 Rc8 27.Nd5 Rc4 28.Qd2 Rc5 29.Ne3 Qc8 30.Rfe1 Nc4 31.Nxc4 Rxe1+ 32.Qxe1 Rxc4 33.g3 g6 34.Qe3 h5 35.Rd2 <h4>

click for larger view

<35...h4> Black has managed to get an "attack" going but with so few pieces remaining it isn't going to amount to much.

5. [-0.05], d=29: 26.Qe4 Rc4 27.Qe2 Qc8 28.Rfe1 Re6 29.Ne3 Rcc6 30.Qd2 Nc4 31.Nxc4 Rxe1+ 32.Qxe1 Rxc4 33.g3 g6 34.Qe3 h5 35.Rd2 h4 36.<Re2>

click for larger view

click for larger view

1. [+0.04], d=30: 26.h3 Rc2 27.Rf2 Rc4 28.Qe3 Qc8 29.Re2 Rc2 30.Red2 Rxd2 31.Qxd2 Ng6 32.Nc3 Qc5+ 33.Qd4 Qf5 34.Ne4 Qf4 35.Qd7 Re7 36.Qd6 Re5 37.Qd8+ Kh7 38.Qd3 Re7 39.Nd6 Re3 40.Qd2 Re7 41.Kh1 Qe5 42.Qd4 <Qxd4>

click for larger view

<42...Qxd4> After 43.Rxd4 there isn't anything here for either side. Black can force the exchange or rooks by 43...Rd2 44.Nf5 Rxd4.

2. [+0.01], d=29: 26.b4 Rc2 27.Qe4 Ra2 28.a4 <Ra3> 29.Qd4 Ra2 30.Rf2 Rxf2 31.Qxf2 Nc6 32.Qc2 Qd6 33.Qc4 Qe6 34.h3 Ne7 35.Nxe7+ Qxe7 36.b5 Rd8 37.Rxd8+ Qxd8 38.f4 Qd2 39.Qe4 g5 40.Qe8+ Kg7 41.Qe5+ f6 42.Qe7+ Kg6 43.Qe8+ Kf5 44.Qc8+ Kxf4 45.Kh2 Qe3 46.<Qc4+>

click for larger view

<28...Ra3> Unfortunately, moving the knight loses to either 29.Nc7 or 29.Ne7+. But this and the next move by Black don't accomplish anything and represent the loss of 2 (maybe 3) tempos. So Black needs to improve here.

<46.Qc4+> Even though White is temporarily down a pawn, he could have restored material equality by 46.Qc7+ and 47.Qxa6, with an advanced 2:1 q-side pawn majority for good measure. But with White's queen now offside, it would be impossible for White's king to avoid the upcoming stream of checks.

3. [-0.02], d=29: 26.b3 Qd6 27.<b4> Rc2 28.Qe3 <Qe6> 29.Qe4 Qg6 30.Qxg6 Nxg6 31.Rf2 Ree2 32.Rxe2 Rxe2 33.Rd3 <Kh7> 34.Kf1 Ra2 35.Rc3 Nh4 36.Ne3 <b5> 37.f4 Kg6 38.Nc2 Nf5 39.g4 Nd6 40.Kg2 Nc4 41.Kg3 Kf6 42.Nd4 <Rxa3> 43.Nxb5 Rxc3+ 44.<Nxc3>

click for larger view

<27.b4> The subtleties of 26.b3 followed by 27.b4 escape me.

<28...Qe6> 28...Re6 protecting the rook and allowing the knight to begin "offensive" operations might be possible, but then Black needs to worry about a knight move, Rd8+ and Qe4+, picking up Black's Rc2.

<33...Kh7> I don't think that chess engines understand the concept of centralizing the king in the endgame, being too concerned with king safety. 33...f6 followed by 34...Kf7 seems better.

<36...b5> And the immediate 36...Kg6 also seems better, starting to centralize the king, protect the Pf7, and keep the b-pawn protected. 36...f5, pretty much forcing 37.f4 is also a possibility.

<42...Rxa3> And here the simple 42...a6 looks better since it's not clear how White can protect his a-pawn. Komodo couldn't figure it out either, the best it can come up with for White is ([-0.30], d=23), although for some unknown reason it took its time capturing the pawn: 43.h4 g6 44.g5+ Kg7 45.Kf3 Rh2 46.Kg3 Rd2 47.Nf3 Rd5 48.Kg4 h5+ 49.Kg3 Kf8 50.Ne1 Rd1 51.Ng2 Ra1 52.Ne3 Rxa3 53.Rxa3 Nxa3 54.Nd5 Nc2 55.f5 Nd4 56.fxg6 fxg6 57.Kf4 Kf7. But the advantage should be enough to win since if White's knight moves then ...Nc2 wins the Pb4. And if White's king tries to move to the q-side then ...Nf5 wins the Ph4.

click for larger view

4. [-0.05], d=29: 26.Ne3 Qc8 27.b4 Rc7 28.Rfe1 Rce7 29.Kf2 Ng6 30.<a4> Re5 31.b5 Qe6 32.Rd3 <h5> 33.Rd2 Nh4 34.Rdd1 <g6> 35.h3 g5 36.Qd2 Kh7 37.Qc3 Kg8 38.<Qd2>

click for larger view

<30.a4> White would do better by exchanging a set of rooks by 30.Nc4 or 30.Ng4.

