Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / Testing of some 7-men tables
1 2 3 4 5 6 Previous Next  
- - By 7-man TB (**) [ru] Date 2012-04-27 19:28 Edited 2012-04-28 17:43
Rook and 2 Pawns  against  Rook and Pawn tables were recently generated ( KRPP - KRP , all possible promotions).
All depending endings ( KR?? - KR? ) were generated too. Before starting next cycle of generation ( KBPP - KBP ) it would be good to make a rough test of the generated tables.

You may post positions you are interested in here and I will provide the answers. Answers will not be instant. I will check the forum several times a day only. If your position is really secret you may send a pm.

Unfortunatelly only part of tables is on Windows computer where the program providing the answers is placed. Most tables are on another computer and can not be used directly without transfering them to work computer. The information returned by the program is very poor. It is number of moves till the longest mate and several moves forward. More suitable procedures will be writen later.

The directly available tables are (post only these positions):
QRP-QR
RBN-RB
RBP-RB
RBP-RN
RBP-RP
RBP-RR
RNP-RB
RNP-RN
RNP-RP
RNP-RR
RPP-RB
RPP-RN
RPP-RP
RPP-RR
RRN-RR
RRP-RB
RRP-RN
RRP-RP
RRP-RR

Surely in EPD format. Other formats will not be considered.

EDIT: The program normalizes positions to provide white to be the strongest side. The definitions of the strongest side are
1) If one side has a pawn and another doesn't have then the side with a pawn is strongest.
2) If both sides have pawns or both have no pawns then
  2a) The side with more pieces is stronger.
  2b) If the number of pieces is equal then the highest pieces are compared one by one till the first mismatch. The side that has stronger mismatched piece is    the strongest.
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-04-27 20:05
Here are a few RPP-RP positions.
 #1 8/r3pk2/1R6/5P2/4PK2/8/8/8 w - - 0 1
 #2 8/2r1kp2/1R6/4PK2/5P2/8/8/8 w - - 0 1
 #3 8/1r3kp1/R7/6P1/5PK1/8/8/8 w - - 0 1
 #4 8/1r3kp1/R7/6P1/5PK1/8/8/8 b - - 0 1
 #5 8/7p/5R2/8/4k1PP/1r6/6K1/8 w - - 0 1
 #6 8/8/5rk1/R6p/4K1p1/6P1/8/8 w - - 0 1
 #7 8/6k1/7p/2R2P2/6PK/8/8/1r6 w - - 0 1
 #8 r7/8/3kP3/3P2p1/2K5/8/8/4R3 w - - 0 1
 #9 7k/6p1/1R3p2/8/r7/6KP/8/8 w - - 0 60
#10 8/5p2/5k2/7R/4P3/5PK1/8/1r6 b - - 0 63
#11 8/8/6p1/R5K1/4k2P/8/p7/r7 w - - 0 1
#12 8/6k1/5p2/8/3P3R/2r1PK2/8/8 w - - 0 46
Parent - - By 7-man TB (**) [ru] Date 2012-04-28 09:30 Edited 2012-04-28 09:56
8/r3pk2/1R6/5P2/4PK2/8/8/8 w - -
Evaluation: DRAW.

8/2r1kp2/1R6/4PK2/5P2/8/8/8 w - -
Evaluation: DRAW.

8/1r3kp1/R7/6P1/5PK1/8/8/8 w - -
Evaluation: 79 half-moves to mate.
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Dadi"]
[Black "79"]
[Result "*"]
[FEN "8/1r3kp1/R7/6P1/5PK1/8/8/8 w - - 0 1"]
[SetUp "1"]

1. f5 $1  1... Rb1 2. Ra7+ $1  2... Kg8 3. Kh5 Rb6 4. Ra8+ Kf7 5. g6+ Ke7 6.
Rg8 Kf6 7. Rf8+ Ke5 8. f6 $1  8... Rxf6 9. Rf7 {[%t bLon] Mate in 32:} 9...
Rf5+ 10. Kg4 $1  10... Rf6 11. Rxg7 Rd6 12. Kg5 Ke6 13. Rf7 Rd3 14. Rf2 Re3
15. g7 Rg3+ 16. Kh6 $1  16... Rh3+ 17. Kg6 Rg3+ 18. Kh7 Rh3+ 19. Kg8 Ke7 20.
Rf4 Rh2 21. Re4+ Kd6 22. Kf7 Rf2+ 23. Kg6 Rc2 24. Rg4 Rc8 25. Kf6 Rg8 26. Rg5
Re8 27. g8=Q Rxg8 28. Rxg8 $1  28... Kc5 29. Rg4 Kd5 30. Rh4 Kc6 31. Rh5 Kd7
32. Rd5+ Kc6 33. Ke6 Kc7 34. Rd6 Kc8 35. Rd7 Kb8 36. Re7 Kc8 37. Kd6 Kd8 38.
Re5 Kc8 39. Rb5 Kd8 40. Rb8# *

8/1r3kp1/R7/6P1/5PK1/8/8/8 b - -
Evaluation: DRAW.

8/7p/5R2/8/4k1PP/1r6/6K1/8 w - -
Evaluation: 101 half-moves to mate.
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Dadi"]
[Black "101"]
[Result "*"]
[FEN "8/7p/5R2/8/4k1PP/1r6/6K1/8 w - - 0 1"]
[SetUp "1"]

1. g5 Ra3 2. Rf7 Ra6 3. Rxh7 Kf4 {[%t bLon] Mate in 48:} 4. Kf1 Kf3 5. Ke1 Kf4
6. Rh8 Re6+ 7. Kd2 Kf5 8. Rf8+ Kg6 9. Kd3 Kg7 10. Rc8 Kg6 11. Kd4 Ra6 12. Kd5
Kf5 13. Rf8+ Kg4 14. Rf6 Ra5+ 15. Ke6 $1  15... Ra7 16. Rh6 $1  16... Ra6+ 17.
Kd7 Ra7+ 18. Kc8 Ra8+ 19. Kb7 Re8 20. g6 Re7+ 21. Kc8 Kf5 22. h5 Kf6 23. Rh8
Rg7 24. Rd8 Kg5 25. Rd5+ Kh6 26. Rd7 Rg8+ 27. Rd8 $1  27... Rg7 28. Rh8+ Kg5
29. Kd8 Ra7 30. Rh7 $1  30... Ra8+ 31. Ke7 Ra7+ 32. Kd6 Ra6+ 33. Kc5 Ra8 34.
g7 Kf6 35. h6 Rc8+ 36. Kd6 Rd8+ 37. Kc7 Ra8 38. Rh8 Ra7+ 39. Kd6 Ra6+ 40. Kc5
Ra5+ 41. Kb4 Ra1 42. g8=Q Rb1+ 43. Kc3 Rc1+ 44. Kd2 Rh1 45. Qg7+ Kf5 46. Rf8+
Ke6 47. Qf6+ Kd5 48. Qg5+ Ke6 49. Rf6+ Kd7 50. Qg7+ Ke8 51. Rf8# *

8/8/6p1/4k1P1/r6P/5RK1/8/8 b - -
Evaluation: DRAW.

8/6k1/7p/2R2P2/6PK/8/8/1r6 w - -
Evaluation: DRAW.

r7/8/3kP3/3P2p1/2K5/8/8/4R3 w - -
Evaluation: DRAW.

8/8/6kp/R7/8/1r3P2/6P1/7K b - -
Evaluation: DRAW.

8/5p2/5k2/7R/4P3/5PK1/8/1r6 b - -
Evaluation: DRAW.

R7/P7/8/4K2p/r5k1/6P1/8/8 b - -
Evaluation: DRAW.

8/6k1/5p2/8/3P3R/2r1PK2/8/8 w - -
Evaluation: DRAW.
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-04-28 16:09 Edited 2012-04-28 16:14
There are some interesting results here. I'll start with position #11:

8/8/6p1/R5K1/4k2P/8/p7/r7 w - - 0 73

[Shown with reversed colors in your results]

This is a position from the ninth game of the World Championship match between Steinitz and Gunsberg, New York 1890-91. The game was drawn, but it has always been thought that Black (Gunsberg) had a winning position here. After 73.Ra4+ he played 73...Kf3, but the endgame manuals state that he could have won with 73...Kd5. See, for instance, Rook Endings (Levenfish & Smyslov) and Analysing the Endgame (Speelman). The same view is posted in the comments to the game on chessgames.com.

