Alaric 707 Gauntlet- Updated Results & PGNs

[geots]

Alaric 707 Gauntlet

Intel Pentium 4 (2.8ghz)
128mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)
40 moves in 62 minutes repeating (adapted to CCRL 40/40)
All engines single cpu 32-bit
Alaric 707 will play 20 games against each opponent
Run by George Speight



Alaric 707

(4.0-3.0) Glaurung 2 epsilon/5
(3.0-4.0) Chess Tiger 2007.1
(3.0-4.0) CM9000 Enforcer
(3.5-3.5) SmarThink 1.00
(2.5-4.5) Scorpio 1.91
(4.0-3.0) Delfi 5.1
(4.5-2.5) SlowChess Blitz WV2.1
(4.5-2.5) CM10th Guardian
(4.5-2.5) Ruffian 2.1.0
(3.0-4.0) WildCat 7
(4.5-2.5) Pro Deo 1.2
(2.0-4.0) Gandalf 6


Games available for download from:
http://kirr.homeunix.org/chess/discussi ... f=7&t=2498

[pichy]

Alaric 707 Gauntlet

Intel Pentium 4 (2.8ghz)
128mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)
40 moves in 62 minutes repeating (adapted to CCRL 40/40)
All engines single cpu 32-bit
Alaric 707 will play 20 games against each opponent
Run by George Speight




Alaric 707

(4.0-3.0) Glaurung 2 epsilon/5
(3.0-4.0) Chess Tiger 2007.1
(3.0-4.0) CM9000 Enforcer
(3.5-3.5) SmarThink 1.00
(2.5-4.5) Scorpio 1.91
(4.0-3.0) Delfi 5.1
(4.5-2.5) SlowChess Blitz WV2.1
(4.5-2.5) CM10th Guardian
(4.5-2.5) Ruffian 2.1.0
(3.0-4.0) WildCat 7
(4.5-2.5) Pro Deo 1.2
(2.0-4.0) Gandalf 6


Games available for download from:
http://kirr.homeunix.org/chess/discussi ... f=7&t=2498

Alaric 707 has made very good progress, here is the result against Movei438

32mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)


Engine Score Al Mo S-B
1: AlaricWB707 36.0/50 ·················································· =0===10==111=111111=1====1110=1101=11=0111=1=111=1 504.00
2: Movei00_8_438 14.0/50 =1===01==000=000000=0====0001=0010=00=1000=0=000=0 ·················································· 504.00

50 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 1/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:

[Uri Blass]

Alaric 707 Gauntlet

Intel Pentium 4 (2.8ghz)
128mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)
40 moves in 62 minutes repeating (adapted to CCRL 40/40)
All engines single cpu 32-bit
Alaric 707 will play 20 games against each opponent
Run by George Speight




Alaric 707

(4.0-3.0) Glaurung 2 epsilon/5
(3.0-4.0) Chess Tiger 2007.1
(3.0-4.0) CM9000 Enforcer
(3.5-3.5) SmarThink 1.00
(2.5-4.5) Scorpio 1.91
(4.0-3.0) Delfi 5.1
(4.5-2.5) SlowChess Blitz WV2.1
(4.5-2.5) CM10th Guardian
(4.5-2.5) Ruffian 2.1.0
(3.0-4.0) WildCat 7
(4.5-2.5) Pro Deo 1.2
(2.0-4.0) Gandalf 6


Games available for download from:
http://kirr.homeunix.org/chess/discussi ... f=7&t=2498

Alaric 707 has made very good progress, here is the result against Movei438

32mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)


Engine Score Al Mo S-B
1: AlaricWB707 36.0/50 ·················································· =0===10==111=111111=1====1110=1101=11=0111=1=111=1 504.00
2: Movei00_8_438 14.0/50 =1===01==000=000000=0====0001=0010=00=1000=0=000=0 ·················································· 504.00

50 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 1/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:

I remember that you posted better result of movei against alaric at 3+1
Alaric is better than movei but your result seems to me too high(I admit that I did not test movei against alaric).

I wonder if you use the same personality that you use in the first match based on the same change files.

http://64.68.157.89/forum/viewtopic.php?t=15758

Uri

[pichy]

Alaric 707 has made very good progress, here is the result against Movei438

32mb hash each
3 & 4 man tablebases
Perfect 12.32.ctg book (with 12 move CCRL limit)


Engine Score Al Mo S-B
1: AlaricWB707 36.0/50 ·················································· =0===10==111=111111=1====1110=1101=11=0111=1=111=1 504.00
2: Movei00_8_438 14.0/50 =1===01==000=000000=0====0001=0010=00=1000=0=000=0 ·················································· 504.00

50 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 1/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:[/quote]



I remember that you posted better result of movei against alaric at 3+1
Alaric is better than movei but your result seems to me too high(I admit that I did not test movei against alaric).

I wonder if you use the same personality that you use in the first match based on the same change files.

http://64.68.157.89/forum/viewtopic.php?t=15758

Uri[/quote]

I used the same personality, but I believe that Movei perform better at 3+1 or 5+1 than 1+1 time control. I am running another short match 0f 20 games at 5+1, just to compare the result

[Uri Blass]

I do not know if this is a possible reason for better results of movei at longer time control but maybe at fast time control movei is wasting too much time on printing the pv.

maybe it is better for fast time control if I tell movei to print pv information only if it searched more than 1,000,000 nodes.

I doubt if there is an improvement at long time control and my experience against fruit2.0 does not suggest better result at longer time control.

Hopefully I will be able to finish the match against fruit2.0 with 40 minutes/30 moves.

The last time when I looked at the result yesterday the result was 23-23 after 46 games and I think that fruit2.0 is one of the program that is relatively good at blitz.

Uri

[pichy]

I do not know if this is a possible reason for better results of movei at longer time control but maybe at fast time control movei is wasting too much time on printing the pv.

maybe it is better for fast time control if I tell movei to print pv information only if it searched more than 1,000,000 nodes.

I doubt if there is an improvement at long time control and my experience against fruit2.0 does not suggest better result at longer time control.

Hopefully I will be able to finish the match against fruit2.0 with 40 minutes/30 moves.

The last time when I looked at the result yesterday the result was 23-23 after 46 games and I think that fruit2.0 is one of the program that is relatively good at blitz.

Uri

20 games is not a good way of testing two engines, but Movei is performing better at Blitz 3/1, or it is just another lucky match for Movei


Engine Score Al Mo S-B
1: AlaricWB707 11.5/20 ···················· 1=101=0110===01==011 93.75
2: Movei00_8_438 8.5/20 0=010=1001===10==100 ···················· 93.75

20 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 3/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:

[pichy]

I do not know if this is a possible reason for better results of movei at longer time control but maybe at fast time control movei is wasting too much time on printing the pv.

maybe it is better for fast time control if I tell movei to print pv information only if it searched more than 1,000,000 nodes.

I doubt if there is an improvement at long time control and my experience against fruit2.0 does not suggest better result at longer time control.

Hopefully I will be able to finish the match against fruit2.0 with 40 minutes/30 moves.

The last time when I looked at the result yesterday the result was 23-23 after 46 games and I think that fruit2.0 is one of the program that is relatively good at blitz.

Uri

20 games is not a good way of testing two engines, but Movei is performing better at Blitz 3/1, or it is just another lucky match for Movei


Engine Score Al Mo S-B
1: AlaricWB707 11.5/20 ···················· 1=101=0110===01==011 93.75
2: Movei00_8_438 8.5/20 0=010=1001===10==100 ···················· 93.75

20 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 3/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:

-------------------------------------------------------------------------------------

I decided to continue the match up to 50 games, and Movei438 scored much better at 3+1 than at 1+1

Engine Score Al Mo S-B
1: AlaricWB707 27.5/50 ·················································· 1=101=0110===01==010101===1==10=10=1=011=0=01=01=1 604.75
2: Movei00_8_438 22.5/50 0=010=1001===10==101010===0==01=01=0=100=1=10=10=0 ·················································· 604.75

50 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 3/1
Hardware: AMD Athlon(tm) Processor 1202 MHz with 512MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)
PGN-File: C:\Program Files\Arena\Arena.pgn
Website:
E-Mail Address:

[Uri Blass]

My opinion is that it is not movei that is better at longer time control but it is arena that cheat against programs who print the pv often.

Movei is one of the program that print the pv often so the interface of arena dislike movei at fast time control.

To make the effect as big as possible on my computer
I made an experiment and decided to play a match of 1+0 between 2 versions of movei(one of them prints the pv and one of them does not print the pv).

I stopped the match after 52 games

result was significant
40-12 for the new version

1: Movei00_8_447 40.0/52 ···················································· 0101==1=01==1=11===10111=11111=11=11=11111111111011= 480.00
2: Movei00_8_446 12.0/52 1010==0=10==0=00===01000=00000=00=00=00000000000100= ···················································· 480.00


Did I make a big improvement in bullet by not printing the pv?
I do not think that it is the case and it seems to me that arena made a big cheat against the version that I told to print the pv often and stole time from it because the time that it take to print the pv is practically almost 0 under winboard.

Unfortunately it is not possible not to print the pv in version 438

Uri

[pichy]

My opinion is that it is not movei that is better at longer time control but it is arena that cheat against programs who print the pv often.

Movei is one of the program that print the pv often so the interface of arena dislike movei at fast time control.

To make the effect as big as possible on my computer
I made an experiment and decided to play a match of 1+0 between 2 versions of movei(one of them prints the pv and one of them does not print the pv).

I stopped the match after 52 games

result was significant
40-12 for the new version

1: Movei00_8_447 40.0/52 ···················································· 0101==1=01==1=11===10111=11111=11=11=11111111111011= 480.00
2: Movei00_8_446 12.0/52 1010==0=10==0=00===01000=00000=00=00=00000000000100= ···················································· 480.00


Did I make a big improvement in bullet by not printing the pv?
I do not think that it is the case and it seems to me that arena made a big cheat against the version that I told to print the pv often and stole time from it because the time that it take to print the pv is practically almost 0 under winboard.

Unfortunately it is not possible not to print the pv in version 438

Uri

Since Arena cheat cheat against programs who print the pv often, I decided to test under Shredder


Athlon, Blitz:5'+1" 0

123456789012345678901234567890123456789012345678901
1 Alaric 707 2760 10100½0100½11011½01½1011½0½11111½½0010½0001½101½000 25.5/51 650.25
2 Movei00_8_438 2700 01011½1011½00100½10½0100½1½00000½½1101½1110½010½111 25.5/51 650.25