<32...h5> Now that White has unwisely allowed his Ne3 to be pinned 32...f5 is a consideration. If 33.g3 then 33...h5 and Black has another target on the k-side.

<34...g6> And here 34...Nf5 looks like a killer, since if 35.Nxf5 then 35...Rxe1 36.Rxe1 Qxe1#.

<38.Qd2> And we've reached a situation where Komodo doesn't know what to do so it starts moving its pieces back and forth without an apparent plan. Engine dithering at work.

5. [-0.05], d=29: 26.Qe4 Qc8 27.b4 Rc4 28.Qb1 Ng6 29.h3 Nf4 30.Nxf4 Rxf4 31.Qb2 Qe6 32.Qd2 Rf5 33.Rfe1 Re5 34.Rxe5 Qxe5 35.Qd4 Qg5 36.Qd2 Qg3 37.Qd7 Re6 38.Qd2 Qe5 39.Qd4 Qe2 40.Qd3 a5 41.Qxe2 Rxe2 42.Rd8+ Kh7 43.Rb8 axb4 44.<axb4>

click for larger view

click for larger view

1. [0.00], d=28: 26.b4 Rc4 27.Qe3 Re6 28.Qe2 Qc8 29.Rd2 Rc1 30.Rxc1 Qxc1+ 31.Rd1 Qxa3 32.Nc7 Re7 33.Nd5 <Re6>

click for larger view

<33...Re6> Quick draw by repetition. What makes it slightly interesting is that White sacs a pawn in order to achieve the draw this quickly.

2. [0.00], d=28: 26.Qe4 Rc4 27.Qe2 Qc8 28.Rfe1 Rc2 29.Qe4 Re6 30.Nf4 Re7 31.Nd5 <Re6>

click for larger view

<31...Re6> An even quicker draw by repetition. And White didn't need to sac a pawn to achieve it.

3. [0.00], d=28: 26.Qa4 Re6 27.Nc3 Qc7 28.Nd5 Qd8

click for larger view

<28...Qd8> And an even quicker draw by repetition than PV #2.

4. [0.00], d=28: 26.h3 Qd6 27.Qa4 Qe6 28.b4 Qc6 29.Qxc6 Rxc6 30.b5 Rc5 31.Rd4 Ng6 32.a4 Rc2 33.Nb4 Rb2 34.Nd3 Ra2 35.Rf2 Ra1+ 36.Rf1 Ra3 37.Rf2 <Ra1+>

click for larger view

<37...Ra1+> Spark showed a remarkable lack of imagination in these first 4 PVs.

5. [-0.07], d=28: 26.<f4> Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 a5 31.Rfd2 <h5> 32.Kh2 Kh7 33.b4 Rb5 34.bxa5 <bxa5> 35.a4 Rb4 36.Rd5 Rxa4 37.Rxh5+ Kg6 38.Rg5+ Kf6 39.Rd6+ Re6 40.Rdd5 Rxf4 41.Kg3 Rfe4 42.Rgf5+ Ke7 43.<Rxa5>

click for larger view

<26.f4> This move seems to consistently give Black a small advantage but not enough to matter.

<31...h5> This was unnecessary and, unless followed immediately by g6, will cost Black, as you will see.

<34...bxa5> It might seem that 34...Rxa5 wins the a-pawn, but White counterattacks: 35.a4 (needed) Ra3 36.Rd7 and if 36...f6 (Spike suggested 36...f5), 37.Rb7 followed by 38.Rdd7 and White will win either Black's Pb7 or Pg7, and probably has the better chances.

click for larger view

1. [0.00], d=35: 26.<Qa4> Qa8 27.Qd4 Qc8 28.Rfe1 Rc4 29.Qe3 Re6 30.Nf4 <Rxf4> 31.Qxf4 Nd3 32.Rxd3 Rxe1+ 33.Kf2 <Rc1> 34.Rd2 Qc5+ 35.Kg3 Qe7 36.Rd8+ Qxd8 37.Qxc1 Qd6+ 38.Qf4 <Qg6+> 39.Qg4 Qc2 40.Qd7 Qg6+ 41.Kf2 Qc2+ 42.<Kg3>

click for larger view

<26.Qa4> Spike, as usual, marches to its own drummer, feinting a play on the q-side before returning to the "usual" play on the center.

<30...Rxf4> This sequence might work in blitz but not in CC, and definitely not with computers in use.