Now we learn from your results that this position is actually drawn, so it has taken 122 years to get to the bottom of this position! :cool:
Parent - By Stonehenge (***) Date 2012-04-28 16:16

> Now we learn from your results that this position is actually drawn, so it has taken 122 years to get to the bottom of this position! :cool:


Fascinating.
It's one of the beautiful aspects of chess that even with a very limited number of pieces some positions are unsolvable for human players.
Parent - By 7-man TB (**) [ru] Date 2012-04-28 16:43
Thanks, Dadi!

Very interesting.

And I forgot to mention that output is always from white point of view for the strongest side.
Parent - - By CSullivan (**) [us] Date 2012-05-02 15:52
Here is an interesting note at Amazon.com by Manuel Vargas in his review of Basic Chess Endings:

I have analyzed in depth the endgame Steinitz-Gunsberg (9th match game for the world title, 1890), No. 723, pages 343-344. Despite the revision by Grandmaster Pal Benko the analysis has a serious error and omission. It is evaluated as a win for black, when it is, in fact, a draw. If 3.Kh6 g5 the text only mentions the losing 4.hxg5?. Instead 4. Ra3! draws. The correct analysis and evaluation of this endgame (attributed to I. Zaitsev)appeared on page 11 of the revised german edition of the Levenfish and Smyslov work on rook endgames (Schachverlag Rudi Schmaus, Heidelberg 1985). Besides, I have discovered that 4. Ra8! also draws. However this error and omission is forgivable because Ra3! or Ra8! are extremely unobvious moves and the (rare) german edition of Levenfish's and Smyslovs's work is probably the only one with updated analysis.  (Manuel Vargas, 17 April 2005)
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-05-06 09:29
Thanks. I didn't know about this review or the revised German edition of Levenfish & Smyslov (and clearly some endgame book authors missed them too!). It seems that Manuel Vargas is talking about the position after 1.Ra4 Kd5 2. Ra5 Kc6 3.Kh6 g5.

8/8/2k4K/R5p1/7P/8/p7/r7 w - - 0 66

Manuel Vargas says that Basic Chess Endings only mentions 4.hxg5? (and so does Speelman in the 1988 revision of his book). Instead, Manuel says, both 4.Ra3 (I. Zaitsev, 1985) and 4.Ra8 (Manuel Vargas) draw. The positions below follow the mainline up to move 4 and the final two positions correspond to 4.Ra3 and 4.Ra8. I have inverted the positions so that White has two pawns instead of Black.
R7/P7/8/r3K2p/6k1/6P1/8/8 w - - 0 1
R7/P7/8/r6p/3K2k1/6P1/8/8 b - - 0 1
R7/P7/8/7p/r2K2k1/6P1/8/8 w - - 0 1
R7/P7/8/7p/r5k1/2K3P1/8/8 b - - 0 1
R7/P7/8/7p/r7/2K3Pk/8/8 w - - 0 1
R7/P7/8/7p/r5P1/2K4k/8/8 b - - 0 1
R7/P7/r7/7p/6P1/2K4k/8/8 w - - 0 1
R7/P7/8/7p/6P1/2K4k/8/r7 w - - 0 1
Parent - - By 7-man TB (**) [ru] Date 2012-05-06 21:39
All positions are draws.
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-05-06 21:54
Thanks. This means that the analysis by I. Zaitsev, quoted in the revised German edition of Levenfish & Smyslov (1985), was correct (4.Ra3 draws) and so was Manuel Vargas about 4.Ra8 drawing.
Parent - By Pia (****) [ru] Date 2012-05-07 10:17 Edited 2012-05-07 10:20
FinalGen was also right saying "black does not lose" (unsure) for these two moves. :razz:
But simple "search for draw" the program failed.
Parent - By Dadi Jonsson (Silver) [is] Date 2012-04-29 09:55
Position #12 occurred in the game Grivas, Efstratios – Soylu, Suat, Varna Balkaniad, 1994.

8/6k1/5p2/8/3P3R/2r1PK2/8/8 w - -

Grivas won the game and later he analyzed this position in his book Practical Endgame Play – Mastering the Basics (Everyman 2008). Karsten Müller wrote about this endgame in his Endgame Corner on ChessCafe.com (Central Pawns):

> Grivas managed to win in an instructive way, but can this position really be called won already? I had my doubts and I sent him a few ideas. He finally came up with a corrected version and with his kind permission I now hand over the microphone to him.


The rest of his column is dedicated to the improved analysis. The position above is drawn as the TB7 results show and Grivas agrees with that in his analysis.

46.Kf4 Rd3 47.Rh1 Kf7 48.Ke4
8/5k2/5p2/8/3PK3/3rP3/8/7R b - - 0 48

48...Rd2

Grivas believed that this was the losing move and Black should have played 49...Ra2 instead. However, a month later Müller published an addendum with analysis from the excellent analyst Charles Sullivan (who occasionally posts here on the forum) where he stated that 48...Rd2 was still playable. This has now been confirmed by TB7.

49.Kd5
8/5k2/5p2/3K4/3P4/4P3/3r4/7R b - - 0 49

49...Ra2?
Grivas doesn't comment on this move, but Sullivan gives it a question mark and says that Black could still have saved himself with 49...Re2. The TB7 results above confirm that 49...Ra2 is a losing move, but I didn't ask for analysis of the position after 49...Re2 to see if that was the saving move (although I wouldn't be surprised if Sullivan is correct).

50.Rc1?
Grivas gives this move an exclamation mark, but Sullivan correctly believes that it is a mistake leading to a draw. Instead he suggests that 50.Rh7+ will win the game and in fact that is the only winning move according to TB7.

50...Ra5
Sullivan writes: "The final mistake. 50...Re2 still holds". TB7 confirms that 50...Ra5 loses, so Sullivan is correct here, too.

51.Rc1
The final position analyzed with TB7 shows that White's shortest win here starts with 51.Kd6, leading to a mate in 38 moves. I don't have the statistics for 51.Rc1.
Parent - - By cma6 (****) Date 2012-04-29 19:28
Dadi,
Could you explain how you generated those FEN position descriptions?
            Thanks, CMA
Parent - By Dadi Jonsson (Silver) [is] Date 2012-04-29 20:08
Assuming that you are using Aquarium, either:

1) Click 'Copy as EPD' in the Home-View tab or

2) Right-click in the notation and select 'Copy position as EPD' (or use the keyboard shortcut shown in the menu)

Paste into your post on the forum and the diagram will be displayed automatically. If you just want to display the FEN string (as I did above), click 'Insert raw text' and paste there.
Parent - By Lukas Cimiotti (Bronze) [de] Date 2012-05-01 16:01
Rybka finds a mate in position #3:

8/1r3kp1/R7/6P1/5PK1/8/8/8 w - - 0 1


Analysis by Rybka 4.1:

1.f5 Rd7 2.Kf4 Rc7 3.g6+ Kf8 4.Kg5 Rb7 5.f6 Rb5+ 6.Kf4
  +-  (1.75)   Depth: 6   00:00:01  0kN
1.f5 Rd7 2.Kf4 Rc7 3.g6+ Kf8 4.Kg5 Rb7 5.Re6 Rb5 6.Rc6
  +-  (1.75)   Depth: 7   00:00:01  0kN
1.f5 Rd7 2.Kf4 Rc7 3.g6+ Kf8 4.Kg5 Rb7 5.Kg4 Rb4+ 6.Kg5 Rb3 7.f6 Rb5+ 8.Kf4 gxf6
  +-  (1.75)   Depth: 8   00:00:01  0kN
1.f5 Rd7 2.Kf4 Rc7 3.g6+ Kf8 4.Kg5 Rb7 5.Kg4 Rb4+ 6.Kg5 Rb3 7.Ra8+ Ke7 8.Rg8 Rg3+ 9.Kf4 Rc3 10.Rxg7+ Kf8
  +-  (1.75)   Depth: 9   00:00:01  0kN
1.f5 Rd7 2.Kf4 Rc7 3.g6+ Ke7 4.Kg5 Rc2 5.Ra7+ Kf8 6.Rf7+ Kg8 7.Rb7 Rg2+ 8.Kf4 Rf2+ 9.Ke4
  +-  (1.75)   Depth: 10   00:00:01  0kN
1.f5 Rd7 2.g6+ Ke7 3.Kf4 Kf8 4.Ra8+ Ke7 5.Ke5 Rd2 6.Ra7+ Kf8 7.Rf7+ Kg8 8.Rc7 Re2+ 9.Kf4 Rf2+ 10.Kg4 Kf8 11.Rf7+
  +-  (1.79)   Depth: 11   00:00:01  1kN
1.f5 Rd7 2.g6+ Ke7 3.Kf4 Kf8 4.Ra8+ Ke7 5.Ke5 Rd2 6.Ra7+ Kf8 7.Rf7+ Kg8 8.Rc7 Re2+ 9.Kf4 Rf2+ 10.Ke4 Kf8
  +-  (1.78)   Depth: 12   00:00:01  1kN
1.f5 Rd7 2.g6+ Ke7 3.Kf4 Rd4+ 4.Ke5 Rd1 5.Ra7+ Kf8 6.Rf7+ Kg8 7.f6 Re1+ 8.Kf5 Rf1+ 9.Kg5 Rg1+ 10.Kf5 Rf1+ 11.Kg5 Rg1+ 12.Kf5 Rf1+ 13.Kg5 Rg1+ 14.Kf5 Rf1+ 15.Kg5 Rg1+ 16.Kf5
  +-  (1.78)   Depth: 13   00:00:02  7kN, tb=10
1.f5 Rd7 2.Kh5 Rc7 3.g6+ Kf8 4.Kg5 Rc8 5.Ra7 Rb8 6.Rf7+ Kg8 7.Rd7 Kf8 8.Kf4 Ra8 9.Rf7+ Kg8 10.Rb7 Kf8 11.Ke5
  +-  (1.79)   Depth: 14   00:00:02  26kN, tb=51
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+ Kf7 5.Rc8 Ra6 6.g6+ Ke7 7.Rc7+ Kf8 8.Rf7+ Kg8 9.Rd7 Ra8 10.Kg5 Kf8 11.Rf7+ Kg8 12.Rb7 Kf8 13.Rf7+ Kg8 14.Rb7 Kf8 15.Rf7+ Kg8 16.Rb7
  +-  (1.90)   Depth: 15   00:00:02  51kN, tb=94
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+ Kf7 5.Rc8 Ra6 6.g6+ Ke7 7.Rc7+ Kf8 8.Rf7+ Kg8 9.Rd7 Ra8 10.Kg5 Kf8 11.Rf7+ Kg8 12.Rb7 Kf8 13.Rf7+ Kg8 14.Rb7 Kf8 15.Rf7+ Kg8 16.Rb7
  +-  (2.02 !)   Depth: 16   00:00:02  64kN, tb=120
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+ Kf7 5.Rc8 Ra6 6.g6+ Ke7 7.Rc7+ Kf8 8.Rf7+ Kg8 9.Rd7 Ra8 10.Kg5 Kf8 11.Rf7+ Kg8 12.Rb7 Kf8 13.Rf7+ Kg8 14.Rb7 Kf8 15.Rf7+ Kg8 16.Rb7
  +-  (2.17 !)   Depth: 16   00:00:02  82kN, tb=155
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+ Kf7 5.Rc8 Ra6 6.g6+ Ke7 7.Rc7+ Kf8 8.Rf7+ Kg8 9.Rd7 Ra8 10.Kg5 Kf8 11.Rf7+ Kg8 12.Rb7 Kf8 13.Rf7+ Kg8 14.Rb7 Kf8 15.Rf7+ Kg8 16.Rb7
  +-  (2.57 !)   Depth: 16   00:00:02  121kN, tb=197
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+ Ke7[] 5.f6+ gxf6[] 6.g6 Rb1 7.g7[] Rh1+ 8.Kg6[] Rg1+[] 9.Kh7[] Rh1+[] 10.Kg8[] Rd1 11.Ra7+ Ke6[] 12.Ra4 f5 13.Ra8 Ke7 14.Ra7+ Kf6 15.Ra8 f4 16.Kf8
  +-  (3.37 !)   Depth: 16   00:00:02  167kN, tb=246
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5 Rb6 4.Ra8+[] Ke7 5.f6+ gxf6 6.g6[] Rb1 7.g7[] Rh1+ 8.Kg6[] Rg1+ 9.Kh7[] Rh1+ 10.Kg8[] Rc1 11.Ra7+ Ke6 12.Ra4 f5 13.Ra6+ Ke5 14.Kf7 Rc8 15.g8Q Rxg8 16.Ra5+
  +-  (4.15)   Depth: 16   00:00:02  258kN, tb=344
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+[] Ke7 5.f6+ gxf6 6.g6[] Rb5+ 7.Kh6[] Rb4 8.Ra7+ Ke6 9.g7 Rh4+ 10.Kg6[] Rg4+[] 11.Kh7 Rh4+ 12.Kg8[] Rc4 13.Ra1 f5 14.Ra6+ Ke5 15.Kf7 Rc8 16.Rc6
  +-  (4.15)   Depth: 17   00:00:02  295kN, tb=412
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+ Ke7 5.f6+ gxf6 6.g6 Rb5+ 7.Kh6 Rb4 8.Ra7+ Ke6 9.g7 Rh4+ 10.Kg6[] Rg4+[] 11.Kh7 Rh4+ 12.Kg8[] Rc4 13.Ra1 f5 14.Ra6+ Ke5 15.Kf7 Rc8 16.Ra5+
  +-  (4.27 !)   Depth: 18   00:00:03  382kN, tb=539
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+ Ke7 5.f6+ gxf6 6.g6 Rb5+ 7.Kh6 Rb4 8.Ra7+ Ke6 9.g7 Rh4+ 10.Kg6[] Rg4+[] 11.Kh7 Rh4+ 12.Kg8[] Rc4 13.Ra1 f5 14.Ra6+ Ke5 15.Ra8 Rg4 16.Ra5+
  +-  (4.42 !)   Depth: 18   00:00:03  501kN, tb=771
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+ Ke7 5.f6+ gxf6 6.g6 Rb5+ 7.Kh6 Rb4 8.Ra7+ Ke6 9.g7 Rh4+ 10.Kg6[] Rg4+[] 11.Kh7 Rh4+ 12.Kg8[] Rc4 13.Ra1 f5 14.Ra6+ Ke5 15.Ra8 Rg4 16.Ra5+
  +-  (4.82 !)   Depth: 18   00:00:03  672kN, tb=1004
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+ Ke7 5.f6+ gxf6 6.g6 Rb5+ 7.Kh6 Rb4 8.Ra7+ Ke6[] 9.g7[] Rh4+ 10.Kg6[] Rg4+[] 11.Kh7 Rh4+[] 12.Kg8[] Rc4 13.Ra1 f5 14.Ra6+ Ke5 15.Ra8 Rg4 16.Ra5+
  +-  (5.62 !)   Depth: 18   00:00:03  915kN, tb=1315
1.f5 Rb1 2.Ra7+ Kf8 3.Kh5[] Rb6 4.Ra8+[] Kf7 5.g6+[] Ke7 6.Rg8[] Kf6 7.Rf8+[] Ke7 8.Rf7+[] Ke8 9.Rxg7 Ra6 10.f6 Kf8 11.Rh7 Rxf6 12.Kh6
  +-  (7.08)   Depth: 18   00:00:04  1240kN, tb=1750
1.f5 Rb1 2.Ra7+[] Kf8 3.Kh5[] Rb6 4.Ra8+ Kf7 5.g6+ Ke7 6.Rg8 Kf6 7.Rf8+ Ke7 8.Rf7+ Kd6 9.Rxg7 Rb2 10.Rf7 Ke5 11.g7 Rg2 12.Kh6 Ke4 13.Rf8
  +-  (7.08)   Depth: 19   00:00:04  1321kN, tb=1944
1.f5 Rb1 2.Ra7+[] Kf8 3.Kh5[] Rb6 4.Ra8+ Kf7 5.g6+ Ke7 6.Rg8 Kf6 7.Rf8+ Ke7[] 8.Rf7+ Kd6 9.Rxg7 Rb2 10.Rf7 Ke5 11.g7 Rg2 12.Kh6 Ke4 13.Rf8 Rh2+ 14.Kg6 Rg2+ 15.Kf6 Kf4 16.Ra8
  +-  (7.08)   Depth: 20   00:00:04  1502kN, tb=2492
1.f5 Rb1 2.Ra7+[] Kf8 3.Kh5[] Rb6 4.Ra8+ Kf7 5.g6+ Ke7 6.Rg8 Kf6 7.Rf8+ Ke7 8.Rf7+ Kd6 9.Rxg7 Rb2 10.Rf7 Rh2+ 11.Kg5 Ke5 12.Re7+ Kd6 13.f6 Rg2+ 14.Kh6 Rh2+ 15.Kg7 Rf2 16.Re8
  +-  (11.77)   Depth: 21   00:00:09  6058kN, tb=6180
1.f5 Rb1 2.Ra7+[] Kf8 3.Kh5[] Rb6 4.Ra8+ Kf7 5.g6+ Ke7 6.Rg8 Kf6 7.Rf8+ Ke7 8.Rg8 Kf6 9.Rf8+ Ke7 10.Rg8 Kf6 11.Rf8+ Ke7 12.Rg8 Kf6 13.Rf8+ Ke7 14.Rg8 Kf6 15.Rf8+ Ke7 16.Rg8
  +-  (12.31)   Depth: 22   00:00:16  13549kN, tb=11141
1.f5 Rb1 2.Ra7+ Kg8 3.Kh5 Rb6[] 4.Ra8+ Kf7[] 5.g6+ Ke7 6.Rg8[] Kf6[] 7.Rf8+[] Ke7[] 8.Rf7+[] Kd6 9.Rxg7[] Rb2 10.f6 Rh2+ 11.Kg4 Rg2+ 12.Kf3 Ra2 13.f7 Kd5 14.f8Q Ra6 15.Rd7+ Kc4 16.g7
  +-  (#39)   Depth: 23   00:00:48  68789kN, tb=27552
1.f5 Rb1 2.Ra7+ Kg8 3.Kh5 Rb6 4.Ra8+ Kf7[] 5.g6+ Ke7 6.Rg8[] Kf6 7.Rf8+ Ke5[] 8.f6 Rxf6 9.Rf7 Ra6 10.Rxg7
  +-  (#40)   Depth: 24   00:01:13  115mN, tb=46530
1.f5 Rb2 2.Ra7+[] Kg8 3.Kh5 Rb6 4.Ra8+ Kf7 5.g6+ Ke7 6.Rg8[] Kf6 7.Rf8+[] Ke5 8.f6 Rxf6 9.Rf7 Ra6 10.Rxg7
  +-  (#40)   Depth: 25   00:01:13  115mN, tb=46604
1.f5 Rb2 2.Ra7+[] Kg8 3.Kh5 Rb6 4.Ra8+ Kf7 5.Ra3 Ke7 6.Ra8 Kf7 7.Ra3 Ke7 8.Ra8 Kf7 9.Ra3 Ke7 10.Ra8 Kf7 11.Ra3 Ke7 12.Ra8 Kf7 13.Ra3 Ke7 14.Ra8 Kf7 15.Ra3 Ke7 16.Ra8
  +-  (#40)   Depth: 26   00:01:13  115mN, tb=46793
1.f5 Rb2 2.Ra7+[] Kg8 3.Kh5 Rb6 4.Ra8+ Kf7 5.g6+[] Ke7 6.Rg8[] Kf6 7.Rf8+[] Ke5 8.f6 Rxf6 9.Rf7 Ra6 10.Rxg7
  +-  (#40)   Depth: 27   00:01:20  129mN, tb=58985
Parent - - By 7-man TB (**) [ru] Date 2012-04-28 10:19
This is the longest mate when one of the white pawns stays not further than on 4th rank. Pawn endings are sliced, so it is easy to discover the longest mate for pawns on specific lines.
The position is interesting due the fact that 2.Re7 and 3.Kb4 are the only moves that win the game. Houdini doesn't see these moves (at least at fast control).

8/1R6/6p1/4r2k/3P4/2KP4/8/8 b - -

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "RPP-RP (4)"]
[Black "268"]
[Result "*"]
[FEN "8/1R6/6p1/4r2k/3P4/2KP4/8/8 b - - 0 1"]
[SetUp "1"]

1... Ra5 2. Re7 $1  2... g5 3. Kb4 $1  3... Ra6 4. d5 g4 5. Kc5 $1  5...
Ra5+ 6. Kd4 g3 7. Rg7 $1  7... Ra4+ 8. Ke3 $1  8... Rg4 9. Rxg4 Kxg4 {[%t
bLon] Mate in 126:} 10. d6 $1  10... Kh3 11. d7 $1  11... g2 12. Kf2 $1  12...
Kh2 13. d8=Q g1=Q+ 14. Kf3 Qg2+ 15. Ke3 $1  15... Qg1+ 16. Ke4 Qg2+ 17. Kd4
Qd2 18. Kc4 Qc2+ 19. Kb5 $1  19... Qb3+ 20. Kc6 Qc2+ 21. Kb7 Qg2+ 22. Ka7 Qg7+
23. Ka6 Qg6+ 24. Kb5 $1  24... Kg3 25. Kc5 $1  25... Qf5+ 26. Kc6 $1  26...
Kf2 27. Qd4+ $1  27... Kf1 28. Kc7 $1  28... Qa5+ 29. Kd7 Qf5+ 30. Ke7 Qh7+
31. Kd8 Qg8+ 32. Kc7 Qf7+ 33. Qd7 $1  33... Qf4+ 34. Qd6 $1  34... Qf7+ 35.
Kc6 Qf3+ 36. Qd5 Qf6+ 37. Kd7 Qg7+ 38. Ke8 Qh8+ 39. Ke7 Qh4+ 40. Kd7 Qg4+ 41.
Kd6 Qf4+ 42. Qe5 Qf8+ 43. Kc7 Qf7+ 44. Kc6 $1  44... Qf3+ 45. Qe4 Qf8 46. Kb7
Qg7+ 47. Ka6 Qf6+ 48. Kb5 Qb2+ 49. Kc4 Qc1+ 50. Kd4 Qb2+ 51. Kd5 Qb5+ 52. Kd6
$1  52... Qb8+ 53. Ke7 Qc7+ 54. Kf6 Qd8+ 55. Kf7 Qc7+ 56. Kg6 Qg3+ 57. Kf6
Qd6+ 58. Kf5 Qf8+ 59. Kg4 Qg7+ 60. Kf3 Qf6+ 61. Qf4 Qc3 62. Ke4+ Ke2 63. Qe3+
Kd1 64. d4 Qc6+ 65. Kf5 Qd7+ 66. Kg6 Qd6+ 67. Kh5 Qh2+ 68. Kg5 Qg2+ 69. Kf6
Qd5 70. Qe5 Qa8 71. Qc5 Qd8+ 72. Kf5 Qd7+ 73. Kf4 Qf7+ 74. Qf5 Qc4 75. Qe4
Qc7+ 76. Qe5 Qc1+ 77. Kg4 Qc3 78. Kh4 Kd2 79. Qf4+ $1  79... Ke2 80. Qe4+ Kd2
81. d5 Qf6+ 82. Kh5 Qh8+ 83. Kg5 Qg7+ 84. Qg6 Qe5+ 85. Qf5 $1  85... Qg3+ 86.
Kf6 Qd6+ 87. Kf7 Qc7+ 88. Ke6 Qc8+ 89. Ke5 Qc5 90. Qf4+ Kd3 91. Qe4+ Kc3 92.
Ke6 Qc8+ 93. Kf7 Qd7+ 94. Kg6 Qd6+ 95. Qe6 Qg3+ 96. Kf7 Qc7+ 97. Ke8 Qb8+ 98.
Kd7 Qa7+ 99. Kd8 Qa5+ 100. Ke8 Qa8+ 101. Kf7 Qa7+ 102. Qe7 Qa2 103. Qe3+ Kb4
104. Qe6 Qf2+ 105. Ke8 Qc5 106. d6 Qh5+ 107. Ke7 Qh4+ 108. Kd7 Ka5 109. Qd5+
Kb6 110. Qb3+ Kc5 111. Qe3+ Kb5 112. Qe8 Qg3 113. Kc7+ Kb4 114. Qc6 Ka3 115.
Kb6 Qb3+ 116. Qb5 Qg8 117. Qc5+ Ka2 118. d7 Qd8+ 119. Kc6 Qf6+ 120. Kb7 Qf3+
121. Ka7 Qf7 122. Qc7 Qf2+ 123. Kb8 Qb2+ 124. Kc8 Qd4 125. d8=Q Qg4+ 126. Qcd7
Qc4+ 127. Kb7 Qb3+ 128. Ka6 Qc4+ 129. Qb5 Qe6+ 130. Qdb6 Qxb6+ 131. Qxb6 Ka3
132. Qb1 Ka4 133. Kb6 Ka3 134. Kc5 Ka4 135. Qa2# *
Parent - - By Stonehenge (***) Date 2012-04-28 14:46
Interesting position, it takes a while (about 10 minutes using 3 cores) for Houdini to find Re7 and to see that it's winning.
After 1...Ra5 the following analysis is produced:
8/1R6/6p1/r6k/3P4/2KP4/8/8 w - -

Engine: Houdini DEV 3T EGTB (2048 MB)
by Robert Houdart

27/54  0:08   +0.71    2.Kc4 Kg5 3.Rf7 Ra4+ 4.Kd5 Ra2 5.Ke6 Re2+
                       6.Kd6 Re3 7.d5 Rxd3 8.Ke5 Re3+ 9.Kd4 Re1
                       10.d6 Kg4 11.d7 Rd1+ 12.Ke5 g5
                       13.Ke4 Re1+ 14.Kd4 Rd1+ 15.Ke5 (54.338.493) 6072
                       TB:23.606

33/60  3:06   +0.39    2.Kc4 Kg5 3.Rf7 Ra4+ 4.Kd5 Ra2 5.Kc6 Rc2+
                       6.Kd7 Rd2 7.d5 Rxd3 8.d6 Kh4 9.Ke6 Re3+
                       10.Kf6 Rd3 11.d7 g5 12.Rh7+ Kg4
                       13.Ke5 Rd1 14.Rf7 Kh4 15.Kf5 (1.185.729.020) 6365
                       TB:762.892

34/64  5:52   +0.27    2.Kc4 Kg5 3.Rf7 Ra4+ 4.Kd5 Ra2 5.Kc6 Rc2+
                       6.Kd7 Rd2 7.d5 Rxd3 8.d6 Kg4 9.Ke6 g5
                       10.d7 Re3+ 11.Kf6 Rd3 12.Ke5 Kh4
                       13.Ke4 Rd1 14.Rg7 g4 15.Kf5 (2.254.936.613) 6397
                       TB:1.499.862

34/64  8:14   +0.44++  2.Re7 (3.212.073.164) 6501
                       TB:1.996.928

34/64  8:57   +0.57++  2.Re7 (3.497.488.156) 6512
                       TB:2.119.133

34/65  9:53   +0.87++  2.Re7 (3.875.234.825) 6525
                       TB:2.281.539

34/73  13:43  +1.82++  2.Re7 (5.448.633.402) 6614
                       TB:3.044.095

34/91  25:16  +2.40    2.Re7 g5 3.Kb4 Rd5 4.Kc4 Rd8 5.d5 Kg6
                       6.Kc5 Rc8+ 7.Kd4 Rd8 8.Re3 Kf5
                       9.Rf3+ Kg6 10.Kc5 g4 11.Re3 Rc8+
                       12.Kb5 Rd8 13.Kc6 Rc8+ 14.Kd7 Rf8
                       15.Rg3 (10.418.821.912) 6870
                       TB:4.597.378
Parent - - By 7-man TB (**) [ru] Date 2012-04-28 14:53
Thanks for the analysis, I was lazy to link 6-men bases and check this ending.

Probably number of threads is not very important for speed when 6-men tables are used.

Funny, that even with tables the engine is not 100% sure that white wins.
Parent - - By Stonehenge (***) Date 2012-04-28 14:54
I forgot to mention that in the analysis above Houdini only used 5-men table bases.
Parent - - By Barnard (Bronze) Date 2012-04-28 15:20

>EGTB


you forgot to mention Robert?why you need to mention?if you are using Gaviota tablebases,and you are using it,Gaviota tablebases only had generated 5 men tablebases,so you dont need mention it Robert

or you mention it to recall about how 'powerful' is analizing Houdini the endgame,only using 5 men tablebase,and not the 6 men tablebase? :wink:
Parent - - By Stonehenge (***) Date 2012-04-28 15:41 Edited 2012-04-28 15:44
Houdini 2 can use both Gaviota (up to 5-men) and Nalimov (up to 6-men) EGTB.
Parent - - By Barnard (Bronze) Date 2012-04-28 15:45
yes,but in your analisis you were using Gaviota,not Nalimov

edit:correct me if im wrong Robert,but i really doubt it
Parent - - By Stonehenge (***) Date 2012-04-28 15:58
You are wrong.

I really don't understand the point of your negative intervention.
I mentioned the 5-men EGTB in reply to a post that assumed that I was using 6-men.
Parent - By Barnard (Bronze) Date 2012-04-28 16:05 Edited 2012-04-28 16:09
my negative intervention?do you think im trying to denostate Houdini?i think you are wrong about that point Robert

im who are supporting Houdini against others opinions playing correspondece chess,and im who are playing a 10 game correspondence  match between 'Houdini' against 'all other engines' (you can see the match under the correspondence subforum,''Barnard-Uly,Houdini against other engines'',im Barnard),just to prove Houdini is stronger than all other engines at very long time controls like correspondence chess

so what is your point?that im trying to benefit other engine?be sure that engine wont be rybka,you can have it for sure
Parent - By siah (***) Date 2012-04-30 07:03
Why don't you answer PMs??
Parent - - By kgburcham (***) [us] Date 2012-04-29 14:55 Edited 2012-04-29 15:08
Robert does the Dev stand for developement version?
Big difference in depth, kns, and total nodes, but the 2.0c version cannot get to the +2.4 score as your new version does.
I tried several times with a reboot each time with same results.

Houdini DEV 3T EGTB (2048 MB)

34/64  8:14   +0.44++  2.Re7 (3.212.073.164) 6501
                       TB:1.996.928
34/64  8:57   +0.57++  2.Re7 (3.497.488.156) 6512
                       TB:2.119.133
34/65  9:53   +0.87++  2.Re7 (3.875.234.825) 6525
                       TB:2.281.539
34/73  13:43  +1.82++  2.Re7 (5.448.633.402) 6614
                       TB:3.044.095
34/91  25:16  +2.40    2.Re7 g5 3.Kb4 Rd5 (10.418.821.912) 6870
                       TB:4.597.378

Houdini 2.0c Pro x64 (4096 MB)

33/71   0:58   +0.62   1.Kc4 Kg5 2.Rf7 Ra4+ 3.Kd5 Ra2 4.Rf1 Rc2 5.Ke6 Re2+ (2.173.853.593) 37070  TB:170.876
33/71   1:36   +0.71   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ (3.657.828.080) 37727  TB:275.444
34/71   2:05   +0.78++  1.Re7 (4.795.932.334) 38242  TB:365.957
34/75   3:12   +0.91++  1.Re7 (7.495.108.105) 38911  TB:544.006
34/75   5:08   +1.03   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ (12.222.521.879) 39593  TB:808.145
35/78   8:58   +1.11++  1.Re7 (21.960.359.902) 40796  TB:1.379.727
35/78  14:56   +1.21   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 g4 5.Kc5 Rc8+ (37.499.234.586) 41813  TB:2.196.754
35/78  16:29   +1.21   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 g4 5.Kc5 Rc8+ (41.510.127.521) 41951  TB:2.362.127

Here is another example to compare to your dev version
Low score
6 threads 980x
Houdini 2.0c Pro x64 (4096 MB)

32/71   1:32   +0.42--  1.Kc4 Kg5 (1.769.837.450) 19213  TB:147.804
32/74   2:13   +0.63   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (2.580.648.085) 19342  TB:213.802
33/74   2:18   +0.56--  1.Re7 g5 (2.672.504.205) 19302  TB:220.398
33/74   2:54   +0.64   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (3.396.003.391) 19421  TB:272.315
34/74   3:53   +0.68   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (4.590.574.565) 19634  TB:349.189
35/74   4:15   +0.68   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (5.009.414.723) 19630  TB:382.869
36/74   4:47   +0.68   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (5.652.872.081) 19651  TB:436.838
37/75   6:34   +0.68   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (7.766.637.432) 19662  TB:570.443
38/75   8:45   +0.68   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (10.318.253.872) 19650  TB:733.936
39/79  18:16   +0.74++  1.Re7 (22.227.979.280) 20276  TB:1.490.947
39/79  37:13   +0.87++  1.Re7 (46.249.353.777) 20711  TB:2.737.798
39/79  37:56   +0.87   1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb6 Rd8 7.Kc6 Rc8+ (47.153.349.447) 20713  TB:2.771.120
Parent - - By Stonehenge (***) Date 2012-04-29 15:49
Yes, it's the latest DEV version (which is whatever happens to be compiled on the server at the time of the test).
I've run the position again with today's DEV version (slightly different from yesterday's) and it produces a high score as well.

