Division 6,4th Edition:16 Engines

[Tony Thomas]

Knightx, Mataczx , Alichess. I say it is too hard to pick, because I want to pick them all.

[swami]

Joker will probably not do very well on P-4. Unfortunately I had no opportunity to make a P-4-proof version yet. OTOH I might be lucky: the most dramatic drop in speed only occurs at very specific search depths (when system stack and move stack collide in the L1 cache), and in 1 min games this depth might never be reached.

Hi,

This is 1+1,I can't call that a bullet since it is more of blitz than a bullet because there is 1 second increment per move.Joker does somewhat okay and I guess the latest version might boost the strength.

[swami]

Knightx, Mataczx , Alichess. I say it is too hard to pick, because I want to pick them all.

Cool,Tony. That's too much of a compliment regarding the level of toughness in my tournament

[Graham Banks]

Matacz
Anechka
Natwarlal

Hi Graham. You pick my guess .
At least some difference
Matacz, KnightX, Natwarlall.

Hi Pavel,

you're allowed to select the same three if you want. Just let Swami know.

Regards, Graham.

[swami]

Matacz
Anechka
Natwarlal

Hi Graham. You pick my guess .
At least some difference
Matacz, KnightX, Natwarlall.

Hi Pavel,

you're allowed to select the same three if you want. Just let Swami know.

Regards, Graham.

You will get an additional 1 point if you replied first and someone picked 3 engines same as yours, 1 point might seem too little but it's worth it in the horse race such as this

[swami]

Code: Select all

[Event "Division 6"]
[Site "ADMIN"]
[Date "2007.07.18"]
[Round "1"]
[White "Joker1108w"]
[Black "Esc 1.16"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "E12"]
[Opening "Queen's Indian"]
[Time "08:54:35"]
[Variation "Petrosian, 5.Nc3 Be7"]
[WhiteElo "2200"]
[TimeControl "60+1"]
[Termination "normal"]
[PlyCount "84"]
[WhiteType "program"]
[BlackType "program"]