<33...Rc1> 33...Qe6 looks more aggressive on the surface but leads to nothing after 34.Qd2 Rb1 35.Rd8+ Kh7 36.Qc2+ Qg6 37.Rd2 Qxc2 38.Rxc2 g5 39.g4 Kg6 40.Ke3 Kf6 41.f4 Re1+ 42.Kf2 Rb1 43.Ke3 and a likely draw by repetition.

<38...Qg6+> White would have a slight advantage after 38...Qxf4+ 39.Kxf4 because of his more active king.

<42.Kg3> And a draw by repetition.

2. [0.00], d=35: 26.Qe4 Qc8 27.Rfe1 Re6 28.Nf4 Rc4 29.Qe3 <Rxf4> 30.Qxf4 Nd3 31.Rxd3 Rxe1+ 32.Kf2 <Rc1> 33.Rd2 Qc5+ 34.Kg3 Qe7 35.Rd8+ Qxd8 36.Qxc1 Qd6+ 37.Qf4 Qg6+ 38.Qg4 Qc2 39.Qd7 Qxb2 40.Qc8+ Kh7 41.Qf5+ Kg8 42.<Qc8+>

click for larger view

<29.Rxf4> Spike seems to really like this continuation and attempted swindle!

<32...Rc1> 32...Rb1 would have avoided the rook exchange but then Black needs to be constantly aware (after Black's queen moves) of the possibility of Rd8+, Qf5+, and Qxb1.

<42.Qc8+> The same draw by repetition as in PV #1.

3. [0.00], d=35: 26.h3 Rc4 27.Qf2 <Rc5> 28.<Qd4>

click for larger view

<27...Rc5> After the passive 27.Qf2 27...Qc8 threatening 28...Rc2 seems more aggressive. If White tries to block 28...Rc2 with 28.Nc3, then 28...a5 followed by 29...b5 and 30...b4 (if allowed) could open up the q-side somewhat and Black may be able to win one of White's q-side pawns.

click for larger view

4. [-0.02], d=35: 26.Ne3 Qc8 27.Rfe1 Nc4 28.Nxc4 Rxe1+ 29.Rxe1 Rxc4 30.Qe5 Kh7 31.Re2 Qd7 32.Kf2 a5 33.Qe3 Qc6 34.g3 <g5> 35.Qe5 Rc5 36.Qe3 Rc1 37.Qd3+ Kg7 38.Qd4+ Qf6 39.Ke3 <Qxd4+> 40.Kxd4 Kf6 41.Kd3 a4 42.Re4 b5 43.Rb4 Rc5 44.Re4 Rd5+ 45.Kc3 Kf5 46.Rd4 <Ke6>

click for larger view

<34...g5> I think this exposes Black's king too much. Better might be 34...g6 followed by ...f5

<39...Qxd4+> Helping White centralize his king doesn't seem like a good idea.

<46...Ke6> Both sides need to be careful that exchanging rooks doesn't leave the other side with a superior king position and allow themselves to get into zugswang.

5. [-0.06], d=35: 26.f4 Ng4 27.Qd2 Nf6 28.Nc3 Qc7 29.Qd6 Qxd6 30.Rxd6 Ne4 31.Nxe4 Rxe4 32.Rd2 f5 33.g3 Kf7 34.h4 Kg6 35.Kg2 Rc6 36.Rf3 Rce6 37.Rfd3 Kh5 38.Kf3 a5 39.Rc3 <g6> 40.Rcd3 a4 41.b4 axb3 42.Rxb3 g5 43.fxg5 hxg5 44.hxg5 <Kxg5>

click for larger view

<39...g6> Getting one's king trapped at the edge of the board, even though it's only temporary, doesn't seem like such a good idea.

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (33-ply):

1. = (0.00): 26...Re6 27.b4 Rcc6 28.Nf4 Red6 29.Rxd6 Qxd6 30.Re1 f6 31.Ng6 Nxg6 32.Qxg6 Qd4+ 33.Kf1 Qc4+ 34.Kg1 Qd4+

2. = (0.00): 26...Qb8 27.Rfe1 Re6 28.Qd4 Qe8 29.Re3 Nc6 30.Rxe6 Qxe6 31.Qd2 Ne5 32.b4 Rc8 33.Kf2 Kf8 34.Ne3 Nc4 35.Nxc4 Qxc4 36.Qd6+ Kg8 37.Qd7 Rc7 38.Qd8+ Kh7 39.Qd3+ Qxd3 40.Rxd3 Kg6 41.Ke3 Rc2 42.Rd7

3. = (0.00): 26...Qa8 27.b4 Rcc8 28.Rfe1 Rcd8 29.Re3 Ng6 30.Qb1 Rxe3 31.Nxe3 Qb8 32.g3 Rxd1+ 33.Qxd1 Qe5 34.Qd3 Qa1+ 35.Kg2 Qa2+ 36.Kh3 Qe6+ 37.Kg2 Qa2+

