ICCA members and other interested in Computer Chess
Hans Berliner
(ICCA) journal there appears an editorial which eulogizes Dr. M. Botvinnik
for his contribution to computer chess. This is followed by an article by
Botvinnik. The very first example of this article contains a number of
flaws, one so serious that in the position in which Bovinnik's program
indicates "white cannot possibly lose and may win", black can immediately
win a rook and the game. At the end of this example Botvinnik opines
"Truly, the manner of operating of a master at the board!". That would
appear to indicate he is extremely pleased with the performance of his
program. Yet three pages later the editors of the "journal" place the
following footnote "Recently, an unfortunate shortcoming in the search tree
--- was discovered. --- This mistake was due to an error that has slipped
into the subroutine that determines fork trajectories. The error will soon
be corrected". This is clearly utter nonsense. If Botvinnik was so pleased
as to praise the performance of his program so highly, how can it be that
the analysis was in obvious error. Undoubtedly there are errors in
programs, but an author does not present an example of performance until it
meets the authors standards.
I took this opportunity to write a 10-page article entitled "Playing
Computer Chess in the Human Style". The article includes an appraisal of
Botvinnik's genuine literature contributions, as well as the contributions
of many others. It also dealt with the fraudulent presentation about what
his program is supposed to be doing. The article was refused publication by
ICCA as a rebuttal of the Botvinnik article, unless I was willing to submit
to major revisions. This raises many questions about whether an author is
entitled to write his own papers, and to what extent an organization may
inhibit rebuttal of falsehoods published in its journals. My article is
quite direct in appraising Botvinnik. If he wishes to put himself forward
as some sort of apostle, and has the help of the ICCA in doing this, then he
and they must be prepared to stand the heat that comes from a fraudulent
presentation. Botvinnik is regarded as a dubious participant by almost
everyone at the forefront of computer chess. That is because he never puts
a program forward to compete even when offered extremely good conditions.
A copy of the article may be obtained by:
1) Ftp'ing to k.gp.cs.cmu.edu
Logging in as "anonymous"
giving any character string as password
doing "cd /usr/berliner/public"
then typing "get Human"
"quit" and you are back to where you were.
2) If that seems hard or you want an electronic or hard copy, reply with your e-mail
or postal address.
Feng-Hsiung Hsu
found at least one major error in his own article which happened
to concern a post of mine here.
The following came from his article that was rejected by ICCA.
>When I published this on the chess net, F. H. Hsu contributed the following
>information. He and DT2 had analysed this position with Kasparov, and DT2
>was unable to see the whole line to the end. However, by piecemeal analysis
>it became clear that white does indeed have a win, and the longest line is
>3.-Qf5+!; 4. Kg1! (not Kg2 because of Qd5+),Qb1+; 5. Kg2,Qe4+; 6. Kh3,Qf5+;
>7. g4,Qf1+; 8. Kg3,Qg1+; 9. Kf3,Qf1+; 10. Ke3!,Qh3+!; 11. Kd4,e5+!; 12. Kd5,
>Qg2+; 13. Kc5,Qg1+; 14. Kc6,Qh1+ (Qg2+ and Qc1+ are also possible but make
>no difference); 15. Kb6,Qg1+; 16. K:a6. The total variation is 31 ply in
>length. It is no wonder DT2 was not able to find such a deep variation.
And the following is what I actually posted.
>Judging from the selective depth reached, Hitech did not see the win, but
>only saw that Rd8 is at least good for a draw. When we discussed the
>position with Kasparov, he had already analyzed the position in great detail.
>The hardest line requires WK to run all the way (30+ plies) to a7 to avoid
>repetition draw. In order to see the win, the selective search depth
>must reach at least 30 plies. On DT-1, this takes an 11-ply nominal
>search. On DT-2, it takes 10 plies (under tournament time for a 6-processor
^^^^^^^^^^^^^^^^^^^^^
>machine) to see the 30+ plies winning line, and 9 plies (under action time,
>about 1/2 min) to play Rd8, seeing at least a draw. This is with a very
>primitive quiescence search and without repetition detection for the last
>4 plies. A comparable machine with a better quiescence search and with
>hardware repetition detection should be able to play Rd8 in blitz time,
>and possibly even seeing it winning.
Under tournament time means under 3 minutes. Also the program that
analyzed with Kasparov was DT-1, not DT-2. DT-2 did not yet exist back
in early 1990.