1. d4 Nf6 2. Nf3 b6 3. c4 Bb7 4. Nc3 e6 5. a3 Be7 6. d5 O-O 7. e4 d6 8. Bd3
Nbd7 9. O-O Ne5 10. Nxe5 dxe5 11. f4 Nd7 12. Qg4 Bc5 13. Kh1 Qe7 14. Qg3
Bd4 15. Nb5 exd5 16. exd5 c6 17. d6 Qd8 18. Nxd4 exd4 19. f5 f6 20. Bh6
{(c1h6 f8f7 h6f4 d8f8 h1g1 a8d8 b2b4 d7e5) +0.34/11 3} Rf7 {(Rf7 Bf4 Nc5
Bc2 Nd7 Rfd1 c5 Bd3 a6 b3) -0.62/9 2} 21. Bf4 {(h6f4 d7e5 f4e5 f6e5 g3e5
f7f6 c4c5 b6c5 d3c4 g8h8 c4e6) -0.44/10 3} Nc5 {(Nc5 Bb1 Qf8 b4 Nd7 Rd1 c5
bxc5 bxc5 Bd3) -0.70/9 2} 22. Bc2 {(d3c2 c5d7 h1g1 c6c5 a1e1 d8c8 b2b3 b7c6
b3b4 c8a6 g3d3) +0.59/10 3} Nd7 {(Nd7 Rfe1 c5 Bd3 Bc6 b3 a5 a4 Rf8 Rad1)
-0.78/9 2} 23. Kg1 {(h1g1 c6c5 a1e1 d8f8 b2b4 b7c6 f1f2 a8d8 e1e6 d8a8)
+0.63/10 2} c5 {(c5 Rfe1 Qf8 Bd3 Rd8 Re6 a6 Rae1 Bc6) -0.64/9 2} 24. Rae1
{(a1e1 d7e5 f4e5 f6e5 e1e5 d8d7 f1e1 a8d8 e5e6 d8f8) +0.66/10 3} Kh8 {(Kh8
Bd3 a6 Rf2 Bc6 b3) -0.68/8 2} 25. b3 {(b2b3 a7a5 f1f2 d7e5 f4e5 f6e5 g3e5
f7d7 f5f6 d7d6 f6g7) +0.70/9 5} Qf8 {(Qf8 Bd1 Ne5 Bxe5 fxe5 Qxe5 Rd8 Bh5
Rfd7) -0.65/8 2} 26. Re6 {(e1e6 h8g8 f1e1 b7c6 b3b4 f8c8 g3f2 a8b8 c2d3)
+0.84/10 1} Rd8 {(Rd8 Rfe1 Nb8 Re7 Nc6 Rc7 Rxc7 dxc7 Rd7 Bd3) -0.79/9 2}
27. Rfe1 {(f1e1 d7e5 f4e5 f6e5 e6e5 b7c8 e5e7 c8f5) +0.57/9 2} Nb8 {(Nb8
Bd1 Qg8 Bh5 Rff8 Re7 Nd7 Qd3 a6) -1.02/9 2} 28. R1e2 {(e1e2 f7d7 e2e1 h8g8
c2d3 b8c6 d3e4 f8f7) +0.85/8 1} Nc6 {(Nc6 Be4 Rfd7 Rb2 Na5 Bd3 Nc6 Qf3 Ne5)
-0.41/9 2} 29. Be4 {(c2e4 c6a5 e4b1 h8g8 b1c2 a5c6 c2e4 c6b8) +0.79/9 1}
Rfd7 {(Rfd7 Bf3 a5 Be4 a4 bxa4 Na5 Bd3) -0.47/8 2} 30. Bf3 {(e4f3 d8a8 g3e1
c6e5 f3b7 e5d3 e1f1 d3f4 f1f4 d7b7) +0.93/9 2} a6 {(a6 Re1 Kg8 Bd5 Kh8 Bf3
Kg8 Qf2 Ne5 Bxb7 Rxb7) -0.46/8 2} 31. b4 {(b3b4 a6a5 b4c5 b6c5 e2b2 d8c8
f4d2 c8a8) +1.01/8 1} a5 {(a5 bxc5 bxc5 Rb2 Rc8 Qe1 Qf7 Be4 Ba6) -0.77/8 2}
32. b5 {(b4b5 c6e5 f3b7 d7b7 f4e5 f6e5 g3e5 h8g8) +1.12/9 1} Ne5 {(Ne5 Bxb7
Rxb7 Bxe5 fxe5 Qxe5 Rf7 Rf2 d3 f6 gxf6 Rfxf6 Rxf6) -1.14/10 2} 33. Bxb7
{(f3b7 d7b7 f4e5 f6e5 g3e5 b7f7 e2f2 f7f6 e6f6 g7f6 e5e7) +0.98/11 1} Rxb7
{(Rxb7 Bxe5 fxe5 R2xe5 a4 Qe1 d3 Re8 Rxe8 Rxe8) -1.46/10 2} 34. Bxe5 {(f4e5
f6e5 g3e5 b7f7 e2f2 f8g8 f5f6 d8f8 f6g7 g8g7 f2f7) +1.32/11 0} fxe5 {(fxe5
R2xe5 a4 Qe1 d3 Re8 Rxe8 Rxe8) -1.46/9 2} 35. Qxe5 {(g3e5 b7f7 e2f2 a5a4
f5f6 g7f6 f2f6 f7f6 e6f6 f8g7 e5f4) +1.40/11 1} d3 {(d3 Rd2 Rf7 g4 a4 Rxd3
Rdd7 Re8) -3.00/10 4} 36. Rd2 {(e2d2 b7f7 g2g4 h8g8 d2d3 h7h5 h2h3 a5a4
d3e3 f7d7 e5d5) +2.58/11 1} Rf7 {(Rf7 g4 a4 Rxd3 Ra7 Re7 Rxd6 Qxg7+)
-3.15/10 2} 37. g4 {(g2g4 f8g8 d2d3 d8d7 d3e3 a5a4 g4g5 f7f8 e5f4 f8f7
f4g4) +3.11/11 1} Qg8 {(Qg8 Rxd3 Rdd7 Re8 Rf8 Rxf8 Qxf8 Qe6 Qd8 h4 a4)
-3.21/10 2} 38. Rxd3 {(d2d3 a5a4 d3e3 f7f8 e5d5 g8f7 d5c6 d8b8 f5f6 g7f6
c6g2) +3.27/11 1} Rfd7 {(Rfd7 Re7 Rxe7 dxe7 Re8 Qe6 a4 Rd6 Qxe6 Rxe6 Kg8
h4) -3.92/10 4} 39. g5 {(g4g5 d8a8 g5g6 g8d8 e6e7 d7e7 d6e7 d8e8 e5d6 a5a4
d6b6) +4.27/11 2} a4 {(a4 g6 h6 Rh3 Qf8 Rxh6+ Kg8 Rh7 Ra8) -3.95/8 2} 40.
Rh3 {(d3h3 d7d6 e6d6 d8d6 e5d6 g8e8 g1f2 e8g8 d6b6 g8c4 b6b8) +8.41/12 2}
Qf8 {(Qf8 Reh6 Qxf5 Qxf5 gxh6 Rxh6 Kg8 Qd5+ Kh8 Qc6 Rg7) -7.61/10 4} 41.
Reh6 {(e6h6 f8g8 g5g6 d8f8 h6h7 g8h7 h3h7 h8g8 e5e6 d7f7 g6f7) +17.92/10 1}
Qg8 {(Qg8 g6 Rf8 Rxh7+ Qxh7 Rxh7+ Kg8 Qe6+ Rff7 Rh8+ Kxh8 Qe8+ Rf8 Qxf8#)
-M1/10 6} 42. g6 {(g5g6 d8f8 h6h7 g8h7 h3h7 h8g8 e5e6 f8f7 e6e8 f7f8 h7h8)
+M497/13 2} Rf8 {(Rf8 Rxh7+ Qxh7 Rxh7+ Kg8 Qe6+ Rff7 Rh8+ Kxh8 Qe8+ Rf8
Qxf8#) -M1/10 2} 43. ... {Black resigns} 1-0