4. = (0.00): 26...Qc8 27.Rfe1 Rc4 28.Qe3 Re6 29.b4 Ng6 30.Qd3 Rxe1+ 31.Rxe1 Rc1 32.Rxc1 Qxc1+ 33.Kf2 Qb2+ 34.Kf1 Qa1+ 35.Kf2 Qb2+

5. = (0.00): 26...Rc4 27.Qe3 Re6 28.b3 Nc6 29.Qf2 Rc5 30.Nc3 Qf8 31.Qb2 Qb8 32.Ne4 Rh5 33.Ng3 Rb5 34.Ne4 Rh5

6. = (0.00): 26...Kf8 27.b4 Rc6 28.f4 Ng6 29.Ne7 Rd6 30.Nxg6+ fxg6 31.Rxd6 Qxd6 32.Qc4 Qd7 33.Rf3 Rc8 34.Qd3 Qxd3 35.Rxd3 Rc4 36.g3 Ke7 37.Re3+ Kd6 38.Kf2 Rc2+ 39.Re2 Rc3 40.Re3 Rc2+

7. = (0.08): 26...Rc8 27.Rfe1 Qd6 28.h3 Qc5+ 29.Kh1 Qc6 30.Qd4 Ng6 31.Rxe8+ Rxe8 32.Nc3 Ne5 33.Re1 Ng6 34.Rxe8+ Qxe8 35.Kh2 Qe5+ 36.Qxe5 Nxe5 37.b3 Nc6 38.Ne2 g5 39.Kg3 Kg7 40.f4 f5 41.b4 b5 42.Kf3

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (34-ply):

1. = (0.00): 26...Qxd4 27.Rxd4 Nc6 28.Rd3 Ne5 29.Rd2 Nc4 30.Nxc4 Rxc4 31.Kf2 f5 32.Rfd1 Kf7 33.Rd7+ Re7 34.R1d2 Rc7 35.R7d5 Kf6 36.f4 Re6 37.Kf3 g5 38.Rd7 Rc4 39.g3 a5 40.Rh7 Kg6 41.Rhd7 Kf6

2. = (0.00): 26...Qc7 27.b4 Nc6 28.Qg4 Rg5 29.Nd5 Qb8 30.Qc4 Ne7 31.Nc3 Nf5 32.Nd5 Ne7

3. = (0.00): 26...Qc8 27.Rfe1 Nc4 28.Nf1 Re6 29.Rxe6 fxe6 30.Ne3 Nxe3 31.Qxe3 Rc2 32.Rd2 Rxd2 33.Qxd2 Qc5+ 34.Kf1 Kf8 35.b4 Qc4+ 36.Kf2 Qh4+ 37.Kg1 Ke7 38.Qc3 Kd6 39.Qd3+ Ke7 40.Qc3

4. = (0.00): 26...Rc8 27.f4 Qxd4 28.Rxd4 Nc6 29.Rd3 Na5 30.g3 Nc4 31.Nxc4 Rxc4 32.Rf2 Re7 33.Kg2 f6 34.Rfd2 Kf7 35.Kf2 g6 36.Rd6 Rc1 37.g4 Rh1 38.Kg2 Rhe1 39.Kf2 Rh1

5. = (0.00): 26...Qe7 27.Rfe1 Qf6 28.f4 Ng4 29.Qxf6 Nxf6 30.g3 g6 31.Kg2 Rb5 32.Rd2 Rb3 33.Rde2 Rd3 34.Nc4 Rc8 35.Ne5 Rdd8 36.Kf3 Kg7 37.g4 Rd4 38.Ke3 Re4+ 39.Kf3 Rd4

6. = (0.20): 26...Qf6 27.f4 Nd7 28.Rfe1 Re7 29.g3 Qxd4 30.Rxd4 Rc8 31.Kf2 Rce8 32.Kf3 g6 33.g4 Kg7 34.Rd6 Nf6 35.Rxf6 Rxe3+ 36.Rxe3 Rxe3+ 37.Kxe3 Kxf6 38.Ke4 Ke6 39.h3 Kd6 40.f5 a5 41.Kd4 gxf5 42.gxf5

7. = (0.24): 26...Qd7 27.Qxd7 Nxd7 28.Kf2 Nf6 29.h4 h5 30.Rd4 Kf8 31.Rfd1 Rce5 32.R1d3 g6 33.b3 Kg8 34.b4 Kg7 35.Rc3 b5 36.Rdd3 a6 37.Rc6 R5e6 38.Rxe6 Rxe6 39.Rd8 Rc6 40.Ke2 Rc3

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (33-ply):

1. = (0.00): 26...Ng4 27.h3 Ne3 28.Nxe3 Qxd4 29.Rxd4 Rxe3 30.Rf2 Re7 31.Rd8+ Kh7 32.Rfd2 Rc1+ 33.Kf2 Kg6 34.R8d6+ f6 35.R6d5 Rce1 36.g3 Rh1 37.Kg2 Rhe1 38.Kf2

2. = (0.04): 26...Nc6 27.Qd3 b5 28.Rf2 Rc4 29.Rfd2 Ree4 30.Qf3 Qd6 31.g3 Rcd4 32.Rxd4 Rxd4 33.Rxd4 Nxd4 34.Qe4 Qc5 35.Kf2 Ne6+ 36.Ne3 Qc1 37.Qb4 a6 38.Qc3 Qxc3 39.bxc3 Kf8 40.Kf3 Ke7 41.c4 bxc4 42.Nxc4

3. = (0.04): 26...Nc4 27.b4 Rc6 28.Rf3 b5 29.Rg3 g6 30.Rgd3 Rce6 31.h3 a6 32.a4 Rd6 33.Qc5 Qb8 34.axb5 axb5 35.Nc7 Rxd3 36.Rxd3 Rc8 37.Rd7 Qb6 38.Kf2 Qxc5+ 39.bxc5 b4 40.Rd4 Na5 41.Nd5 Rxc5 42.Nxb4

4. = (0.16): 26...Rc4 27.Qe3 Nd7 28.Qg3 Rce4 29.Qc3 Qc8 30.Qxc8 Rxc8 31.Kf2 Nc5 32.g3 Kf8 33.Nc3 Ree8 34.Kf3 Rcd8 35.Rxd8 Rxd8 36.Rd1 Rxd1 37.Nxd1 Ke7 38.b4 Nb7 39.Ke4 Nd6+ 40.Kd5 Nb5 41.a4 Nc7+ 42.Kd4

5. = (0.24): 26...Ng6 27.Qd3 Qd6 28.g3 Qc6 29.Nc3 Nf8 30.Rfe1 Rxe1+ 31.Rxe1 Ne6 32.Rd1 g6 33.h4 Kg7 34.Kf2 Rc4 35.Qe3 Rc5 36.Qf3 Qc8 37.Qe4 h5 38.Rd6 Qc7 39.Rd5 Rxd5 40.Qxd5 Qc5+

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (32-ply):

1. = (0.00): 26...Rc2 27.Rf2 Qg5 28.Qe4 Rc1 29.Rff1 Rc4 30.Qe2 Qg6 31.Rd2 Kh7 32.Rfd1 Rcc8 33.Qf2 Nc4 34.Rd3 Ne5 35.R3d2

2. = (0.00): 26...Rc4 27.Qe3 Rc2 28.f4 Nd7 29.Qd3 Ree2 30.Ne3 Rb2 31.Rfe1 Rxe1+ 32.Rxe1 Qh4 33.Rf1 Nf8 34.Nd5 Ng6 35.h3 Kh7 36.Kh2 Ra2 37.Qf5 Qd8 38.Qd3 Qh4

3. = (0.00): 26...Rc6 27.Rfe1 Rce6 28.Re4 Qg5 29.Re3 Nc6 30.Rxe6 Rxe6 31.Qd2 Qh4 32.Nc7 Re5 33.Nd5 Qc4 34.Kf2 Qh4+ 35.Kg1

4. = (0.16): 26...Rc8 27.f4 Ng4 28.Rf3 Rc6 29.h3 Nf6 30.b5 Rc5 31.Nxf6+ Qxf6 32.Qxf6 gxf6 33.Rd7 Rxb5 34.Rxa7 Rb2 35.Rd3 h5 36.f5 Kg7 37.Rg3+ Kf8 38.Kh2 Ree2 39.Ra8+ Re8 40.Rg8+ Kxg8 41.Rxe8+ Kg7 42.Rb8

5. (3.23): 26...Rxd5 27.Qxd5 Qxd5 28.Rxd5 g5 29.f4 gxf4 30.Rxf4 Re7 31.Kf2 Kg7 32.Kg3 Ng6 33.Rff5 Re3+ 34.Rf3 Re7 35.Rff5

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (34-ply):

1. = (0.00): 26...Nc4 27.b4 Rc6 28.Rfe1 Ree6 29.a4 Rxe1+ 30.Rxe1 Rd6 31.Rd1 Nb2 32.Qxb2 Rxd5 33.Rxd5 Qxd5 34.Qc3 h5 35.a5 h4 36.axb6 axb6 37.Qe3 Qd1+ 38.Kf2 Qc2+ 39.Qe2 Qg6 40.Qd2 Qg3+ 41.Kf1 Qc7 42.Ke2

2. = (0.00): 26...Nc6 27.Qd3 Ne5 28.Qd4

3. = (0.00): 26...Qd6 27.b4 Rc2 28.Rf2 Rcc8 29.a4 Ng6 30.Rfd2 Ne5 31.Rf2

4. = (0.00): 26...Rc4 27.Qd2 Qc8 28.Rfe1 Rc2 29.Qd4 Rc4 30.Qf2 Rc2 31.Qd4

5. = (0.00): 26...Qc8 27.Qa4 Ra5 28.Qd4 Rc5

6. = (0.00): 26...Qb8 27.f4 Nc6 28.Qd3 Ne7 29.Nxe7+ Rxe7 30.Rfe1 Rxe1+ 31.Rxe1 Rc8 32.Qf5 Rc5 33.Qd3

7. = (0.00): 26...Rc2 27.Qa4 Qc8 28.Rf2 Rxf2 29.Kxf2 Nc4 30.Qb4 a5 31.Qb5 Nd6 32.Qd3 Nc4 33.Qd4 b5 34.a4 Nd6 35.Nb6 Qc2+ 36.Qd2 Qc5+ 37.Qd4 Qc2+

8. = (0.00): 26...Kf8 27.f4 Nc4 28.b4 Rxd5 29.Qxd5 Qxd5 30.Rxd5 Ne3 31.Rd7 Nxf1 32.Kxf1 Re7 33.Rd8+ Re8 34.Rxe8+ Kxe8 35.Ke2 Kd7 36.Kd3 Ke6 37.Ke4 g6 38.Kd4 Kd6

9. = (0.00): 26...Kh8 27.Ne3 Qxd4 28.Rxd4 Nc6 29.Rd3 Ne5 30.Rd2 g6 31.Kf2 Nc4 32.Nxc4 Rxc4 33.Rfd1 Kg7 34.g3 Kf6 35.f4 Re6 36.Kf3 Rce4 37.Rd3 Re1 38.R1d2 a6 39.Rc3 Rf1+ 40.Kg2 Rfe1 41.Kf2 Rh1

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (29-ply):

1. = (0.00): 26...Qxd4+ 27.Rxd4 Nc4 28.b4 Rcc8 29.Nb5 a6 30.Rxc4 Rxc4 31.Nd6 Rec8 32.Nxc4 Rxc4 33.Rd1 Rc3 34.Rd6 Rxa3 35.Rxb6 Ra2 36.Rb7 Rb2 37.Rb6 Ra2

2. = (0.00): 26...Qc8 27.Ne4 Rc4 28.Qd5 Re6 29.Nd6 Qc5+ 30.Qxc5 Rxc5 31.Rf2 g6 32.Rfd2 Nc4 33.Nxc4 Rxc4 34.Kf2 Kg7 35.Rd7 Kf6 36.R1d2 Re7 37.Kg3 Rc6 38.Rd8 Rec7 39.Kf4 Rc4+ 40.Ke3 Re7+ 41.Kf2 Rc6

3. = (0.00): 26...Rc8 27.Nd5 Rc4 28.Qf2 Qc8 29.Rfe1 Rc2 30.Qd4 Rc4 31.Qf2

4. = (0.08): 26...Qb8 27.Rfe1 Rc4 28.Qf2 Rcc8 29.Re2 Nc6 30.h3 Rcd8 31.Rxe8+ Rxe8 32.f4 Rd8 33.Rxd8+ Qxd8 34.Qf3 Qd4+ 35.Kh2 Nd8 36.Qe4 Qxe4 37.Nxe4 Nc6 38.Kg3 Kf8 39.Kf3 Ke7

5. = (0.08): 26...Qc7 27.Rfe1 Rc4 28.Qf2 Qc8 29.Qg3 Rc5 30.Re4 Ng6 31.Rxe8+ Qxe8 32.Qf2 Qe6 33.f4 Ne7 34.Re1 Qd7 35.Qe2 Nd5 36.Qe8+ Qxe8 37.Rxe8+ Kh7 38.Nxd5 Rxd5 39.Kf2 Rd2+ 40.Re2

6. = (0.16): 26...Qe7 27.Rfe1 Qf8 28.Ne4 Rc6 29.Qd5 Rce6 30.Re2

7. = (0.16): 26...Nc6 27.Qxd8 Rxd8 28.Rxd8+ Nxd8 29.Rd1 Ne6 30.Kf2

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (32-ply):

1. = (-0.16): 26...Qd6 27.b4 Rc4 28.Qe3 Rc2 29.Rfe1 Qg6 30.Nf4 Qg5 31.Kh1 Rcc8 32.h3 Kf8 33.Qe4 Ng4 34.Qd4 Red8 35.Nd5 Nf2+ 36.Qxf2 Rxd5 37.Rxd5 Qxd5 38.Qe3 Qd6 39.Qe2 Qf6 40.Qa6 Qc3 41.Rd1