8/1R6/6p1/r6k/3P4/2KP4/8/8 w - -

Engine: Houdini DEV 3T EGTB (2048 MB)
by Robert Houdart

33/64  5:30   +0.25    1.Kc4 Kg5 2.Rf7 Ra4+ 3.Kd5 Ra1 4.Ke5 Re1+
                       5.Kd6 Rd1 6.d5 Rxd3 7.Ke5 Re3+ 8.Kd4 Re1
                       9.d6 Kg4 10.d7 Rd1+ 11.Ke5 g5 12.Ke4 Kh4
                       13.Kf5 Rf1+ 14.Ke5 (2.300.435.618) 6967
                       TB:1.331.275

33/65  6:45   +0.30++  1.Re7 (2.835.600.824) 6990
                       TB:1.573.972

33/65  7:01   +0.39++  1.Re7 (2.953.577.272) 6999
                       TB:1.611.490

33/74  7:28   +0.67++  1.Re7 (3.140.169.947) 7008
                       TB:1.672.430

33/101 11:37  +1.58++  1.Re7 (4.999.279.293) 7165
                       TB:2.437.884

33/101 19:27  +3.07    1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6
                       5.Kc5 Rc8+ 6.Kb5 Rb8+ 7.Kc6 Rc8+
                       8.Rc7 Ra8 9.d6 Kf6 10.Kd5 g4 11.Rc4 g3
                       12.Rf4+ Kg5 13.Rf7 Rg8 14.d7 (8.600.333.160) 7367
                       TB:3.609.490

As yesterday analysis with 3 threads of Core i5-750, using 2048 MB hash and 5-men EGTB.
Compared to Houdini 2.0c I haven't made any significant changes in the handling of rook end games, this must be a random fluke or a positive side-effect of other changes.
Parent - - By kgburcham (***) [us] Date 2012-04-29 15:58
33/101

what is the 101 called, Extensions?
that is one difference in comparing Dev to 2.0c.
Parent - - By Stonehenge (***) Date 2012-04-29 20:18
Correct, 101 is the deepest position in the search tree - 101 plies or 50 full moves from the starting position.
The fact that it's different from your Houdini 2.0c analysis is probably just random - SMP analysis is not very predictable or repeatable.
Parent - - By Carl Bicknell (*****) [gb] Date 2012-05-02 16:51
Well even though you haven't made any big changes in R&P endings I take one lesson away from this: Software triumphs over Hardware.

Despite much faster hardware and great depths, I couldn't match your analysis:

Analysis by Houdini 2.0c Pro x64 (12 cores)

  +/=  (0.51 --)   Depth: 29/64   00:00:15  423mN, tb=15090
1.Kc4 Kg5 2.d5 Kf6 3.Rb2 Ra4+ 4.Kc5 Ra5+ 5.Kc6 Ra6+ 6.Rb6 Ra7 7.d4 g5 8.d6 Ra4 9.Rb5 Rxd4 10.Rd5 Rc4+ 11.Kb6 Rb4+ 12.Kc7 Rc4+ 13.Kb7 Rb4+ 14.Kc6 Rc4+ 15.Kb6
  +/=  (0.53)   Depth: 29/64   00:00:21  621mN, tb=25092
1.Kc4 Kg5 2.d5 Kf6 3.Rb2 Ra4+ 4.Kc5 Ra5+ 5.Kc6 Ra6+ 6.Rb6 Ra7 7.d4 g5 8.d6 Ra4 9.Rb5 Rxd4 10.Rd5 Rc4+ 11.Kb6 Rb4+ 12.Kc7 Rc4+ 13.Kb7 Rb4+ 14.Kc6 Rc4+ 15.Kb6
  +/=  (0.53)   Depth: 30/64   00:00:23  673mN, tb=28103
1.Kc4 Kg5 2.d5 Kf6 3.Rb2 Ra4+ 4.Kc5 Ra5+ 5.Kc6 Ra6+ 6.Rb6 Ra7 7.d4 g5 8.d6 Ra4 9.Rb5 Rxd4 10.Rd5 Rc4+ 11.Kb6 Rb4+ 12.Kc7 Rc4+ 13.Kb7 Rb4+ 14.Kc6 Rc4+ 15.Kb6
  +/=  (0.53)   Depth: 31/64   00:00:28  798mN, tb=37608
1.Kc4 Kg5 2.d5 Kf6 3.Rb2 Ra4+ 4.Kc5 Ra3 5.d4 g5 6.Re2 Ra5+ 7.Kc6 Ra4 8.Rf2+ Kg6 9.Rd2 Kf7 10.d6 Ke8 11.Rb2 Ra6+ 12.Kd5 Ra3 13.Rb8+ Kd7 14.Rb7+ Kd8 15.Ke6 Re3+ 16.Kf5 Rd3 17.Ke4 Rc3 18.Rg7 Rg3 19.Ke5 Re3+ 20.Kf5 g4 21.Rxg4 Kd7 22.Rg6 Re8 23.Rf6 Re1 24.Rh6 Re2 25.d5 Rd2 26.Ke5 Re2+ 27.Kd4 Rd2+ 28.Ke4 Re2+ 29.Kd3 Rf2 30.Kd4 Rd2+ 31.Ke4
  +/=  (0.54)   Depth: 32/64   00:00:44  1292mN, tb=68023
1.Kc4 Kg5
  +/=  (0.48 --)   Depth: 33/65   00:01:00  1785mN, tb=108190
1.Kc4 Kg5 2.d5 Kf6 3.Rb6+ Kf5 4.Rb2 Ra4+ 5.Kc5 Ra5+ 6.Rb5 Ra7 7.Rb1 Ra5+ 8.Kc6 Ra6+ 9.Kb5 Ra3 10.Kc4 Ra4+ 11.Kc5 Ra5+ 12.Kc6
  +/=  (0.42)   Depth: 33/65   00:01:44  3184mN, tb=215849
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Rc7 Ra8 9.d6 Kf6 10.Rb7 Ra6+ 11.Rb6 Ra3 12.d4 Ra4 13.Rb5 Rxd4 14.Rd5 Rc4+ 15.Kb6 Rb4+ 16.Kc7 Rc4+ 17.Kb7 Rb4+ 18.Ka7 Ra4+ 19.Kb8 Rb4+ 20.Kc7 Rc4+
  +/=  (0.53)   Depth: 33/65   00:02:35  4828mN, tb=299090
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Rc7 Ra8 9.d6 Kf6 10.Rb7 Ra6+ 11.Rb6 Ra3 12.d4 Ra4 13.Rb5 Rxd4 14.Rd5 Rc4+ 15.Kb6 Rb4+ 16.Kc7 Rc4+ 17.Kb7 Rb4+ 18.Ka7 Ra4+ 19.Kb8 Rb4+ 20.Kc7 Rc4+
  +/=  (0.53)   Depth: 34/65   00:02:41  5020mN, tb=309933
1.Re7
  +/=  (0.60 ++)   Depth: 35/67   00:03:27  6535mN, tb=389925
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Rc6 Ra1 18.Rh6 Rc1 19.Re6 Ra1 20.Kc4 Ra2 21.Rf6 Ke7 22.Re6+ Kd7 23.Rf6
  +/=  (0.62)   Depth: 35/69   00:04:23  8402mN, tb=468141
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Rc6 Rb1+ 18.Kc4 Ra1 19.Rh6 Ra5 20.Re6 Ra2 21.Kd4 Ra5 22.Rh6 Ra3 23.Re6 Ra5 24.Rh6
  +/=  (0.62)   Depth: 36/71   00:04:56  9484mN, tb=520263
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Ra6 Rc8 18.Rh6 Ra8 19.Re6 Rb8+ 20.Kc5 Rb3 21.Kc4 Ra3 22.Rf6 Ke7 23.Rg6 Ra1 24.Re6+ Kd7 25.Rh6 Ra5 26.Re6 Ra2 27.Kd4 Ra5 28.Rh6 Ra3 29.Re6 Ra5 30.Rh6
  +/=  (0.62)   Depth: 37/71   00:06:00  11631mN, tb=614870
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Ra6 Rc8 18.Rh6 Ra8 19.Re6 Rb8+ 20.Kc5 Rb3 21.Kc4 Ra3 22.Rf6 Ke7 23.Rg6 Ra1 24.Re6+ Kd7 25.Rh6 Ra5 26.Re6 Ra2 27.Kd4 Ra5 28.Rh6 Ra3 29.Re6 Ra5 30.Rh6
  +/=  (0.62)   Depth: 38/72   00:07:52  15349mN, tb=802237
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Ra6 Rc8 18.Rh6 Ra8 19.Re6 Rb8+ 20.Kc5 Rb3 21.Kd4 Ra3 22.Rh6 Kc7 23.Rg6 Kd7 24.Re6 Ra5 25.Rh6 Ra3 26.Re6 Ra5
  +/=  (0.62)   Depth: 39/74   00:10:43  21166mN, tb=1064435
1.Re7 g5 2.Kb4 Rd5 3.Kc4 Rd8 4.d5 Kg6 5.Kc5 Rc8+ 6.Kb5 Rd8 7.Kc6 Rc8+ 8.Kb7 Rc5 9.Re5 g4 10.Kb6 Rc3 11.Re3 Kf7 12.Kb5 g3 13.Rxg3 Ke7 14.Kb4 Rc2 15.Rg6 Rc1 16.Re6+ Kd7 17.Ra6 Rc8 18.Rh6 Ra8 19.Re6 Rb8+ 20.Kc5 Rb3 21.Kd4 Ra3 22.Rh6 Kc7 23.Rg6 Kd7 24.Re6 Ra5 25.Rh6 Ra3 26.Re6 Ra5
  +/=  (0.62)   Depth: 40/80   00:14:57  29962mN, tb=1449575
1.Re7
  +/=  (0.68 ++)   Depth: 41/84   00:32:32  68059mN, tb=3039079

(Bicknell,  02.05.2012)
Parent - - By Nelson Hernandez (Gold) [us] Date 2012-05-13 12:16
Now, wasn't that a super-effective, yet devilishly subtle, way of compelling us to buy Houdini 3?  Robert is a man who knows his audience.
Parent - - By Carl Bicknell (*****) [es] Date 2012-05-13 15:53
The funny thing is I'm a complete sucker for this sort of thing even when I know people are doing a 'sell' on me. When Vas talks about Rybka I want to buy it, when Robert talks about Houdini it's the same...
Parent - By Nelson Hernandez (Gold) [us] Date 2012-05-13 22:17
Yep, me too, I can't resist "positive side effects of other changes" either.
Parent - - By Arpad Rusz (*) [hu] Date 2012-04-28 15:15 Edited 2012-04-28 15:42
8/4k3/3R4/2r2Rr1/8/4N3/8/4K3 w - - 0 1
(Can you confirm that this is a win with only winning move 1.Rdd5!)
8/4k3/8/2rR1Rr1/8/4N3/8/4K3 b - - 0 1
(The previous position after 1.Rdd5. This should be win for white.)
8/4k3/3R1R2/2r3r1/8/4N3/8/4K3 b - - 0 1
(And after 1.Rff6. This should be draw.)
Parent - - By 7-man TB (**) [ru] Date 2012-04-28 16:03
1) Yes, it is the only mate. { Evaluation: 133 half-moves to mate. }

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[FEN "8/4k3/3R4/2r2Rr1/8/4N3/8/4K3 w - - 0 1"]
[SetUp "1"]

{Evaluation: 133 half-moves to mate.}  1. Rdd5 $1  1... Rc1+ 2. Kd2 $1  2...
Rgg1 3. Rde5+ Kd6 4. Re4 Ra1 5. Rf7 Rg6 6. Ree7 Rg5 7. Rh7 Ra4 8. Rd7+ Kc6 9.
Rc7+ Kb5 10. Rh6 Ra2+ 11. Kd3 Ra3+ 12. Ke4 Ra4+ 13. Kf3 Re5 14. Rc3 Rc5 15.
Rb3+ Rb4 16. Rd3 Rc6 17. Rh5+ Rc5 18. Nd5 Rb2 19. Rh8 Rcc2 20. Rd1 Rf2+ 21.
Ke4 Rfe2+ 22. Kd3 Rh2 23. Rd8 Ka6 24. Rf1 Kb5 25. Rc1 Rbd2+ 26. Kc3 $1  26...
Ra2 27. Rb1+ Kc6 28. Kb3 Kb7 29. Rd6 Rac2 30. Rd7+ Kb8 31. Rg7 Rhg2 32. Rh7
Rh2 33. Rd7 Rhe2 34. Rd8+ Kb7 35. Ka3+ Kc6 36. Nb4+ Kc7 37. Rbd1 Rc3+ 38. Ka4
Kb7 39. Rh1 Rc8 40. Rd6 Rc7 41. Kb5 Rg2 42. Rh8 Rcg7 43. Rb6+ Kc7 44. Nd5+ Kd7
45. Ra6 Rc2 46. Ra7+ Ke6 47. Re8+ Kxd5 48. Rxg7 {[%t bLon] Mate in 20:} 48...
Rb2+ 49. Ka4 Rb1 50. Rd7+ Kc6 51. Rd4 Ra1+ 52. Kb3 Rb1+ 53. Ka2 Rb5 54. Rc8+
Kb7 55. Rc3 Ra5+ 56. Kb2 Rb5+ 57. Ka3 Ra5+ 58. Ra4 Rb5 59. Rb4 Ka6 60. Rcb3
Rxb4 61. Rxb4 Ka7 62. Rb5 Ka8 63. Kb4 Ka7 64. Ka5 Ka8 65. Kb6 Kb8 66. Rc5 Ka8
67. Rc8# *

2) see 1

3) Draw

4) This is not in the list of endings above, but occasionally the ending is available already.
Draw.
Parent - - By Arpad Rusz (*) [hu] Date 2012-04-28 16:50 Edited 2012-04-28 16:52
Thanks!

Here are some RPP/RP positions from endgame studies:

8/5k2/5p2/8/P1p2KR1/3r4/8/8 w - - 0 1
(Becker 2009 1.Rg2! =)
8/5K2/8/6k1/1P1p3r/8/3PR3/8 w - - 0 1
(Gonzales 2009 1.d3! +-)
8/3p4/5p2/3P4/8/6RK/5k2/4r3 w - - 0 1
(Becker-Akobia 2007 1.Rg6! =)
8/1p2r3/P6R/8/1k6/3Kp3/8/8 w - - 0 1
(Matous 2000 1.Rb6! +-)
Parent - By 7-man TB (**) [ru] Date 2012-04-28 17:12
Tables confirmed studies.
Studies confirmed tables.

