Kornrade Test

Kornrade · Post by **Kornrade** » Wed Sep 12, 2018 12:27 pm

Hello all,

I have finally succeeded in making an up-to-date list of my studies on the personal site Chichitza
For all my compositions, I tried to add an element of difficulty for the engines (be it fortress, zugzwang, long-term strategy that goes beyond the horizon, etc., or a combination of those).

So, here is my Kornrade Test (for humans and machines):

Code: Select all

01 Constrictor  +- [FEN "3k4/2pq3p/pp5R/4P3/P6P/2PN4/1PN3K1/8 w - - 0 1"]
02 Mirror       +- [FEN "6Nk/pp2Np1p/2p2Pp1/2R2bP1/7K/P7/1q1n1Q2/5n2 w - - 0 1"]
03 Walk         +- [FEN "5B1q/3N3p/p3p2p/7k/B7/5PP1/K6P/8 w - - 0 1"]
04 Detour       +- [FEN "7q/b1p5/1p1Npkb1/pPP2ppP/P1P5/3B2P1/5P1R/K3R3 w - - 0 1"]
05 Stubbornness == [FEN "8/b4kp1/2p1p3/5pB1/p1PP4/1p6/3K1P2/3N4 w - - 0 1"]
06 Guard Change +- [FEN "2k4N/Q1np4/2p2Bpp/1p1P4/pPP1p2P/P7/7q/1K6 w - - 0 1"]
08 Slowness     +- [FEN "q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - 0 1"]
09 Standstill   == [FEN "1N4K1/1pp1p3/7k/2r3pP/p2pp1P1/P6N/8/2B5 w - - 0 1"]
10 Shelter      == [FEN "4B3/k1K1pp2/p3bq2/2p1p3/1p2P1P1/4P2Q/nR6/8 w - - 0 1"]
11 Chase        +- [FEN "4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - 0 1"]
12 Desperation  == [FEN "2b5/1pr4p/3bp1pk/1p6/1PpN2PP/K1P1n3/P3N1R1/3R4 w - - 0 1"]
13 Shield       +- [FEN "8/4Pr1N/8/6PN/8/3K1ppp/4prkb/4Bbnn w - - 0 1"]
14 Fork-field   +- [FEN "3B2r1/n3Pp2/P7/5k1p/q3NP2/7K/5P2/4NB2 w - - 0 1"]
15 Harassment   +- [FEN "8/q1p1N1p1/3p1p1p/pk5P/1p2BB2/8/PK3P2/8 w - - 0 1"]
16 Smothering   +- [FEN "6kq/2p4p/6P1/5nN1/K7/2P5/7B/3B4 w - - 0 1"]
* Smothering II +- [FEN "5k2/2p4p/6P1/1p3n2/3q4/PB4BN/KP5P/8 w - - 0 1"]
17 Fishing Rod  == [FEN "8/1B3b1p/7P/p2p4/P2pp1kp/Q7/1N1Pq2P/7K w - - 0 1"]
18 Stronghold   == [FEN "6k1/1qr1p2p/ppN5/3p1N2/6p1/1P1P2Pp/PP2PP1P/2RK4 w - - 0 1"]
19 Evil Eye     == [FEN "8/pr1Pq3/br1kP3/1R2RQ2/5p2/1nPK4/5P2/8 w - - 0 1"]
20 Roundabout   +- [FEN "4b3/5p2/2p4p/P1p1PN2/2pn2P1/P2k4/Pr1npKN1/B7 w - - 0 1"]

Further information (e.g. co-authors, links) and solutions to the studies are available on Chichitza.
Study 07 is intentionally missing, it has been shown to be cooked and is currently (still) under repair...
Study 16(II) is the submitted (and not accepted) version of study 16, which I consider more challenging for the engines than the published version.

I know that some of the positions have been the subject of discussions in this forum and I am grateful for your opinions and insights on those threads. Therefore, my questions here are a little more general:

How would you rate the difficulty of the studies for engines, considering your favourite engine and available hardware?
0 Engines solve it immediately
1 Engines are slightly confused
2 Engines are very confused
3 Engines are utterly confused

If you don't mind, would like to update the difficulty iconograms on the site according to these evaluations.

Also, from a human perspective, how entertaining would you consider the studies?

Thank you and have fun

!

peter · Post by **peter** » Wed Sep 12, 2018 3:13 pm

Hi Mihai!

Kornrade wrote: ↑Wed Sep 12, 2018 12:27 pm I have finally succeeded in making an up-to-date list of my studies on the personal site Chichitza

Good to see the new update of your site online again!

As you know my opinion about most of your fine studies already, I'll have al look just at the ones, I don't have stored in my private collection till now. There are several of your ones among my most beloved pieces.

Being rather busy at the moment, it could take a while, but I'll come back here then.

Best regards

peter · Post by **peter** » Wed Sep 12, 2018 9:27 pm

peter wrote: ↑Wed Sep 12, 2018 3:13 pm I'll come back here then.

Curiosity was too big, had to look at the one from this year, Roundabout, as a matter of fact the only one, I hadn't stored till now.

One more masterpiece of a Zugzwang- position, still one of the last weaknesses of engines.
At least 3 points, "utterly confused".

As for human entertainment- factor, very funny for me too, I've learned to cope with not fully understanding positions that even modern engines don't "understand" at all neither.

Even at least to keep the winning eval and moves in hash at backwarding the solution-line you have to use engines with nullmove to be switched off, here's Michael Byrne's McCaine with unchecked nullmove- option and hash full from going back many moves with Zugzwang very deep in the tree again and again, which to solve from the starting position and evaluate it correctly as won, even nullmove- less engines don't have good chances still:

[d]4b3/5p2/2p4p/P1p1PN2/2pn2P1/P2k4/Pr1npKN1/B7 w - -

Engine: McCain-v2 -> the MaverickS (32768 MB)
by dedicated to John S McCain, an American

47/83 9:07 +1.88 1.Ne1+ Kc3 2.Ng3 Bd7 3.a6 Bc8 4.a7 Bb7
5.Kg2 Ba8 6.g5 hxg5 7.Kh3 Bb7 8.e6 fxe6
9.Kg4 Ba8 10.Kh5 Bb7 11.Kg6 Ba8
12.Kf7 Bb7 13.Ke8 Ba8 14.Kd8 (12.322.281.580) 22524

...

48/87 12:00 +1.80 1.Ne1+ Kc3 2.Ng3 Bd7 3.a6 Bc8 4.a7 Bb7
5.Kg2 Ba8 6.g5 hxg5 7.e6 fxe6 8.Kh3 Bb7
9.Kg4 Ba8 10.Kh5 Bb7 11.Kg6 Ba8
12.Kf7 Bb7 13.Ke8 Ba8 14.Kd8 (16.249.903.401) 22557

49/86 12:26 +1.90++ 1.Ne1+ (16.829.602.605) 22554

49/91 13:48 +2.00++ 1.Ne1+ (18.699.440.453) 22568

49/93 14:56 +2.14++ 1.Ne1+ (20.227.365.842) 22572

Best regards

Paloma · Post by **Paloma** » Wed Sep 12, 2018 11:30 pm

Engine: McCain-v2 -> the MaverickS (32768 MB)
by dedicated to John S McCain, an American

47/83 9:07 +1.88 1.Ne1+ Kc3 2.Ng3 Bd7 3.a6 Bc8 4.a7 Bb7
5.Kg2 Ba8 6.g5 hxg5 7.Kh3 Bb7 8.e6 fxe6
9.Kg4 Ba8 10.Kh5 Bb7 11.Kg6 Ba8
12.Kf7 Bb7 13.Ke8 Ba8 14.Kd8 (12.322.281.580) 22524

Peter,
what compile of McCain used you? LMonge or JStanback?

because my JStanback exe produced an error message >>libwinpthread-1.dll is missed on your computer<<
same as JStanback exe of McBrain-99

Vincent has see same problem.

see here:
viewtopic.php?f=2&t=68366#p773323

peter · Post by **peter** » Thu Sep 13, 2018 12:05 am

Paloma wrote: ↑Wed Sep 12, 2018 11:30 pm what compile of McCain used you? LMonge or JStanback?

because my JStanback exe produced an error message >>libwinpthread-1.dll is missed on your computer<<
same as JStanback exe of McBrain-99

LMonge works in Shredder13 at my PC, JStanback doesn't, same error message.
BTW LMonge doesn't work in Fritz14 neither, no error message, just no name nor UCI- parameters shown.

Vinvin · Post by **Vinvin** » Thu Sep 13, 2018 12:53 am

EPD simple list :

3k4/2pq3p/pp5R/4P3/P6P/2PN4/1PN3K1/8 w - - 0 1
6Nk/pp2Np1p/2p2Pp1/2R2bP1/7K/P7/1q1n1Q2/5n2 w - - 0 1
5B1q/3N3p/p3p2p/7k/B7/5PP1/K6P/8 w - - 0 1
7q/b1p5/1p1Npkb1/pPP2ppP/P1P5/3B2P1/5P1R/K3R3 w - - 0 1
8/b4kp1/2p1p3/5pB1/p1PP4/1p6/3K1P2/3N4 w - - 0 1
2k4N/Q1np4/2p2Bpp/1p1P4/pPP1p2P/P7/7q/1K6 w - - 0 1
q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - 0 1
1N4K1/1pp1p3/7k/2r3pP/p2pp1P1/P6N/8/2B5 w - - 0 1
4B3/k1K1pp2/p3bq2/2p1p3/1p2P1P1/4P2Q/nR6/8 w - - 0 1
4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - 0 1
2b5/1pr4p/3bp1pk/1p6/1PpN2PP/K1P1n3/P3N1R1/3R4 w - - 0 1
8/4Pr1N/8/6PN/8/3K1ppp/4prkb/4Bbnn w - - 0 1
3B2r1/n3Pp2/P7/5k1p/q3NP2/7K/5P2/4NB2 w - - 0 1
8/q1p1N1p1/3p1p1p/pk5P/1p2BB2/8/PK3P2/8 w - - 0 1
6kq/2p4p/6P1/5nN1/K7/2P5/7B/3B4 w - - 0 1
5k2/2p4p/6P1/1p3n2/3q4/PB4BN/KP5P/8 w - - 0 1
8/1B3b1p/7P/p2p4/P2pp1kp/Q7/1N1Pq2P/7K w - - 0 1
6k1/1qr1p2p/ppN5/3p1N2/6p1/1P1P2Pp/PP2PP1P/2RK4 w - - 0 1
8/pr1Pq3/br1kP3/1R2RQ2/5p2/1nPK4/5P2/8 w - - 0 1
4b3/5p2/2p4p/P1p1PN2/2pn2P1/P2k4/Pr1npKN1/B7 w - - 0 1

zullil · Post by **zullil** » Thu Sep 13, 2018 12:14 pm

17 Fishing Rod == [FEN "8/1B3b1p/7P/p2p4/P2pp1kp/Q7/1N1Pq2P/7K w - - 0 1"]

[d]8/1B3b1p/7P/p2p4/P2pp1kp/Q7/1N1Pq2P/7K w - - 0 1

Stockfish-dev solves this one, at about depth 60 on 20 cores. So I'd rate this one a 2; Stockfish seemed very confused for a long time.

Pretty study!

lech · Post by **lech** » Thu Sep 13, 2018 1:21 pm

Kornrade wrote: ↑Wed Sep 12, 2018 12:27 pm ...
For all my compositions, I tried to add an element of difficulty for the engines (be it fortress, zugzwang, long-term strategy that goes beyond the horizon, etc., or a combination of those).

So, here is my Kornrade Test (for humans and machines):

Code: Select all

01 Constrictor  +- [FEN "3k4/2pq3p/pp5R/4P3/P6P/2PN4/1PN3K1/8 w - - 0 1"]
02 Mirror       +- [FEN "6Nk/pp2Np1p/2p2Pp1/2R2bP1/7K/P7/1q1n1Q2/5n2 w - - 0 1"]
03 Walk         +- [FEN "5B1q/3N3p/p3p2p/7k/B7/5PP1/K6P/8 w - - 0 1"]
04 Detour       +- [FEN "7q/b1p5/1p1Npkb1/pPP2ppP/P1P5/3B2P1/5P1R/K3R3 w - - 0 1"]
05 Stubbornness == [FEN "8/b4kp1/2p1p3/5pB1/p1PP4/1p6/3K1P2/3N4 w - - 0 1"]
06 Guard Change +- [FEN "2k4N/Q1np4/2p2Bpp/1p1P4/pPP1p2P/P7/7q/1K6 w - - 0 1"]
08 Slowness     +- [FEN "q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - 0 1"]
09 Standstill   == [FEN "1N4K1/1pp1p3/7k/2r3pP/p2pp1P1/P6N/8/2B5 w - - 0 1"]
10 Shelter      == [FEN "4B3/k1K1pp2/p3bq2/2p1p3/1p2P1P1/4P2Q/nR6/8 w - - 0 1"]
11 Chase        +- [FEN "4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - 0 1"]
12 Desperation  == [FEN "2b5/1pr4p/3bp1pk/1p6/1PpN2PP/K1P1n3/P3N1R1/3R4 w - - 0 1"]
13 Shield       +- [FEN "8/4Pr1N/8/6PN/8/3K1ppp/4prkb/4Bbnn w - - 0 1"]
14 Fork-field   +- [FEN "3B2r1/n3Pp2/P7/5k1p/q3NP2/7K/5P2/4NB2 w - - 0 1"]
15 Harassment   +- [FEN "8/q1p1N1p1/3p1p1p/pk5P/1p2BB2/8/PK3P2/8 w - - 0 1"]
16 Smothering   +- [FEN "6kq/2p4p/6P1/5nN1/K7/2P5/7B/3B4 w - - 0 1"]
* Smothering II +- [FEN "5k2/2p4p/6P1/1p3n2/3q4/PB4BN/KP5P/8 w - - 0 1"]
17 Fishing Rod  == [FEN "8/1B3b1p/7P/p2p4/P2pp1kp/Q7/1N1Pq2P/7K w - - 0 1"]
18 Stronghold   == [FEN "6k1/1qr1p2p/ppN5/3p1N2/6p1/1P1P2Pp/PP2PP1P/2RK4 w - - 0 1"]
19 Evil Eye     == [FEN "8/pr1Pq3/br1kP3/1R2RQ2/5p2/1nPK4/5P2/8 w - - 0 1"]
20 Roundabout   +- [FEN "4b3/5p2/2p4p/P1p1PN2/2pn2P1/P2k4/Pr1npKN1/B7 w - - 0 1"]

...

It is wonderful that you are here and I hope I will never feel alone.

peter · Post by **peter** » Fri Sep 14, 2018 7:37 am

Hi Louis!

zullil wrote: ↑Thu Sep 13, 2018 12:14 pm Pretty study!

Fishing Rod is the easiest one for engines, I guess, followed by Chase probably, Walk and two or three others are engine- solveable too, if you try enough engines and settings and give them enough hardware- time.
Two strong solvers are Sting and SF MateFinder, at least for backward solving, yet both still without any chance for the most of the Chichitza- positions with empty hash, as for not only "finding" first move but giving a correct according output- line and eval too.

One of my all time favorites is Stubbornness, nr.5 in the set.

[d]8/b4kp1/2p1p3/5pB1/p1PP4/1p6/3K1P2/3N4 w - - 0 1

One of the fortresses, the engines have to get to the full 50 moves boundary, to see the draw, with or without full Syzygys. Having a last black pawn move after about 60 moves, this has to be about 110 here all in all.

zullil · Post by **zullil** » Fri Sep 14, 2018 11:14 am

peter wrote: ↑Fri Sep 14, 2018 7:37 am Hi Louis!
zullil wrote: ↑Thu Sep 13, 2018 12:14 pm Pretty study!
Fishing Rod is the easiest one for engines, I guess,

Hi Peter,

I chose to start with Fishing Rod . . . simply because I was intrigued by the name! Was happy that Stockfish solved it, but now you've warned me that other solutions will be hard to come by. Might return to these on and off as time permits, but I'll lower my expectations.

Kornrade Test

Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test

Re: Kornrade Test