Critter 0.52b - time managment bugfix

[rvida]

Hi,

I recieved several emails asking specifically for this bugfix.

So here we are: http://www.mediafire.com/?nudydmxn2yd

With some GUIs there is no need for this fix, while in others it cures the following symptom:

Also there seems to be a problem with time management when playing 40
moves in x minutes it uses al lot off time and then play the last moves very quick, sometimes it ruines a good position then.

Anyway, I'm happy to see people are interested in my engine, it gives me much motivation.

---
Richard

[mariaclara]

thanks for the fix.

will do new tournaments.

Hi,

I recieved several emails asking specifically for this bugfix.

So here we are: http://www.mediafire.com/?nudydmxn2yd

With some GUIs there is no need for this fix, while in others it cures the following symptom:

Also there seems to be a problem with time management when playing 40
moves in x minutes it uses al lot off time and then play the last moves very quick, sometimes it ruines a good position then.

Anyway, I'm happy to see people are interested in my engine, it gives me much motivation.

---
Richard

[alpha123]

I'll start testing this version.

I'm glad you're motivated, stay motivated. Critter is one of my favorite engines, though I'm not quite sure why.....

Thanks,
Peter

[Werner]

Hi Richard,
thanks for the update. For your info: I made a comparison match with the same openings (top25.pgn from Sedat). First match with pgn and second match with epd startpositions:

Code: Select all

Critter2 pgn
1   Hiarcs 12 SP 1CPU  2200  +17/-11/=22 56.00%   28.0/50
2   Critter 0.52 x64   2200  +11/-17/=22 44.00%   22.0/50

Code: Select all

Critter2 epd
1   Hiarcs 12 SP 1CPU  2200  +17/-9/=24 58.00%   29.0/50
2   Critter 0.52 x64   2200  +9/-17/=24 42.00%   21.0/50

I think - it´s no real difference - but looking at the games short before move 40 you can see no more moves in 1 sec
I think we will see with too this new version.

[Werner]

Here is a short game with Critter 0.52 x64.
Have a look at move 8....Bxf3?

[d]r3k2r/2p2ppp/p1n5/1pb1P3/4B1b1/2p2N2/PP3PPP/R1BR2K1 b kq - 0 8

Werner

[Event "TwistedLogic"]
[Site "I7"]
[Date "2010.02.03"]
[Round "21"]
[White "Twisted Logic 20100131 x64"]
[Black "Critter 0.52 x64"]
[Result "1-0"]
[BlackElo "2200"]
[SetUp "1"]
[FEN "r2qkb1r/2p2ppp/p1n1b3/1p1pP3/4n3/1B3N2/PPP2PPP/RNBQ1RK1 w kq - 0 1"]
[PlyCount "38"]
[EventDate "2010.??.??"]
[TimeControl "40/480:40/480:40/480"]

1. Nbd2 {+0.28/17 53} Nc5 {-0.60/17 24} 2. c3 {+0.25/16 13} Bg4 {-0.45/16 11}
3. Re1 {+0.38/17 26} Be7 {-0.53/17 11} 4. Bc2 {+0.45/17 30} d4 {-0.73/16 11} 5.
Nb3 {+0.87/16 11} dxc3 {-0.05/17 11} 6. Nxc5 {+0.66/16 10} Bxc5 {0.00/17 11} 7.
Be4 {+0.51/15 10} Qxd1 {+0.22/18 11} 8. Rxd1 {+0.62/16 10} Bxf3 {+0.33/19 11}
9. gxf3 {+2.60/16 10} Bxf2+ {-3.02/18 11} 10. Kxf2 {+6.50/15 9} O-O {-7.67/18
11} 11. Bxc6 {+6.72/16 9} Rad8 {-8.19/20 11} 12. Rxd8 {+6.74/16 9} Rxd8 {-8.10/
19 2} 13. bxc3 {+6.91/16 9} Rd1 {-8.74/20 22} 14. Rb1 {+7.03/16 9} g6 {-8.42/19
15} 15. Be4 {+7.47/15 9} Rh1 {-9.05/18 20} 16. Bg5 {+8.32/15 9} Rxb1 {-9.05/18
11} 17. Bxb1 {+8.68/14 9} Kf8 {-9.33/17 11} 18. Be4 {+9.21/15 9} c5 {-9.80/16
16} 19. Be3 {+9.38/15 9} Ke7 {-10.19/16 11 Schwarz gibt auf} 1-0

[rvida]

Here is a short game with Critter 0.52 x64.
Have a look at move 8....Bxf3?

[d]r3k2r/2p2ppp/p1n5/1pb1P3/4B1b1/2p2N2/PP3PPP/R1BR2K1 b kq - 0 8

Hm, very strange, I dont remember changing anything related to gameplay, but here seems 0.52b analysis differs from 0.52a

Critter 0.52b 32-bit:

Code: Select all

  2/5	00:00	       2.681	-3,07	Bg4xf3 g2xf3 Bc5xf2+ Kg1xf2
  2/5	00:00	       2.784	-0,86	Bg4d7 b2xc3
  3/4	00:00	       3.193	-0,67	Bg4d7 b2xc3 OOO
  4/5	00:00	       3.997	-0,72	Bg4d7 b2xc3 OOO Ra1b1
  5/6	00:00	       6.625	-0,56	Bg4d7 b2xc3 OOO a2a4 Bd7e6
  6/7	00:00	       9.556	-0,56	Bg4d7 b2xc3 OOO a2a4 Bd7e6 Rd1xd8+ Nc6xd8
  7/12	00:00	      29.214	-0,73	Bg4d7 Nf3g5 Rh8f8 b2xc3 h7h5 Ra1b1 OOO
  8/11	00:00	      46.557	-0,66	Bg4d7 Nf3g5 Bc5e7 b2xc3 Ra8d8 Ng5f3 h7h6 Ra1b1
  9/12-	00:00	      98.440	-1,05	Bg4d7
  9/13	00:00	     114.712	-0,85	Bg4d7 Nf3g5 Bc5e7 b2xc3 Be7xg5 Bc1xg5 Ra8b8 f2f4 h7h6 Bg5h4
 10/15	00:00	     240.673	-0,77	Bg4d7 Rd1d3 OOO Rd3xc3 Bc5b4 Rc3c2 Bd7g4 Bc1f4 Bg4xf3 Be4xf3 Nc6d4
 11/15	00:00	     399.571	-0,80	Bg4d7 Rd1d3 OOO Rd3xc3 Bc5b4 Rc3c2 Bd7g4 Bc1g5 Bg4xf3 Be4xc6 Rd8d1+ Ra1xd1 Bf3xd1 Rc2c1
 12/20	00:01	   1.418.216	-0,48	Bg4d7 b2xc3 OOO Bc1f4 Nc6e7 Nf3g5 Bd7e6 Rd1xd8+ Rh8xd8 Ng5xe6 f7xe6 Be4xh7 Rd8f8 Bf4g3
 13/17	00:02	   2.050.372	-0,49	Bg4d7 b2xc3 OOO Bc1f4 Nc6e7 Nf3g5 Bd7e6 Rd1xd8+ Rh8xd8 Ng5xe6 f7xe6 Be4xh7 Rd8f8 Bf4g3 g7g5
 14/20	00:04	   3.708.903	-0,50	Bg4d7 b2xc3 OOO Bc1f4 Nc6e7 Nf3g5 Bd7e6 Rd1xd8+ Rh8xd8 Ng5xe6 f7xe6 Be4xh7 Ne7d5 Bf4d2 Kc8d7 Bh7g6

Critter 0.52 32-bit:

Code: Select all

  2/5	00:00	       2.681	-3,07	Bg4xf3 g2xf3 Bc5xf2+ Kg1xf2
  2/5	00:00	       2.784	-0,86	Bg4d7 b2xc3
  3/4	00:00	       3.193	-0,67	Bg4d7 b2xc3 OOO
  3/5	00:00	       3.907	-0,18	Bg4xf3 Be4xf3 OOO Rd1xd8+ Nc6xd8
  4/8	00:00	       5.377	+0,24	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3
  5/9	00:00	       7.241	+0,46	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Kc8d7
  6/10	00:00	      10.742	+0,32	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 f3f4
  7/12	00:00	      16.895	+0,28	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 f3f4 f7f6
  8/13	00:00	      27.871	+0,51	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 f3f4 f7f6 a2a4
  9/14	00:00	      48.001	+0,48	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 Be4f5+ Kc8b7 f3f4 g7g6 Bf5e4+ Kb7b6
 10/15	00:00	      91.379	+0,25	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 Be4f5+ Kc8b8 f3f4 g7g6 Bf5d7 Re8e7
 11/17	00:00	     214.993	+0,34	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 Be4f5+ Kc8b7 f3f4 g7g6 Bf5e4+ Kb7b6 Kg1g2 f7f6
 12/19	00:00	     451.743	+0,08	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 f3f4 g7g6 a2a4 f7f6 Be4d3 b5xa4 Ra1xa4 f6xe5
 13/19	00:01	     915.298	+0,07	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 Rh8e8 f3f4 g7g6 a2a4 f7f6 a4xb5 a6xb5 Be4d3 f6xe5
 14/21	00:02	   1.841.999	+0,13	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 f7f6 e5xf6 g7xf6 a2a4 Kc8d7 Be4d3 Kd7c6 Bd3e4+ Kc6b6 a4xb5 a6xb5
 15/21	00:03	   3.349.350	+0,13	Bg4xf3 g2xf3 OOO Rd1xd8+ Nc6xd8 b2xc3 f7f6 e5xf6 g7xf6 a2a4 Kc8d7 a4xb5 a6xb5 Kg1f1 Bc5d6 f3f4 Nd8e6

This looks like in 0.52a something goes wrong with hashing. But the sources are absolutely intentical except UCI handling & time managment code. I guess it was some comiler/optimization related glitch.

Edit: note that after Bxf3 gxf3 the castling move shown in PV is illegal (!)

Richard

[Frank Quisinsky]

Hi Richard,

so I will try Critter for the SWCR rating list from the German newspaper SCHACHWELT too.

SCHACHWELT
http://www.schach-welt.de/index.php?opt ... &Itemid=79

So I need in the next times an interview details about your engine (to the playing style) for the other pages of SWCR.

But all in 10 days, because 3 other update which comes before are to test.

THANKS for Critter !!

Best
Frank

[swami]

Critter 0.52b JA
by Richard Vida, Slovak Republic.

Strategic Test Suite Conditions:

Core2Quad 32 bits, Q6600, 2 GB RAM, 2.4GHZ
10 seconds per position
900 positions
Engine uses 128 Mb Hash.
Single CPU
Gradual Test


Overall Performance:

• Total Score: 644/900 [.....] Average : 71.55% [.....] Grade: A- [.....] Total Rated Time: 42.66/150 minutes [2560 Seconds/9000 Seconds]= 0.284

Subject-wise Scores:

STS (v1.0) - Undermining:
78/100, Grade: A

STS (v2.1) - Open Files and Diagonals:
71/100, Grade: A-

STS (v3.0) - Knight Outposts/Centralization/Repositioning:
73/100, Grade: A-

STS (v4.1) - Square Vacancy:
75/100, Grade: A

STS (v5.0) - Bishop vs Knight:
77/100, Grade: A

STS (v6.0) - Re-Capturing:
78/100, Grade: A

STS (v7.0) - Offer of Simplification:
61/100, Grade: B

STS (v8.1) - Advancement of f/g/h Pawns:
64 /100, Grade: B

STS (v9.0) - Advancement of a/b/c Pawns:
67/100, Grade: B+

Best Wishes,
Swami

[mhalstern]

I am running a gauntlet with the latest 64 bit version on my I7-920. Look for the games and crosstable in the "Computer Chess Club: Tournaments and Matches." The tourney should be finished by late Sunday Evening. I'll post as soon as it is finished.