Killer Books
A.Mader
A few weeks ago a computer tournament was held in Wels (Austria) using
Chrilly Donningers autoplayer. Every participant played two games
against every other participant and....
- drum roll -
... the opening books were disabled!!!
I do not know the exact result (if enough people are intereted I will
try to get it), but:
(1) Fritz won the tournament.
(2) M-Chess was in the second half of the tournament table.
(3) Nimzo became the king of draws.
Just another piece to complete the puzzle.....
Best wishes
Andreas Mader
"The last good thing written in C was Schuberts Symphony Number 9"
mclane
would be quoting out of it.
Here again the same rule:
can somebody show/post us the complete text file of
games so that we could study ourselves
If
Mchess
played
so
weak
without
opening
book.
Komputer Korner
A.Mader wrote:
>
> A few weeks ago a computer tournament was held in Wels (Austria) using
> Chrilly Donningers autoplayer. Every participant played two games
> against every other participant and....
>
> - drum roll -
>
> ... the opening books were disabled!!!
>
> I do not know the exact result (if enough people are intereted I will
> try to get it), but:
>
> (1) Fritz won the tournament.
> (2) M-Chess was in the second half of the tournament table.
> (3) Nimzo became the king of draws.
>
> Just another piece to complete the puzzle.....
>
> Best wishes
> Andreas Mader
>
> "The last good thing written in C was Schuberts Symphony Number 9"
Do you know what the time control was. I can't believe that Fritz
would win at a long time control?
--
Komputer Korner
The komputer that couldn't kompute the square root of
36^n.
Robert Hyatt
Komputer Korner (kor...@netcom.ca) wrote:
Then you don't understand "statistics". Remember Fritz won the *unlimited*
world computer chess championship in 1995, at 40/2 time control. It is
*not* a slouch, although I'd agree with others that if you played the Jakarta
event 4 times, you'd possibly have 4 different winners. There's still luck
in any rating based on < 20 or 30 games...
A.Mader
Komputer Korner wrote:
>
> A.Mader wrote:
> >
> > A few weeks ago a computer tournament was held in Wels (Austria) using
> > Chrilly Donningers autoplayer. Every participant played two games
> > against every other participant and....
> >
> > - drum roll -
> >
> > ... the opening books were disabled!!!
> >
> > I do not know the exact result (if enough people are intereted I will
> > try to get it), but:
> >
> > (1) Fritz won the tournament.
> > (2) M-Chess was in the second half of the tournament table.
> > (3) Nimzo became the king of draws.
> >
> > Just another piece to complete the puzzle.....
> >
> > Best wishes
> > Andreas Mader
> >
> > "The last good thing written in C was Schuberts Symphony Number 9"
>
> Do you know what the time control was. I can't believe that Fritz
> would win at a long time control?
I tried to phone Franz Wiesenecker who is responsible for this
tournament, but he is on holidays. As soon as he is back I am able to
get more detailed information. I don't know exactly, but I think that
the time controls were 40/120 and 20/60. You must not overestimate one
single tournament, maybe Fritz was just "lucky" or maybe his opening
book is weak so Fritz had an advantage when they were disabled, or....
Best wishes
A. Mader
A.Mader
Thorsten Czub alias McLane wrote:
>AGAIN: I don't want you to do the list for me. But I like to discuss
about
>the list like some other guys like to discuss about
>Mchess and Marty Hirsch! You got the point?!
This is OK, but I think it is not WHAT you are discussing, it is the way
HOW you are discussing.
>>
>> You are quite right! PLEASE, don't change anything in the making of the
>> Swedish rating list!
>
>Brilliant, don't change doubles and tripples.
>Let the programs play always the same games again and
>DON'T CHANGE ANYTHING!!
You have got the point!
I posted an article in rgcc and explained why I think that removing
doubles will be contraproductive.
>> I learned that most of the guys who argue against the SSDF don't even
>> know what the margin of error means. Do you expect profound critics from
>> such people?
>
>Brilliant!
>Don't answer to the people below IQ < 60 !
Yes! I meant exactly you with my "profound critics" statement! How did
you recognize?
>Next conclusion:
>Don't answer to people wearing JEANS...
No, my next conclusion will be: Don't answer people hiding behind other
names. They will have their reasons to do so!
>> Everybody who thinks he can make it better is free to play 50000 or more
>> games according to his prefered rules!
>> The authors of chess programs are free to implement learning features so
>> that older programs can "defend" against killer openings.
>
>That is exactly what I want to explain Moritz and Dirk and others for
>many days now.
I am glad that we agree at least in some aspects!
>So you see: without knowing much about error margins (you have those
>PROFUND KNOWLEDGE, I know) we come to the same result!!
Maybe you came to the right result for the wrong reasons?! This is a
well known aspect of computer chess programs!
>Despite the fact that I would never give the advice: Don't change
>anything!
>Such a result is nice if the listener is shortly before dying, or - at
>least - dead.
Another profound criticism copyrighted by Thorsten Czub.
>> Not long ago Ed Schroeder argued strongly against the list, now he is
>> the one and only number 1 with his Rebel 8.0 and he is using the list as
>> an advertising argument.
>
>
>What can you do else with such a list.
>It's just a tool. So it is used like a tool.
>Don't understand what you wanted to say to us with the above "Ed
>Schroeder"
OK, I will try to be more precise:
Schroeder, NOT being Number 1:
SSDF is bad! The list is bad! Says nothing about the real strength!
Schroeder, BEING Number 1:
Look! I am Number 1 in the SSDF list! Good list!
You see the inconsistency?
Either is a tool of no use or it is very useful. When the SSDF list was
bad when Rebel was not number 1, how can it be good when Rebel IS number
1?
>> >
>> > But we are pleased that so many people around the world like what we >have
>> > done in the past and are doing in the present. We are indeed listening to
>> > them, and we try to answer questions about our work.
>> >
>> > For the moment there are many, many letters in the r.g.c.c. about killerbooks
>> > and repeated games. We are reading all the letters, and we are discussing
>> > this subject as much as we have time for. We have also discussions with several
>> > of the programmers, amongst them Ed Schroeder and Marty Hirsch.
>> >
>>
>> Don't change anything! Don't stop counting doubles! Hopefully this will
>> force the programers to develop intelligent learning features.
>
>Aha. You want to vehicle/transpot/make evolution!
What else? Isn't this your opinion that the SSDF list is just a tool?
You have to be consequent!
>>So you want KILLER-BOOKS because they force the programmers to program
>>intelligent algos that react on other programs, and learn BTW ?
I don't want killer books and I don't want the SSDF people to do the job
for the programmers.
>> <snip>
>> >
>> > Everybody must have understood, that there is no easy solution with making a
>> > list, that can satisfy all parts. Indeed it must be impossible.
>> >
>>
>> Fully agreed!
>>
>> Best wishes
>> Andreas Mader
>
>Also best wishes, but still not believing in the list and still not
>convinced because you are Herr Mader.
Sorry to hear that, Herr Czub.
>I know you too!! From several published articles.
>I found the PC-Schach and the MODULE also always very strange!!
There would have been something wrong with "Modul" and "PC-Schach", if
YOU wouldn't have found it strange!
BTW: I am sharing some of your opinions (not that ones regarding the
SSDF) but your writing stile makes it very difficult for me to "confess"
this.
Best wishes
Andreas Mader
A.Mader
The time control of this tournament is/was 60 moves / 2 hours.
When F. Wiesenecker has the time he will add new programs to the
tournament and that is why there are no "final" standings. Although
every program has played 2 games against every other program, the number
of participants may grow. Hopefully Rebel 8 and M-Chess 6 (also without
opening libraries) will be added.
The scores of the programs are very close so the entrance of a new
participant will change the order! Please don't overestimate one single
tournament!
1. Fritz 3 19.0
2. Quest 3 17.5
Rebel 7 17.5
4. M-Chess 4 16.5
5. Genius 2 16.0
Kallisto 16.0
Nimzo 3 16.0
8. Genius 3 15.5
9. Rebel 6 15.0
10. Hiarcs 4 14.5
M-Chess 5 14.5
12. W-Chess 14.0
13. Hiarcs 3 13.0
14. Gandalf 3.0
15. Diogenes 2.0
The games were played via AUTO232 with two PCs with the same hardware
and the same configuration.
Andreas Mader
Tord Kallqvist Romstad
A.Mader (andrea...@siemens.at) wrote:
: 1. Fritz 3 19.0
: 2. Quest 3 17.5
What is the difference between Fritz 3 and Quest 3?
Tord
Komputer Korner
A.Mader wrote:
> snipped info on non opening book tourney in Austria.
If these tournaments become popular, we will have to be careful of
hidden camouflaged opening books. What a monstrous thought!!!!
--
Komputer Korner
The komputer that couldn't keep a password safe from prying eyes and
mcl...@prima.ruhr.de
In article <327FC3...@siemens.at>, "A.Mader" <andrea...@siemens.at>
writes:
> 1. Fritz 3 19.0
> 2. Quest 3 17.5
> Rebel 7 17.5
> 4. M-Chess 4 16.5
> 5. Genius 2 16.0
> Kallisto 16.0
> Nimzo 3 16.0
> 8. Genius 3 15.5
> 9. Rebel 6 15.0
>10. Hiarcs 4 14.5
> M-Chess 5 14.5
>12. W-Chess 14.0
>13. Hiarcs 3 13.0
>14. Gandalf 3.0
>15. Diogenes 2.0
>
>The games were played via AUTO232 with two PCs with the same hardware
>and the same configuration.
>
>Andreas Mader
Where can we get the games? I want to replay them!
thanks.
Tim Mirabile
Komputer Korner <kor...@netcom.ca> wrote:
>A.Mader wrote:
>> snipped info on non opening book tourney in Austria.
>
>If these tournaments become popular, we will have to be careful of
>hidden camouflaged opening books. What a monstrous thought!!!!
Or, god forbid, an algorithm that produces good opening moves!!!!
+----------------------------------------------------------------------+
| Tim Mirabile <t...@mail.htp.com> http://www.webcom.com/timm/ |
| TimM on FICS - telnet://fics.onenet.net:5000/ PGP Key ID: B7CE30D1 |
+----------------------------------------------------------------------+
A.Mader
Tord Kallqvist Romstad wrote:
>A.Mader (andrea...@siemens.at) wrote:
>: 1. Fritz 3 19.0
>: 2. Quest 3 17.5
>
>What is the difference between Fritz 3 and Quest 3?
>
>Tord
I asked F. Wiesenecker the same question and he answered that Quest is
an experimental version with some improvements. Sorry, don't know more
about it.
Andreas Mader
A.Mader
Thorsten Czub alias mclane wrote:
:In article <327FC3...@siemens.at>, "A.Mader"
:<andrea...@siemens.at>
:writes:
:
:>A.Mader wrote:
:
<snip>
:
:>The time control of this tournament is/was 60 moves / 2 hours.
:>When F. Wiesenecker has the time he will add new programs to the
:tournament and that is why there are no "final" standings. Although
:>every program has played 2 games against every other program, the
number
:>of participants may grow. Hopefully Rebel 8 and M-Chess 6 (also
without
:>opening libraries) will be added.
:>
:>The scores of the programs are very close so the entrance of a new
:>participant will change the order! Please don't overestimate one
single
:>tournament!
:>
:> 1. Fritz 3 19.0
:> 2. Quest 3 17.5
:> Rebel 7 17.5
:> 4. M-Chess 4 16.5
:> 5. Genius 2 16.0
:> Kallisto 16.0
:> Nimzo 3 16.0
:> 8. Genius 3 15.5
:> 9. Rebel 6 15.0
:>10. Hiarcs 4 14.5
:> M-Chess 5 14.5
:>12. W-Chess 14.0
:>13. Hiarcs 3 13.0
:>14. Gandalf 3.0
:>15. Diogenes 2.0
:>
:>The games were played via AUTO232 with two PCs with the same hardware
:>and the same configuration.
:>
:>Andreas Mader
:
:Where can we get the games? I want to replay them!
:
:thanks.
Franz Wiesenecker sended all games to CSS. Maybe they are on the
"Servicediskette". I think the Nimzo games that came in later are not on
the disc. Hope this helps.
Andreas Mader
Gernot Holm
> >: 2. Quest 3 17.5
> >
> >What is the difference between Fritz 3 and Quest 3?
> >
> >Tord
>
> I asked F. Wiesenecker the same question and he answered that Quest is
> an experimental version with some improvements. Sorry, don't know more
> about it.
>
> Andreas Mader
Actually Quest is a (maybe experimental) version from the AEGON -
Tournament 1995. It identifies itself as "Fritz 3.1 / Quest 3.1" and
is running in text mode.
Gernot Holm
Harald Faber
Hello tordro # ifi.uio.no,
I think I have an interesting aspect to tell you:
TKR> A.Mader (andrea...@siemens.at) wrote:
TKR> : 1. Fritz 3 19.0
TKR> : 2. Quest 3 17.5
TKR>
TKR> What is the difference between Fritz 3 and Quest 3?
Quest is an experimental version of fritz.
Ciao and see ya
Harald
--