1) Draw

2) {Evaluation: 85 half-moves to mate.} 
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Arpad Rusz"]
[Black "85"]
[Result "*"]
[FEN "8/5K2/8/6k1/1P1p3r/8/3PR3/8 w - - 0 1"]
[SetUp "1"]

1. d3 $1  1... Rh3 2. Re5+ $1  2... Kh6
3. Re4 Rxd3 4. Kf6 {[%t bLon] Mate in 40:} 4... Rf3+ 5. Ke5 $1  5... d3 6. Rd4
$1  6... Kg5 7. Ke4 Rf8 8. Rd5+ Kg6 9. Rxd3 Rh8 10. Rd4 Rh5 11. Kd3 Rh8 12. b5
Kf5 13. Kc4 Rc8+ 14. Kd5 Rd8+ 15. Kc5 Rc8+ 16. Kb6 Ke5 17. Rd3 Rf8 18. Kc7
Rf7+ 19. Rd7 Rf8 20. b6 Rf2 21. b7 Rc2+ 22. Kd8 Rb2 23. Rc7 Rb6 24. Kc8 Re6
25. b8=Q Re8+ 26. Kb7 $1  26... Rxb8+ 27. Kxb8 $1  27... Kd6 28. Rc8 Ke5 29.
Re8+ Kd6 30. Kb7 Kd5 31. Kc7 Kc4 32. Rd8 Kb5 33. Rd5+ Kb4 34. Kb6 Kc4 35. Rd6
Kb3 36. Rc6 Ka4 37. Kc5 Kb3 38. Kb5 Ka2 39. Ka4 Kb2 40. Kb4 Ka1 41. Rc2 Kb1
42. Kb3 Ka1 43. Rc1# *

3) Draw

4) {Evaluation: 75 half-moves to mate.} 
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[FEN "8/8/3kP3/1K6/8/p6r/1P2R3/8 b - - 0 1"]
[SetUp "1"]

1... Rb3+ $1  2. Ka4 axb2 3. e7 {[%t bLon] Mate in 36:} 3... Rb4+ $1  4. Ka5 Rb5+ $1  5. Ka6 Rb6+ $1  6. Ka7 Rb7+
$1  7. Ka8 Rb8+ $1  8. Kxb8 b1=Q+ $1  9. Ka7 Kc7 10. e8=N+ Kc6 11. Re6+ Kd7 $1
12. Rb6 Qa2+ 13. Ra6 Qf2+ 14. Rb6 Kxe8 15. Ka6 Kd7 16. Kb5 Qe3 17. Ka5 Kc7
18. Rb4 Kc6 19. Ka4 Qe2 20. Rb3 Kd5 21. Rc3 Kd4 22. Kb3 Qd1+ 23. Kb2 Qd2+ 24.
Rc2 Qb4+ 25. Ka2 Kd3 26. Rc1 Qb5 27. Ka3 Kd2 28. Rh1 Qa6+ 29. Kb4 Qb7+ 30. Kc5
Qxh1 31. Kd6 Qf3 32. Ke7 Qc6 33. Kf8 Qd7 34. Kg8 Ke3 35. Kh8 Kf4 36. Kg8 Kg5
37. Kh8 Kh6 38. Kg8 Qe8# *
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-04-28 16:50
The following positions are of interest in connection with position #12 above.
8/5k2/5p2/8/3PK3/4P3/3r4/7R w - - 0 49
8/5k2/5p2/3K4/3P4/4P3/3r4/7R b - - 0 49
8/5k2/5p2/3K4/3P4/4P3/r7/7R w - - 0 50
8/5k2/5p2/3K4/3P4/4P3/r7/2R5 b - - 0 50
8/5k2/5p2/r2K4/3P4/4P3/8/2R5 w - - 0 51
Parent - - By 7-man TB (**) [ru] Date 2012-04-28 17:00 Edited 2012-04-28 17:08
1) Draw
2) Draw
3) {Evaluation: 71 half-moves to mate.} 
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Dadi"]
[Black "71"]
[Result "*"]
[FEN "8/5k2/5p2/3K4/3P4/4P3/r7/7R w - - 0 1"]
[SetUp "1"]

1. Rh7+ $1  1... Kg6 2. Re7 f5 3. Ke5
Ra5+ 4. d5 Rb5 5. Kd6 Ra5 6. e4 fxe4 7. Rxe4 {[%t bLon] Mate in 30:} 7... Kf7
8. Re2 Ra6+ 9. Kc7 Ra7+ 10. Kb6 Rd7 11. Kc6 Ra7 12. d6 Ra6+ 13. Kc7 Ra7+ 14.
Kb6 Ra4 15. d7 Rd4 16. Kc7 Rc4+ 17. Kd8 Rc3 18. Rf2+ Kg6 19. Rf4 Rc2 20. Ke7
Re2+ 21. Kd6 Rb2 22. d8=Q Rd2+ 23. Ke7 Rxd8 24. Kxd8 $1  24... Kg5 25. Rf7 Kg6
26. Ke7 Kg5 27. Ke6 Kg4 28. Rf5 Kg3 29. Kf6 Kh4 30. Rg5 Kh3 31. Rg6 Kh4 32.
Kf5 Kh3 33. Kf4 Kh2 34. Kf3 Kh1 35. Kf2 Kh2 36. Rh6# *

ADDED:
4) Draw
5) {Evaluation: 75 half-moves to mate.}
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[FEN "8/5k2/5p2/r2K4/3P4/4P3/8/2R5 w - - 0 1"]
[SetUp "1"]

1. Kd6 Ra6+ 2. Rc6 Ra3 3. Rc7+ Kg6 4.
Re7 f5 5. d5 Ra6+ 6. Kc5 Ra5+ 7. Kc6 Ra6+ 8. Kb5 Ra8 9. Kb6 Rd8 10. Kc6 Rc8+
11. Kd7 Rc1 12. e4 f4 13. Re8 f3 14. Rf8 Re1 15. Rxf3 Rxe4 {[%t bLon] Mate in
23:} 16. d6 Rd4 17. Ke7 Re4+ 18. Kd8 Re2 19. d7 Rd2 20. Rf4 Kg5 21. Re4 Rh2
22. Ke7 Rh8 23. d8=Q Rxd8 24. Kxd8 $1  24... Kf5 25. Re7 Kf6 26. Re2 Kg5 27.
Rf2 Kg6 28. Kc7 Kh5 29. Rg2 Kh4 30. Rg8 Kh5 31. Kd8 Kh6 32. Ke7 Kh5 33. Kf6
Kh4 34. Kf5 Kh3 35. Kf4 Kh2 36. Kf3 Kh1 37. Kf2 Kh2 38. Rh8# *
Parent - By Dadi Jonsson (Silver) [is] Date 2012-04-28 17:03
Oops, you were too fast :lol: I edited the post above and added two positions.
Parent - By 7-man TB (**) [ru] Date 2012-04-28 21:08
Just to mention, 3 more endings are available. These endings are known as the endings with very long mates. So they were generated first.
RBN-QN
QBN-QB
QBN-QN
Parent - - By EmilV [cz] Date 2012-04-29 04:38
Hello, 7-men!

I would like to report about your work in EG.
EG is a special magazine for endgame studies and I am editor of its computer column.

Thanks, Emil   emil@vlasak.biz
Parent - By 7-man TB (**) [ru] Date 2012-04-29 06:57
Thanks, Emil!

I sent you some information to pm.
Parent - - By Highendman (****) Date 2012-04-29 12:04
That's incredible! Can you please share some info on the process of generating these? How long it took? The size it takes? Your plans for the future? Are you planning on releasing the code / data?
Parent - - By Fulcrum2000 (****) [nl] Date 2012-04-29 12:14
In another thread there is a bit of info given by Victor:
http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=413381#pid413381
Parent - By Highendman (****) Date 2012-04-29 14:01
Thank you!
Parent - - By mbourzut (*) [us] Date 2012-04-29 14:36
Below a full-point mutual zugzwang, where whoever is to move loses.  How many such positions does your database identify?

8/8/2P5/2r1p3/RK1kP3/8/8/8 w - - 0 1

8/8/2P5/2r1p3/RK1kP3/8/8/8 b - - 0 1
Parent - By 7-man TB (**) [ru] Date 2012-04-29 18:31
Wow, this is very interesting! Thanks for the idea.

There is no code for detection zugzwang positions, but I see that it is easy to implement this. Will be done soon. But not next week. There are many other troubles as for generation so for usual work.

I need to say that generation and related tasks is the background process for the team.
Parent - - By cma6 (****) Date 2012-04-29 19:26
I have a 7-man position that fits your tables. Is there a way to generate the EPD file with Aquarium or does one do that by hand? Is EPD the same as the FEN representation that Dadi used in this thread?
           Thanks, CMA
Parent - - By 7-man TB (**) [ru] Date 2012-04-29 19:40
1) Press Ctrl\  in Aquarium to have a position in clipboard.
2) Press Ctrl-V to place the text into your message

EPD/FEN are the same for all needs
Up Topic The Rybka Lounge / Computer Chess / Testing of some 7-men tables
1 2 3 4 5 6 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill