Which Programs Find the Tacking Move, 17. Qf5!

[tano-urayoan]

But the engines analysis first choice after Qh3 is dxe5 not Kg7 as you post. How different is that for your analysis?

The point Pedro makes is for engines the moves Qf5 or Qf3 or a3 all give the same advantage so any of the them is a good candidate move.

[Terry McCracken]

But the engines analysis first choice after Qh3 is dxe5 not Kg7 as you post. How different is that for your analysis?

The point Pedro makes is for engines the moves Qf5 or Qf3 or a3 all give the same advantage so any of the them is a good candidate move.

18..ef?! 19. Qxh6 with a winning position. In fact 19. f5!!..Kg7 is even better! I maintain 17. Qf5!! is clearly best. Qf3 and a3 are certainly not equal, they are simply poor choices.

If you and Pedro can't see this, then it's pointless trying to explain it.

[Eelco de Groot]

But the engines analysis first choice after Qh3 is dxe5 not Kg7 as you post. How different is that for your analysis?

The point Pedro makes is for engines the moves Qf5 or Qf3 or a3 all give the same advantage so any of the them is a good candidate move.

18..ef?! 19. Qxh6 with a winning position. In fact 19. f5!!..Kg7 is even better! I maintain 17. Qf5!! is clearly best. Qf3 and a3 are certainly not equal, they are simply poor choices.

If you and Pedro can't see this, then it's pointless trying to explain it.

Hi Terry,

Some more computer analysis, I don't think I understand the position very well also can't say I have analyzed it myself but coincidence has it that I was trying to do some programming and I was working on scoring pawns in the King attack, and that seems to be going on in this position as well! So a good testposition for the new Beta!

Toga 1.3.4 counts pawns in the attack on the King, but Toga 1.3.1 does not. The first results with counting code from Toga II 1.3.4 into the 3.1.2 were pretty bad. I don't really know why that was but now I have some intermediate form of counting, that does count pawns attacking close to the King but their attack-weight is still zero, so it only helps because the number of attackers is increased, not the weight of the number of attacks.

Anyway, this Toga Checkov Beta is not tuned yet and it now adds up lots of zeroes which is just a waste of time of course, but it should see maybe just a little better if White is getting some advantage, at least a little more than the 1.24 I got in the last analysis, because it also counts the pawns it scores the attack with f5 a little higher I think. At least that was the assumption but it does not yet get a high score here, in fact the eval is dropping a little in the last PV, or climbing from Black's point of view;

After 1. Qf5 g6 2. Qh3:

r3r1k1/pbp1qp2/3p2pp/2pnP3/5P2/3B3Q/PPPN2PP/4RRK1 b - -

Engine: Toga Checkov Beta 2f (256 MB)
by Thomas Gaksch and Fabien Letouzey

7/22 0:00 -1.64-- 2...h5 3.f5 Qg5 4.fxg6 fxg6 5.Ne4 Qd8
6.Qg3 Kh8 7.exd6 (32.741)

7/31 0:00 -1.35 2...Kh7 3.f5 Nb4 4.Bb5 Nxc2 5.fxg6+ fxg6
6.Bxe8 Qxe8 7.Ne4 Bxe4 8.Rxe4 (113.876)

7/31 0:00 -1.29 2...dxe5 3.fxe5 Kg7 4.e6 fxe6 5.Bb5 c6
6.Bd3 Rad8 7.Qg3 Kh8 8.Qxg6 (166.840)

8/31 0:00 -1.20 2...dxe5 3.fxe5 Kg7 4.Rf2 Rad8 5.e6 Nb4
6.Qg3 Nxd3 7.cxd3 fxe6 8.Nc4 (339.392)

9/31 0:01 -1.21 2...dxe5 3.Nc4 h5 4.Qf3 Rad8 5.f5 Qf6
6.fxg6 Qxf3 7.Rxf3 f6 8.Bf5 (748.743) 551

10/34 0:03 -1.38 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 Kxg6 8.Nxc5 Qf7
9.Ne4 Qxa2 (1.675.678) 543

10/34 0:04 -1.36 2...Qf8 3.Ne4 Kg7 4.Qg3 dxe5 5.fxe5 Kh8
6.c4 Nb4 7.Nf6 Nxd3 8.Qxd3 Rad8 (2.220.862) 540

11/34 0:06 -1.40 2...Qf8 3.Ne4 Kg7 4.exd6 f5 5.Nxc5 Qxd6
6.Nxb7 Qb6+ 7.Kh1 Qxb7 8.Bc4 Kh7
9.Bxd5 Qxd5 (3.454.854) 540

12/41 0:18 -1.60 2...Qf8 3.Ne4 Kg7 4.Bc4 Nb6 5.Bb5 c6
6.Be2 dxe5 7.f5 Bc8 8.g4 (10.016.701) 541

12/42 0:28 -1.28 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 Rad8 8.g3 Nd5 9.Qd2 Kxg6
10.g4 (15.353.056) 540

13/46 0:39 -1.31 2...dxe5 3.fxe5 Kg7 4.e6 f5 5.Qg3 Qg5
6.Nb3 Qxg3 7.hxg3 Bc6 8.Nxc5 Re7
9.g4 fxg4 10.Re4 Nf6 (21.410.647) 541

14/46 1:13 -1.34 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 c6 8.Nxc5 Nd5 9.Qc4 Kxg6
10.Ne4 Kh7 11.Qxc6 (39.951.070) 546

15/63 2:35 -1.53 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 Rad8 8.g3 Nxg6
9.Rxf6 Rf8 10.Rxf8 Rxf8 11.Nxc5 Qd6
12.g4 (86.010.378) 552

16/63 7:25 -1.56 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 Rad8 8.g3 Nxg6
9.Rxf6 Rf8 10.Ra6 Kh7 11.Rxa7 Rd5
12.Kh1 (250.352.349) 562

17/63 20:31 -1.48 2...dxe5 3.f5 Kg7 4.Be4 Nf4 5.Qc3 Bxe4
6.Nxe4 f6 7.fxg6 Qe6 8.g3 Nd5 9.Qxc5 f5
10.c4 Ne7 11.Nc3 Nxg6 12.Kh1 Qd6
13.Qxd6 cxd6 14.Rxf5 (695.440.832) 564

18/74 64:32 -1.36 2...dxe5 3.f5 Kg7 4.Ne4 Qf8 5.Bb5 c6
6.fxg6 Nf4 7.Rxf4 exf4 8.Bc4 f6
9.Qf3 Re7 10.Bf7 Qd8 11.Rd1 Rd7
12.Rf1 Ba6 13.Nxc5 (2.179.587.999) 562

Eval seems a bit unstable here and there, this is just the second position I a running with this version so really a bit early to make any judgments about it.. but I think I'll make some changes...

Regards, Eelco

[YL84]

I can't resist to add my engine to this wonderful list
Milady : 1.7s (analysis below) on a PIII 450MHz. Very nice and complex position, thanks
Yves

"depth 1 score 200 temps 0 nbnod 48 B_F :47 : e5d6"
"depth 2 score 150 temps 6 nbnod 238 B_F :13,7840487520902 : e5d6 e7f8"
"depth 3 score 90 temps 14 nbnod 1281 B_F :10,1413266033008 : c2c4 d5b4 e5d6"
"depth 4 score 100 temps 61 nbnod 6688 B_F :8,57509800675182 : d3b5 c7c6 e5d6 e7e1"
"depth 5 score 200 temps 170 nbnod 22332 B_F :6,90032524311108 : h3f5 g7g6 f5h3 c5c4 e5d6"
"depth 6 score 210 temps 371 nbnod 42453 B_F :5,21524753937114 : h3f5 g7g6 f5h3 d5b4 d3b5 b7c6"
"depth 7 score 210 temps 687 nbnod 98952 B_F :4,77378621309543 : h3f5 g7g6 f5h3 d5b4 d3b5 b7c6 e5d6"
"depth 8 score 180 temps 1522 nbnod 209141 B_F :4,26842131757651 : h3f5 g7g6 f5h3 d5b4 d3b5 b7c6 e5d6 e7f6"
"depth 9 score 150 temps 4614 nbnod 715963 B_F :4,30400560973236 : h3f5 g7g6 f5g4 d5b6 e5d6 e7d6 e1e8 a8e8 d3b5"
"depth 10 score 150 temps 12401 nbnod 1709226 B_F :3,97838149163255 : h3f5 g7g6 f5g4 d5b6 d3b5 c7c6 b5d3 d6d5 e5e6 f7f5"
"depth 11 score 90 temps 99196 nbnod 13909232 B_F :4,40772559497313 : h3f5 g7g6 f5h3 d5b4 h3h6 b4d3 c2d3 a8d8 a2a3 d6e5 f4e5"