Author: Uri Blass
Date: 10:00:58 09/16/02
Go up one level in this thread
On September 16, 2002 at 03:55:47, Brian Katz wrote:
>Is is my imagination, or does Deep-Fritz 7 on a single processor play horribly?I
>have found Deep Fritz 7 leaving it's King in the center and getting crushed,
>where as I was getting execllent results with the Fritz 7 engine after the first
>few engine upgrades.
>Perhaps it is just the Deep Fritz 7 opening book which is the problem, I really
>do not know. What I do know is, that the results and style of play seem to be
>quite different.
>Has anyone had similar results?
>Also, someone recently mentioned that his Fritz 7 engine seemed to be playing at
>only 1900+ strength after the most recent Chessbase upgrade. Has anyone had
>similar resutls?
>All replies welcome.
>Regards
>Brian
Your results suggest that Deep Fritz7 is better than Fritz7
There is not enough evidence to prove it because you did not play hundreds of
games but DF7 scored better than F7.
I give your results again with some comment of me inside the **
I assume that the posts later may be deleted by the moderators because accusing
chessbase of lying with no evidence should be deleted so I copied your post with
the data that suggest that Deep Fritz7 is better.
Copied from your previous post:
I have had some engine tournaments where Deep-Fritz did not perform well at all.
I'll send the crosstables and time controls used. As far as opeing books are
concerned, each program used their own opening book, since programs are supposed
to perform better when using their own books, so I am told.
DF7 vs F7-2 Eatontown 2002
1234
1 Deep Fritz 7 0½1½ 2.0/4 4.00 This was a Game 30 32 mb hash
2 Fritz 7 1½0½ 2.0/4 4.00
DF7 vs F7 Eatontown 2002
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
1 Deep Fritz 7 ½ ½ 0 ½ ½ 1 ½ 1 1 ½ 0 1 ½ 0 ½ 1 ½ 0 ½ 1 11.0/20
2 Fritz 7 ½ ½ 1 ½ ½ 0 ½ 0 0 ½ 1 0 ½ 1 ½ 0 ½ 1 ½ 0 9.0/20
This was a Game 2 where DF7 did ok 16 mb hash
Blitz:5' Eatontown 0
12345678901234567890
1 Fritz 7 1½½½01½½1½100½½½½½½½ 10.5/20
2 Deep Fritz 7 0½½½10½½0½011½½½½½½½ 9.5/20 This was a Game 5 16 mb Hash
Blitz:5'+12" 0
123456
1 Deep Fritz 7 ½10½½1 3.5/6
2 Fritz 7 ½01½½0 2.5/6 This was ok for DF7 32 mn hash
5 12 32MB Eatontown 2002
1 2 3 4 5
1 Deep Fritz ** ½1 ½½ 11 1½ 6.0/8
2 Fritz 6 ½0 ** 10 11 1½ 5.0/8
3 Deep Fritz 7 ½½ 01 ** 0½ ½½ 3.5/8
4 Fritz 7 00 00 1½ ** 1½ 3.0/8
5 Hiarcs 8 0½ 0½ ½½ 0½ ** 2.5/8
This was a disaster for Fritz 7 and Deep Fritz 7
*No it was not a disaster
Deep Fritz7 scored better than Fritz7
Your claim was about comparison between Deep Fritz7 and Fritz7 so the tournament
suggest that Deep Fritz7 is better*
2 MIN EXTRAVAGANZA 16 MB HASH-1 2002 Game 2min+2 sec incr.
1 2 3 4 5 6 7 8 9
1 Deep Fritz ** ½1 ½1 1 00 1 01 1 1 8.0/12
2 Fritz 7 ½0 ** ½½ 01 ½ 1½ 01 1 1 7.5/13 43.00
3 Fritz 6 ½0 ½½ ** 1 0½ 1 ½ 11 1 7.5/12 41.50
4 Hiarcs 8 0 10 0 ** 1 ½½ 1 11 ½1 7.5/12 37.50
5 Deep Fritz 7 11 ½ 1½ 0 ** 0½ 11 0 0 6.5/12
6 Shredder 6.02 0 0½ 0 ½½ 1½ ** ½ ½1 0½ 5.5/13 31.25
7 Shredder 6 10 10 ½ 0 00 ½ ** 0½ 11 5.5/13 30.00
8 Crafty 18.14 0 0 00 00 1 ½0 1½ ** 1 4.0/12
9 Crafty 18.15 0 0 0 ½0 1 1½ 00 0 ** 3.0/11
The above tournament was a disaster for DF7
*No it was not a disaster,DF7 played against strong opponents and scored more
than 50% 6.5/12 and 7.5/13 are almost the same*
This next tournament started out badly for DF7 so I endend it and tried some
other controls. This was a game 2 no increments. 16 mb. I have used lower hash
tables as well.
2 MIN EXTRAVAGANZA 16 MB HASH 2002
1 2 3 4 5 6 7 8 9
1 Crafty 18.14 ** 1 1 2.0/2 2.00
2 Crafty 18.15 ** 1 1 2.0/2 1.00
3 Fritz 7 ** 0 ½ 1 1.5/3
4 Shredder 6 0 0 1 ** 1.0/3 1.50
5 Deep Fritz ½ ** ½ 1.0/2 1.00
6 Fritz 6 0 0 ** 1 1.0/3 0.50
7 Hiarcs 8 ** 1 1.0/1 0.00
8 Deep Fritz 7 ½ 0 ** 0.5/2
9 Shredder 6.02 0 0 ** 0.0/2
*0.5/2 is result that you can expect often when one player is less than 100 elo
better than the opponents*
This went very well for DF7 at 5/12 32 mn hash
DF7vsF7 32mb 5-12 2002
123456
1 Deep Fritz 7 1½½1½1 4.5/6
2 Fritz 7 0½½0½0 1.5/6
Deep Fritz 7 has had some good results too.
I have noticed however, that it has been leaving it's King in the center in a
few games and getting crushed by playing superficial attacks. Advancing the h
pawn as white to h4 and h5 etc while it's king is still in the center. Trying
attacks on both wings while it's king is still in the center.
*Total result is simply better for Deep Fritz7 and even if in some positions
Fritz7 plays better it does not mean that Deep Fritz7 is no imporovement.
You can expect maybe improvement of 20-30 elo from Deep Fritz7.
Chessbase never claimed that it is 100 elo better
Uri*
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.