Author: Karsten Bauermeister
Date: 16:25:53 03/24/98
Go up one level in this thread
On March 24, 1998 at 18:58:22, Dirk Frickenschmidt wrote:
>Hi Karsten,
>
>first of all thanks for playing such a nice tournament and for all the
>interesting games you put here!
>
>One question:
>How did the considerable hardware advantage of Mchess on PII-333Mhz over
>the second, Fritz5, on PII-233Mhz and all other programs occur (Mchess
>being nearly 50% faster than many others)? Who brought the fast Mchess
>machine?
>
>Now sorry I have o correct some comment of yours:
>
>[Event "Mensch-Computer-Turnier"]
>[Site "Aufsess"]
>[Date "1998.??.??"]
>[Round "1"]
>[White "Fritz 5 PII-233"]
>[Black "Tasc R40 Arm2-30"]
>[Result "1-0"]
>[ECO "B34"]
>[PlyCount "89"]
>[EventDate "1998.03.??"]
>{100032 KB Hash Tables, Power-Books} 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4.
>Nxd4
>Nf6 5. Nc3 g6 6. Nxc6 bxc6 7. e5 Ng8 8. Bc4 Bg7 9. Qf3
>{Ab hier muessen beide Rechner selbst nachdenken
>translation: from here both machines have to calculate their moves}
>
>This seems to be wrong: as far as I can see Fritz5 still is in powerbook
>with 10. Bf4.
>Only after the real black fault 10...f6? (which has not been played by
>humans before for good reasons) white is finally out of book.
>
>But you wrote:
>
>9... e6 {Der erste errechnete Zug und gleich der spielentscheidende
>Fehler!
>-1.1: Die gefuerchtete Killerbibliothek?
>translation: the first calculated move and at the same time the fault
>deciding the game
>-1.1: the feared killer-book?}
>
>This comment seems to be quite misleading in several ways:
>
>a) In contrast to some other programs the Fritz5 powerbook does not
>contain any "feared killerbook", at least not as far as I have seen and
>heard until now.
>I have only seen some allegations from Ossi Weiner so far, which were
>missing *any* evidence for raising such serious suspicion and which I
>until now have to consider as quite dirty unless finally proven by
>facts.
>
>b) the position is well known, from the game Leonhard-Tartakower in
>Karlsbad 1907 up to Suetin - Kortchnoi (played in an URS championship in
>1954) and even during the the 80s of our century played by players like
>Nigel Short and Judit Polgar.
>All in all a good database should show more than 80 games with the moves
>up to Qf3, from the beginning of the century up to the late 90s.
>So asking if this was a a killer line is far from reality from this view
>again.
>The contrary is true.
>
>c) It may indeed be questionable if black should choose such a system at
>all, paying with very slow development for this kind of fianchetto, and
>indeed gaining not more than 32% in my database.
>
>But I really doubt if 9...e6 is the move to be criticized as the losing
>move. At this point black already had taken back the knight to g8,
>resulting in a position which no modern chess program would prefer for
>black, being that much back in development.
>Instead e6, in fact being the main human choice in this position
>(besides ...f5 and ...f6) hardly can be blamed: it's only the logical
>consequence of what has happened before - and besides was preferred by
>players like Tartakower or Kortchnoi. Both lost, but hardly for ...e6.
>
>10. Bf4 f6 *This* is the first move unknown in my database, and probably
>not the best.
>
>
>Hope I could help.
>
>
>Kind regards from Dirk
Hello Dirk,
sorry but you set wrong preferences in your first point! Fritz was
playing with special, non-commerical Power-Books, provided by ChessBase
especially for this tournament to the operator! So you can't compare
them to the normal Power-Books.
In your second point you may be right. These comments were taken in a
very short analysis when getting the game into ChessBase for the
bulletin. Some of my comments are a little bit ironic ("Killer-Books")
and shouldn't be taken as "gods words". For me and for Fritz 5 9...e6
seemed to be the fault. The black squares c5, d6, e7 are weak and
therefore the black king will not come to castle!?
But as I said, this was a very short and superficial analysis.
The fast machine of M-Chess was brought by Klaus Fuhrwerk, the organizer
of this tournament. He bought it only a few days before the first round.
M-Chess was doing 35.000 - 45.000 positions per second in the middlegame
on it!
Karsten
This page took 0.01 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.