Good game by Joker in blitz

[Tony Thomas]

Knightx, Mataczx , Alichess. I say it is too hard to pick, because I want to pick them all.

Cool,Tony. That's too much of a compliment regarding the level of toughness in my tournament

It is tough considering that the strength differences are rather small, and the only fact deciding the outcome is the programs ability to play blitz.

[hgm]

I noted that some engines one some platforms have a big problem playing blitz, because of the time it takes to allocate memory for their hash table. Sometimes the OS decides to create a swap file for this, which can take dozens of seconds for a 100MB hash table.

Problem is that the OS does this only on first use: in the task manager you can see that the engine starts with a very small memory foortprint, and then blows up as it starts using the hash table. But if the engine only starts using the hash table after it gets out of book, it is basically dead while the OS is allocating the space. But the GUI doen't know that, and just keeps its clock running...

In Joker I now take care to access the complete hash table before announcing myself to the GUI. Otherwise some GUIs might just start the clock or declare a time loss before the enegine even starts, if the OS decides to shelf the engine for some time while saving memory contents from previous processes it still deems useful to disk. Especially under Arena this seemed a big problem (I was told).

[swami]

I noted that some engines one some platforms have a big problem playing blitz, because of the time it takes to allocate memory for their hash table. Sometimes the OS decides to create a swap file for this, which can take dozens of seconds for a 100MB hash table.