2. = (-0.04): 26...Ng6 27.b4 Rc6 28.Rd2 Qc8 29.Rfd1 Ne7 30.Nxe7+ Rxe7 31.Kf2 Kh7 32.Qd8 Qxd8 33.Rxd8 Rc2+ 34.R1d2 Rxd2+ 35.Rxd2 Kg6 36.Rd6+ Kf5 37.Rd4 Rc7 38.Ke3 Rc3+ 39.Rd3 Rxd3+ 40.Kxd3

3. = (-0.04): 26...Kf8 27.Qd2 Ng6 28.f4 Ne7 29.Nxe7 Qxd2 30.Rxd2 Rxe7 31.Rf3 Rec7 32.Re3 g5 33.fxg5 hxg5 34.Kf2 Kg7 35.g4 Rc2 36.Rxc2 Rxc2+ 37.Kg3 Kg6 38.h3 Rc6 39.Rf3 Rd6 40.b4 Re6 41.Rc3 a6

4. = (0.00): 26...Nc6 27.Qd3 Re5 28.Nf4 Qe8 29.b4 Re3 30.Qd2 Rc4 31.Rfe1 Re5 32.Rxe5 Qxe5 33.Ne2 a6 34.f4 Qe6 35.Qd7 Qxe2 36.Qc8+ Kh7 37.Qf5+ Kg8 38.Qc8+

5. = (0.00): 26...Re6 27.Qd2 Re8 28.Qd4

6. = (0.00): 26...Qa8 27.a4 Nc6 28.Qd3 Ne5 29.Qd4

7. = (0.00): 26...Qc8 27.Rfe1 Ng6 28.Rxe8+ Qxe8 29.Qd3 Qc8 30.b4 Rc1 31.g3 Rxd1+ 32.Qxd1 Qc4 33.f4 Qa2 34.Qd3 h5 35.Ne3 h4 36.f5 hxg3 37.hxg3 Ne5 38.Qd8+ Kh7 39.Qh4+ Kg8 40.Qd8+

8. = (0.00): 26...Rc2 27.Qe4 Rb2 28.b4 Ra2 29.a4 Ra3 30.Qd4 Ra2 31.Qe4

9. = (0.00): 26...Kh8 27.h3 Nc6 28.Qd3 Ne5 29.Qe4 Nc6 30.Qd3

click for larger view

Analysis by Stockfish 2.0.1 JA 64bit (31-ply):

1. = (-0.24): 26...Qxd4+ 27.Rxd4 Kf8 28.Rfd1 Rcc8 29.R1d2 Nc4 30.Rc2 Ne3 31.Rxc8 Rxc8 32.Rd3 Rc1+ 33.Kf2 Nd1+ 34.Kg3 Nxb2 35.Rd7 g6 36.Rxa7 Rc5 37.f4 Nc4 38.Kf3 Rb5 39.Nd3 Rb3 40.Ke4 Rxa3 41.Rxa3 Nxa3 42.Nb4

2. = (-0.08): 26...Qf6 27.Nd5 Qg6 28.Rf2 Qe6 29.Nc3 Rcc8 30.Qe4 Nc4 31.Qxe6 Rxe6 32.Nd5 Kh7 33.Rc2 Rc5 34.a4

3. = (-0.08): 26...Qc8 27.Rfe1 Re6 28.Re3 Nc4 29.Rxe6 Qxe6 30.Nd3 Rc6 31.Nf4 Qe7 32.Qd8+ Qf8 33.Qxf8+ Kxf8 34.Rc1 Ke7 35.Kf2 Kd7 36.Nd3 f6 37.Rc2 Nd6 38.Nb4 Rxc2+ 39.Nxc2 Ke6 40.Ke2 Kd5 41.Kd3 a5 42.Ne3+

4. = (0.00): 26...Qe7 27.Rfe1 Qf6 28.Kh1 Rc4 29.Nd5 Qd8 30.Qd2 Nc6 31.Qf2 Rxe1+ 32.Qxe1 Rd4 33.Rxd4 Nxd4 34.Qe4 Ne6 35.h3 Qd6 36.Kg1 Kf8 37.Kf2 Nc5 38.Qd4 Ne6 39.Qe4

5. = (0.00): 26...Qa8 27.Nd5 Qc6 28.Nc3 Rc4 29.Qf2 Rc5 30.Qd4

click for larger view

All draws by repetition from Critter, and I've indicated a distinct draw by adding a "D" to the [0.00] eval to distinguish it from a [0.00] eval representing equal chances for both sides. But I think that Critter's PV#4 based on ...Rc2 gives us our best practical chances for anything other than a draw.

1. [0.00]D, d=28: 26...<Nc4> 27.b4 Rc6 28.a4 <Ree6> 29.Rf2 Red6 30.Qe4 Re6 31.<Qd4>

click for larger view

<26...Nc4> Blocking the rook and moving the knight before White's f3-f4 requires it doesn't seem like the right thing to do. White has given us a tempo and we should try to find a more aggressive use for it.

<28...Ree6> This looks like a wasted tempo as well. White's knight is his best placed piece and maybe this is the time to eliminate it with 28...Ne3 and the following forced (I think) sequence: 29.Nxe3 Qxd4 30.Rxd4 Rxe3. But Critter evaluates the resulting positions as somewhat advantageous for White, [+0.38], d=27: 31.Rd8+ Kh7 32.Rd7 a6 33.Rxf7 Rb3 34.Rf4 b5 35.axb5 axb5 36.Rd2 Rc4 37.Rf7 Rc6 38.Rd4 Rb2 39.Rg4 Rg6 40.Rff4 Rxg4 41.Rxg4 g5 42.Rd4 Kg6 43.h4 Kf6 44.Rd6+ Kg7 45.hxg5 hxg5 46.Rd5 Rxb4 46.Rxg5+ although Black may be able to hold the draw. But who wants that?

<31.Qd4> An insipid draw by repetition.

2. [0.00]D, d=28: 26...Qd6 27.Qa4 Qb8 28.f4 Nc4 29.Rfe1 b5 30.Rxe8+ Qxe8 31.Qxa7 Rxd5 32.Rxd5 Qe1+ 33.Kh2 Ne3 34.Rxb5 Nf1+ 35.Kg1 Ne3+ 36.<Kh2>

click for larger view

<36.Kh2> And a draw by repetition. But this looks unnecessarily risky to me. If we're satisfied with a draw there's a lot of less nerve-wracking ways to achieve it!

3. [0.00]D, d=28: 26...Rc4 27.Qf2 Rc5 28.<Qd4>

click for larger view

4. [0.00], d=28: 26...<Rc2> 27.Rf2 Rc4 28.Qe3 <Qc8> 29.Rfd2 Ng6 30.Qf2 Rc1 31.f4 Ne7 32.Nxe7+ Rxe7 33.Rxc1 Qxc1+ 34.Kh2 <Re4> 35.Qf3 <Qxd2> 36.Qxe4 <Qxb2> 37.Qe8+ Kh7 38.Qxf7 <Qxa3> 39.Qf5+ Kg8 40.Qe6+ Kh8 41.<Qc8+>

<26...Rc2> This is the line I currently favor, starting the pressure on the c-file and an attempt to put pressure on White's Pb2, although at the moment it doesn't lead anywhere but to a draw. But at least it gives us some possibilities to "slug it out". I'm not sure whether the immediate 26...Rc2 is better for Black to "force" White's 27.Rf2 and take that square away from White's queen or the intermediate 26...Rc4 followed by 27...Rc2 to force White's queen to the relatively passive square on f2. It all depends on whether we can take advantage of the White queen's somewhat exposed position on e3.

<28...Qc8> Unfortunately 28...Ng4 doesn't work after 29.Nf6+ Nxf6 30.Rxd8 Rxd8. The Black rook is pinned so it can't recapture 30...Rxe3.

<34...Re4> Another opportunity to improve on Black's play. 34...Re1 comes to mind to prevent the exchange of rooks but maybe the hesitation 34...Re4 and if 35.Qf3 then 35...Re1 now that White's queen has vacated the e1-h4 diagonal. If 34...Re1 immediately then Critter (d=26) can't find anything better than a draw by repetition after 35.Qd4 Rf1 36.Qe3 Kh7 37.Kg3 Re1 38.Qd3+ Kg8 39.Qd8+ Kh7 40.Qd3+

<35...Qxd2> Transposing into a queen endgame which may be our best chance to bring our presumably superior computing power into play since only deep, deep ply analysis would show whether Black might have a winning advantage. But if we do that we need to be prepared for what could be a long, complex, and possibly "boring" endgame.

<36...Qxb2> This may lead to a draw after 37.Qe8+ Kh7 38.Qxf7 but it does achieve the goal of upsetting the pawn structure symmetry and may represent one of our best chances.

click for larger view

With our 2:1 q-side majority if queens are exchanged, provided that we can centralize our king in time, the K+P ending may be won for Black. This gives us a slight practical advantage since we can push White's queen around knowing that GMVA would not be able to tolerate the exchange of queens. But better endgame players than me need to give us their opinion.

<38...Qxa3> However, this leads straight to a draw by repetition. But I can't see how we can both protect our Pa7 and prevent 39.Qf5+. 38...Qf6 39.Qxa7 Qxf4+ just restores the pawn structure symmetry that we achieved via 35...Qxb2. However, it does lead me to think that if we manage to get in ...a6 or ...a5 before we get into this line then we might be able to get into a queen endgame with a pawn advantage. But this will still not be a guaranteed win unless we succeed in exchanging queens.

<41...Qc8+> And a draw by repetition. But this entire line is still a work in progress since at least there are some possibilities along the way.

5. [0.00]D, d=28: 26...Nc6 27.Qd3 Ne5 28.<Qd4>