Engine Score
Al Mo S-B
1: AlaricWB707 27.0/50
··················································
0=011101=1011==000111100111=01100000=1==1=0=110101 609.00
2: Movei00_8_438 23.0/50
1=100010=0100==111000011000=10011111=0==0=1=001010
·················································· 609.00
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:
______________________________________________________________________
Engine Score Mo Co S-B
1: Movei00_8_438 29.5/50 ·················································· 1=1==1=1=1101011=10=110==1010=001==10==010111===== 604.75
2: Colossus2007c 20.5/50 0=0==0=0=0010100=01=001==0101=110==01==101000===== ·················································· 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:
-------------------------------------------------------------------------------------
Engine Score Co Mo S-B
1: Colossus2007c 25.0/40 ·················································· 1110110111=0110=110110=10011111000=10110 375.00
2: Movei00_8_425 15.0/40 0001001000=1001=001001=01100000111=01001 ·················································· 375.00
40 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:
-----------------------------------------------------------------------
This was the last time that I tested Collosus 2007a, since then it has made a great progress specially versus Movei
I got this result of Colossus 2007a vs Movei403 progress 4 6 10.
Colossus 2007a is a very strong program at Blitz
Engine Score Mo Co S-B
1: Movei00_8_403 18.5/30 1=01=1011001==1==0==1==1=11==1 212.75
2: Colossus2007a 11.5/30 0=10=0100110==0==1==0==0=00==0 212.75
30 games played / Tournament finished
Name of the tournament: Arena tournament
Site/ Country: Jorge, United States
Level: Blitz 5/1
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:
Movei438 is a great improvement over previous Movei.........
Moderator: Ras
-
- Posts: 2564
- Joined: Thu Mar 09, 2006 3:04 am
-
- Posts: 10872
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Movei438 is a great improvement over previous Movei.....
It may be possible to get even bigger improvement by changing personality but it needs to be tested
comments
1)I still allow the progress personality but I doubt if with hash something different than 0 0 0 will be better.
2)I change some numbers relative to default movei and I forgot to change them back and it is possible that mobilitydivider 64 is not the best
or that you need first to change the knight and bishop square tables to make it works.
3)less_reduce 1 that is in the file that I sent you mean that movei does not reduce pass pawn pushes or extended moves.
It seems to help based on my tests so every personality should include less_reduce 1
4)There are some parameters that by default are 0 that mean not used
specailly:
rook_factor 0
a_weaker 0
extreme_pawn_array 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0history_verification_number 1
It may be possible to get even bigger improvement by changing personality but it needs to be tested
bigger number for rook_factor mean increasing the value of the rook in the endgame
bigger number for a_weaker mean less value for extreme a or h pawn in the opening and more value for them in the endgame
big history_verification number mean less pruning based on history information when 1 means that history information does not prevent pruning.
comments
1)I still allow the progress personality but I doubt if with hash something different than 0 0 0 will be better.
2)I change some numbers relative to default movei and I forgot to change them back and it is possible that mobilitydivider 64 is not the best
or that you need first to change the knight and bishop square tables to make it works.
3)less_reduce 1 that is in the file that I sent you mean that movei does not reduce pass pawn pushes or extended moves.
It seems to help based on my tests so every personality should include less_reduce 1
4)There are some parameters that by default are 0 that mean not used
specailly:
rook_factor 0
a_weaker 0
extreme_pawn_array 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0history_verification_number 1
It may be possible to get even bigger improvement by changing personality but it needs to be tested
bigger number for rook_factor mean increasing the value of the rook in the endgame
bigger number for a_weaker mean less value for extreme a or h pawn in the opening and more value for them in the endgame
big history_verification number mean less pruning based on history information when 1 means that history information does not prevent pruning.
-
- Posts: 10872
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Movei438 is a great improvement over previous Movei.....
My result so far with that version are the following results:
1)Movei-Fruit2.0 1+1 55-45
2)Movei-Glaurung1.2.1 3/40 13-11(FRC)
3)Movei-Fruit2.0 5/40 21.5-18.5
Note that it is very dangerous to get conclusion based on small number of games.
In the first match against fruit if you divide the match to 5 matches of 30 games results were as follow:
1)12-8
2)10-10
3)6.5-13.5
4)13.5-6.5
5)13-7
It seems that movei is better than fruit2.0 but
you can see that if you use 20 games(match number 3) or 40 games(match numbers 2-3) you get win for fruit2.0
Even 60 games(matches number 2-4) may give a misleading score of 30-30.
1)Movei-Fruit2.0 1+1 55-45
2)Movei-Glaurung1.2.1 3/40 13-11(FRC)
3)Movei-Fruit2.0 5/40 21.5-18.5
Note that it is very dangerous to get conclusion based on small number of games.
In the first match against fruit if you divide the match to 5 matches of 30 games results were as follow:
1)12-8
2)10-10
3)6.5-13.5
4)13.5-6.5
5)13-7
It seems that movei is better than fruit2.0 but
you can see that if you use 20 games(match number 3) or 40 games(match numbers 2-3) you get win for fruit2.0
Even 60 games(matches number 2-4) may give a misleading score of 30-30.