Problem is that the OS does this only on first use: in the task manager you can see that the engine starts with a very small memory foortprint, and then blows up as it starts using the hash table. But if the engine only starts using the hash table after it gets out of book, it is basically dead while the OS is allocating the space. But the GUI doen't know that, and just keeps its clock running...

In Joker I now take care to access the complete hash table before announcing myself to the GUI. Otherwise some GUIs might just start the clock or declare a time loss before the enegine even starts, if the OS decides to shelf the engine for some time while saving memory contents from previous processes it still deems useful to disk. Especially under Arena this seemed a big problem (I was told).

Hi HG,

Joker did just great in my tournament and there was no issue or any kind of problem with either the usage of hash tables or an operating system.
Windows task manager did show correct usage of hash table for all the engines and I don't get any crashes or losses on time, and GUI is responding well to the engine's choice of moves, When the engine gets out of its book,it immediately starts calculating without a loss of time.Maybe you were usign older betas or one of the older official versions of Arena?

Regards.

[swami]

TOURNAMENT FINISHED:
Division 6

Code: Select all

Engine          Score   WC Jo Ma Sc Na Es Ch Ro An Al EX Kn Gi Do Fl Ne    S-B
01: WChessUCI       23.5/30 ·· 1= 01 =1 11 10 11 11 =1 =1 11 =1 11 1= 11 =0  338.25
02: Joker1108w      22.0/30 0= ·· 11 01 01 11 11 == 11 11 01 11 11 =0 =1 =1  313.00
03: Matacz_1.3_HT37 18.5/30 10 00 ·· 1= 10 01 01 1= 11 11 01 =0 1= =1 =1 =1  259.50
04: Scidlet         16.0/30 =0 10 0= ·· 00 00 10 1= 11 01 10 10 11 1= 1= 1=  223.75
05: Natwarlal       15.0/30 00 10 01 11 ·· 01 10 01 =0 10 10 00 01 =1 11 10  215.50
06: Esc 1.16        15.0/30 01 00 10 11 10 ·· =0 0= =0 == =0 01 01 =1 =1 11  213.50
07: Chispa403-p4    14.5/30 00 00 10 01 01 =1 ·· =0 10 10 11 11 =0 10 10 10  200.75
08: Rotor 0.2       14.0/30 00 == 0= 0= 10 1= =1 ·· 0= =0 01 10 1= 1= =1 0=  199.00
09: Anechka008      14.0/30 =0 00 00 00 =1 =1 01 1= ·· 11 == 01 01 =0 =0 11  192.25
10: AliUci408       14.0/30 =0 00 00 10 01 == 01 =1 00 ·· =1 1= =1 =1 00 11  190.75
11: EXchess_console 13.5/30 00 10 10 01 01 =1 00 10 == =0 ·· 00 01 10 1= 11  192.50
12: Knightx192      13.0/30 =0 00 =1 01 11 10 00 01 10 0= 11 ·· 00 =1 00 01  191.00
13: Gibbon          12.5/30 00 00 0= 00 10 10 =1 0= 10 =0 10 11 ·· 1= 01 10  169.00
14: Dorky 3.48      12.0/30 0= =1 =0 0= =0 =0 01 0= =1 =0 01 =0 0= ·· =1 10  181.25
15: Flux            12.0/30 00 =0 =0 0= 00 =0 01 =0 =1 11 0= 11 10 =0 ·· 01  168.00
16: Neurosis        10.5/30 =1 =0 =0 0= 01 00 01 1= 00 00 00 10 01 01 10 ··  163.50

240 games played / Tournament is finished
Name of the tournament: Division 6
Site/ Country: ADMIN, United States
Level: Blitz 1/1
Hardware: Intel(R) Pentium(R) 4 CPU 2.60GHz  with 248 MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: D:\Arena 1.99\Tournaments\Division 6.pgn
Website: 
E-Mail Address: