SWISSTEST: 1.RobboLito - 2.Rybka - 3.Stockfish

[Glarean]

Hi Walter
He seems to think that two positions of your test are not (fully) correct.
Kurt

Kurt, du langweilst entsetzlich mit solchen Pauschal-Vorwürfen!

Entweder du bringst konkrete Varianten, oder du bist ruhig!

Von dieser bekloppten Troll-Seite kenne ich dich eigentlich gar nicht...
Willst du dich hier einfach "interessant" machen, oder was?!

Walter

.

[Albert Silver]

Hi Walter
He seems to think that two positions of your test are not (fully) correct.
Kurt

Kurt, du langweilst entsetzlich mit solchen Pauschal-Vorwürfen!

Entweder du bringst konkrete Varianten, oder du bist ruhig!

Von dieser bekloppten Troll-Seite kenne ich dich eigentlich gar nicht...
Willst du dich hier einfach "interessant" machen, oder was?!

Walter

After this I admit I have no desire to pursue any discussion. Being rude to Kurt for no reason, and calling CCC, a site for trolls was uncalled for. In any case, in my experience, when people resort to personal attacks and insults, it is because the strength of their arguments is weak.

Your posted games is quite interesting. It would never occur to me the engines could not win it. Especially Rybka. As a counter-argument, I played the position out after Bd7 against Rybka 3 also in 15 minutes (64-bit version on 2 cores at 3.6GHz and 1024 MB hash) and had no trouble winning:

[Event "Computer chess game"]
[Site "ALBERT-PC"]
[Date "2009.12.08"]
[Round "?"]
[White "Albert"]
[Black "Rybka 3"]
[Result "*"]
[Time "14:37:18"]
[WhiteElo "2400"]
[TimeControl "900"]
[SetUp "1"]
[FEN "5k2/5ppp/p7/3p1b2/5P2/5P2/PPBK3P/8 b - - 0 1"]
[Termination "unterminated"]
[PlyCount "47"]
[WhiteType "human"]
[BlackType "program"]

1. ... Bd7 2. Kc3 Ke7 {(Kf8e7 Kc3d4 Ke7d6 Bc2d3 a6a5 Bd3xh7 g7g6 h2h4 Kd6e6 f4f5+ Ke6xf5 h4h5 Bd7e6 f3f4 Kf5xf4 h5xg6 f7xg6 Bh7xg6 Kf4f3 Bg6h5+ Kf3f2 b2b3 Be6g8 Bh5e8 Kf2e2 Be8c6 Ke2d2 Bc6xd5 Bg8h7) -0.81/19 15} 3. Bd3 a5 {(a6a5 b2b4 a5xb4+ Kc3xb4 Ke7d6 a2a4 Bd7h3 a4a5 Kd6c6 Bd3xh7 Bh3g2 Bh7d3 Bg2xf3 Bd3b5+) -0.77/17 20} 4. b4 axb4+ {(a5xb4+ Kc3xb4 Ke7d6 a2a4 g7g6 a4a5 Bd7e6 Bd3b5 Be6f5 a5a6 Bf5c8 h2h3 d5d4 h3h4 f7f6 a6a7) -0.91/19 23} 5. Kxb4 Kd6 {(Ke7d6 a2a4 g7g6 a4a5 f7f6 a5a6 f6f5 Bd3e2 Bd7c6 h2h3 Bc6e8 Be2b5) -0.91/20 22} 6. a4 g6 {(g7g6 a4a5 f7f6 a5a6 f6f5 Bd3e2 Bd7c6 h2h3 Bc6e8 Be2b5) -0.91/21 11} 7. f5 gxf5 {(g6xf5 a4a5 h7h5 f3f4 h5h4 a5a6 Kd6c7 Kb4c5 Bd7e6 Bd3e2) -1.47/19 68} 8. f4 Bc8 {(Bd7c8 a4a5 Kd6c6) -1.58/19 17} 9. a5 Kc6 {(Kd6c6 Bd3b5+ Kc6d6 a5a6 h7h6 h2h4 d5d4 Bb5e2 f7f6 h4h5 d4d3 Be2xd3 Kd6c6 Kb4a5 Bc8e6 Bd3c2 Kc6c7 Bc2b1 Be6d7 Bb1d3 Bd7e6 Ka5b5) -1.58/22 11} 10. Bb5+ Kd6 {(Kc6d6 a5a6 h7h6 h2h4 d5d4 Bb5c4 d4d3 Bc4xd3 Kd6c6 Kb4a5 Bc8e6 Bd3c2 f7f6 Bc2d3 Be6d5 Bd3xf5 Bd5f3 Bf5d3 Bf3d5 Bd3c2) -2.20/24 63} 11. a6 h6 {(h7h6 h2h4 d5d4 Bb5c4 d4d3 Bc4xd3 Kd6c6 Kb4a5 Bc8e6Bd3c2 Be6d5 Bc2xf5 Bd5f3 Bf5b1 f7f6 Bb1a2 Bf3e4 Ba2e6 Kc6d6 Be6g4 Kd6c7) -2.20/24 20} 12. h4 d4 {(d5d4 Bb5c4 d4d3 Bc4xd3 Kd6c6 Kb4a5 Bc8e6 Bd3c2 Be6d5 Bc2xf5 Bd5f3 Bf5b1 f7f6 Bb1a2 Bf3e4 Ba2e6 Kc6d6 Be6g4 Kd6c7) -2.20/24 15} 13. Kc4 d3 {(d4d3) -3.03/20 8} 14. Kxd3 Kc5 {(Kd6c5 a6a7 Bc8b7 Bb5d7 Kc5b6 Kd3d4 f7f6 Bd7c8 Bb7c6 Bc8xf5 Kb6xa7 Kd4c5 Ka7b7 Kc5d6 Kb7b6 Bf5g6 Kb6b5 Kd6e6 f6f5 Bg6xf5 Bc6f3 Ke6e5 Kb5c5 Bf5e4) -4.38/22 24} 15. a7 Bb7 {(Bc8b7) -4.38/21} 16. Bc4 Be4+ {(Bb7e4+ Kd3c3 f7f6 Bc4e6 Kc5b6 Be6xf5 Be4c6 Kc3d4 Kb6xa7 Kd4c5 Ka7b7 Kc5d6 Kb7b6 Bf5g6 Kb6b5 Kd6e6 f6f5 Bg6xf5 Bc6f3 Ke6e5 Kb5c5 Bf5e4 Bf3d1 f4f5) -4.38/22 84} 17. Kc3 f6 {(f7f6) -4.38/22 71} 18. Be6 Kb6 {(Kc5b6 Be6xf5) -4.38/21 60} 19. Bxf5 Bc6 {(Be4c6 Kc3d4) -5.12/20 51} 20. Kd4 Kxa7 {(Kb6xa7) -4.38/18} 21. Kc5 Kb7 {(Ka7b7 Kc5d6) -5.12/18 43} 22. Kd6 Kb6 {(Kb7b6) -4.38/16 0} 23. Ke7 Kc7 {(Kb6c7 Ke7xf6) -5.12/13 4} 24. Kxf6 Kd6 {(Kc7d6 Bf5g6) -5.12/12 36} *

I stopped the game here as the result is obvious.

I didn't bother playing against other engines.

[Jouni]

I took 35 engines, which have CEGT/CCRL ratings and got this list:

solved ELO

52 3133
49 2987
48 2993
46 2953
46 2981
44 2918
44 2981
43 2937
43 2979
43 2889
43 2872
42 2906
42 2758
42 2853
41 2699
41 2902
41 2850
40 2883
38 2855
37 2818
36 2810
35 2761
34 2718
34 2751
33 2642
32 2649
32 2582
31 2689
31 2698
30 2499
29 2670
29 2678
28 2568
27 2651
26 2565

This gives correlation of 0,921! I don't know better short (<100 positions) testsuite. BTW, when I calculated correlation with used time instead of solved positions correlation was to my surprise less = 0,855.

Jouni

[Glarean]

After this I admit I have no desire to pursue any discussion.

Yeah, of course...

Being rude to Kurt for no reason, and calling CCC, a site for trolls was uncalled for.

You didn't read correctly: He attacked without any justification.

But you don't understand this thing. I have known Kurt for many years,
he is a chess friend with many merits, and a friendly man from Switzerland, like me
But sometimes he has a little blackout...
Tomorrow he will be o.k. again ...

[Albert Silver]

After this I admit I have no desire to pursue any discussion.

Yeah, of course...

Being rude to Kurt for no reason, and calling CCC, a site for trolls was uncalled for.

You didn't read correctly: He attacked without any justification.

You are correct. I didn't read it. He attacked? Where? Was the post deleted?

[lmader]

He posted it in German.

Oops, never mind, I misunderstood

[Guenther]

After this I admit I have no desire to pursue any discussion.

Yeah, of course... ;-)

Being rude to Kurt for no reason, and calling CCC, a site for trolls was uncalled for.

You didn't read correctly: He attacked without any justification.

You are correct. I didn't read it. He attacked? Where? Was the post deleted?

Kurt did not attack at all, he just tried to act as an interpreter,
because he had the impression W. Eigenmann missunderstood
your point. This is of course no thankful task, because his behaviour
and stubbornness is wellknown through all chess fora.
I am sorry Albert, but I think you'll never convince him of anything,
at least it never happened in the past and just ended up in mockery.

Guenther

[bob]

I'm not a fan of these types of positions, as if your program is strong enough, it may well see that everything loses, and then choosing the best move is really irrelevant, if you can "see".

I understand this point, but at the same time I would prefer to have the program play to extend the game as long as possible when losing, or to win as quickly as possible when winning. What to do when the position is equal is more interesting.

Here's why: I have personally witnessed hundreds of games where the final result was altered by one side's flag falling. This doesn't just occur at fast time controls either-- it's a real possibility anytime there's a zero increment game.

I should add that a significant percentage of those are not really due to time management issues, but to unexpectedly long games where the result is obvious but altered anyway.

I don't disagree. Obviously, actually, since I originated the "swindle mode" idea years ago which does exactly this (if a position is a forced draw, keep the draw as complicated as possible to give the opponent a chance to make a mistake.) But for dead lost positions, there is always a chance that if all moves are lost, with enough analysis they are all lost in about the same number of moves. Proving that is hard, but a very smart program might see that what was considered an inferior move is actually about as good as the rest...

I don't like the lose in 50 vs lose in 100 move type positions, as they are too deep for me to measure whether they are correct or not.

[Glarean]

Hi Walter
He seems to think that two positions of your test are not (fully) correct.
Kurt

Kurt, du langweilst entsetzlich mit solchen Pauschal-Vorwürfen!

Entweder du bringst konkrete Varianten, oder du bist ruhig!

Von dieser bekloppten Troll-Seite kenne ich dich eigentlich gar nicht...
Willst du dich hier einfach "interessant" machen, oder was?!

Walter

.

Sorry Kurt, I have read wrong your posting: You have just "translated".
Mea maxima culpa!
Nix für ungut!!

Gruss Walter

[Glarean]

Kurt did not attack at all, he just tried to act as an interpreter,
because he had the impression W. Eigenmann missunderstood your point.
This is of course no thankful task, because his behaviour
and stubbornness is wellknown through all chess fora.
I am sorry Albert, but I think you'll never convince him of anything,
at least it never happened in the past and just ended up in mockery.

Yeah, Günther Simon, we've known each other for 10 years, or not?
I remember you: whenever a fight is going on, you come around
and want to join - just like now.
You haven't improved you - pity...

.