Test your engine!! A very very complicated position

[dr_zied_haddad]

Hi All,
What will you play in this position?

8/1k6/2bq1N2/3B2P1/2Qn3r/2R5/8/3K4 w - -

Many lines seems to be interesting but will probably finish in a draw. I think only one line can be promising for white, but winning will be very very difficult.

[jarkkop]

8/1k6/2bq1N2/3B2P1/2Qn3r/2R5/8/3K4 w - - 0 1

Analysis by Toga

1.Rc1 Rf4 2.Rb1+ Kc8 3.Rb6 Kd8 4.Rxc6 Nxc6 5.Qxc6 Qxc6 6.Bxc6 Rf5 7.Ne4 Kc7 8.Ba4 Re5 9.Bc2 Re6
+- (2.66) Depth: 12/42 00:00:11 24560kN
1.Rc1 Rf4 2.Rb1+ Kc8 3.Rb6 Kd8 4.Rxc6 Nxc6 5.Qxc6 Qxc6 6.Bxc6 Rf5 7.Ne4 Kc7 8.Be8 Rf4 9.Bg6 Rf3 10.Nf6
+- (2.63) Depth: 13/50 00:00:30 69740kN
1.Rc1 Rf4 2.Rb1+ Kc8 3.Qa6+ Kd8 4.Rb7 Rf1+ 5.Qxf1 Bxb7 6.Qh3 Bc6 7.Qh8+ Kc7 8.Ne8+ Bxe8 9.Qxd4 Bg6 10.Qa7+ Kd8 11.Qa8+ Kd7
² (0.67) Depth: 14/50 00:01:26 204mN
1.g6 Ka7 2.Qc5+ Qxc5 3.Rxc5 Bxd5 4.Rxd5 Ne6 5.Rd7+ Ka8 6.Nd5 Rg4 7.Ne7 Nc5 8.Rd5 Ne4 9.Kc2 Kb7 10.Rb5+ Kc7 11.Kd3 Kd7
± (0.93) Depth: 14/59 00:05:04 747mN
1.g6 Ka7 2.Qc5+ Qxc5 3.Rxc5 Bxd5 4.Rxd5 Ne6 5.Rd7+ Ka8 6.Nd5 Rg4 7.Ne7 Nc5 8.Rd5 Ne4 9.Kc2 Kb7 10.Kd3 Kc7 11.Rb5 Kd7
± (0.93) Depth: 15/65 00:07:47 1171mN
1.g6 Ka7 2.Qc5+ Qxc5 3.Rxc5 Bxd5 4.Rxd5 Ne6 5.Rd7+ Ka8 6.Nd5 Rg4 7.Ne7 Nc5 8.Rd5 Ne4 9.Rd8+ Kb7 10.Rd7+ Ka6 11.Ke2 Kb5 12.Rd5+ Kc4 13.Ke3 Nc5
± (0.92) Depth: 16/65 00:13:16 2034mN
1.g6 Ka7 2.Qc5+ Qxc5 3.Rxc5 Bxd5 4.Rxd5 Ne6 5.Rd7+ Ka8 6.Nd5 Rg4 7.Ne7 Nc5 8.Rd8+ Kb7 9.Rd5 Kb6 10.Rd6+ Kc7 11.Rf6 Rd4+ 12.Kc1 Rc4+ 13.Kb2 Rb4+ 14.Ka2 Ra4+ 15.Kb1 Rb4+ 16.Kc2
± (0.80) Depth: 17/68 00:25:47 4003mN
1.g6 Ka7 2.Qc5+ Qxc5 3.Rxc5 Bxd5 4.Rxd5 Ne6 5.Rd7+ Ka8 6.Rh7 Rf4 7.Nh5 Rg4 8.Rh6 Kb7 9.Ke1 Rg2 10.Kf1 Rg5 11.Kf2 Kc7 12.Ke3 Re5+ 13.Kd3 Rd5+ 14.Ke4
² (0.66) Depth: 18/69 01:00:27 9582mN

[dr_zied_haddad]

The only line that could provide white with advantage is Rc1!! I analysed this position which is issued of one of my correspondance games. I spent 3 days analysing i also used F11, DS11, Rybka2.3.2a and H11.2MP to see which engine finds the right lines. All i can say is that as far as the analysis goes, the engines seems to be completely lost in evaluating this position. I'll reach an ending with Q+B vs Q where finding winning moves isn't very easy because the opponents queen's checks. I thought first that i'll reach an ending with Q+B+P vs Q which is winning for me, but i overlooked a queen sacrifice leading to a stalemate!! What a shame.

Really keep your engine pondering for many hours, especially if you don't have a 4 or 8 threads you'll see what i'm talking about.

I think that this position could be added to a personal testset database +++

[dr_zied_haddad]

After analysing many days, i finally must convince my self that i'm not going to win this position with white. In fact, this endgame will finish as a draw if Black plays well. Knowing my opponents strength, he's gonna achieve this task.

[Dann Corbit]

The only line that could provide white with advantage is Rc1!! I analysed this position which is issued of one of my correspondance games. I spent 3 days analysing i also used F11, DS11, Rybka2.3.2a and H11.2MP to see which engine finds the right lines. All i can say is that as far as the analysis goes, the engines seems to be completely lost in evaluating this position. I'll reach an ending with Q+B vs Q where finding winning moves isn't very easy because the opponents queen's checks. I thought first that i'll reach an ending with Q+B+P vs Q which is winning for me, but i overlooked a queen sacrifice leading to a stalemate!! What a shame.

Really keep your engine pondering for many hours, especially if you don't have a 4 or 8 threads you'll see what i'm talking about.

I think that this position could be added to a personal testset database +++

Can you post the analysis that you have created? I would like to see it.

[GeorgeLyapko]

The only line that could provide white with advantage is Rc1!! I analysed this position which is issued of one of my correspondance games. I spent 3 days analysing i also used F11, DS11, Rybka2.3.2a and H11.2MP to see which engine finds the right lines. All i can say is that as far as the analysis goes, the engines seems to be completely lost in evaluating this position. I'll reach an ending with Q+B vs Q where finding winning moves isn't very easy because the opponents queen's checks. I thought first that i'll reach an ending with Q+B+P vs Q which is winning for me, but i overlooked a queen sacrifice leading to a stalemate!! What a shame.

Really keep your engine pondering for many hours, especially if you don't have a 4 or 8 threads you'll see what i'm talking about.

I think that this position could be added to a personal testset database +++

The main problem for programs in such positions is null-move. You should definitely switch it off to get more or less adequate analysis. Maybe somebody knows how to turn null-move verification off in Yace?

[dr_zied_haddad]

Hi Dann
Thanks for your interests in this analysis.
I'll post my analysis (and not computer analysis), thus i must insert the variations manually. This will take a few hours (not much free time because of collegues holidays for Xmas).
I'll also post some interesting computer analysis ++ later.