Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: DB will never play with REBEL, they simple are afraid no to do well

Author: Ratko V Tomic

Date: 20:40:28 10/13/99

Go up one level in this thread


> I have read several of your post where you say that you can positionally
> outplay several programs, At 2100 uscf this seems unlikely,

Positionally outplaying doesn't mean winning a game. While my USCF rating is
quite old (from 1982-83 while I was a graduate student in theoretical physics at
Brown University) I had to quit playing (and working on chess all together) due
to lack of time, possibly before my rating in USA maxed out (my younger brother
had achieved master rating, and still plays in his local team competitions (in
Yugoslavia) and my younger sister used to be a strong player as well. Before
coming to school in USA, I played often with master candidates (which would be
near USCF master) in the clubs over there and didn't feel they had that much of
an edge. One neighbour (Branko Damjanovic) was  a junior champion of Yugoslavia,
and IM at the time I played friendly games with him (he later became a GM and
one of the leading players there). Now that was a strong player and, frankly,
none of the programs comes close in the strength I can perceive. He could, of
course, see tactically everything I could see, plus much that I hadn't a clue
about until it would strike, and also positionally he would outplay me
completely in every game, no matter how much beer he might have consumed that
evening. With programs, they generally see tactically more than I do, but not
always, depends on type of tactics, and strategically, provided I look up the
book moves (that programs display, so I don't blow the opening, I haven't
studied openings last 15 years), I judge that most of the time I come out for
good part of the middle game positionally ahead, until some tactical shot turns
the situation around. With Hiarcs and somewhat with Rebel 10b, there is more
equal positional struggle (and Hiarcs can completely outplay me positionally in
maybe 20-30 percent of games), but Fritz 5.32, Junior 5 and CM6k, they're weak
positionally in my judgment (even though I have a strong negative score against
all of them, I maybe win 2 points out of 10 games; the time settings are usually
15-30 seconds/move on average, i.e. without sharp time limit; my score is worse
against Rebel and Hiarcs, better against the others).

> although you
> may be a expert at anti-computer play if there is any real definition
> of this term ,


I wouldn't say that I am an anti-computer "expert", although I have played
almost exlusively against computers since mid 1980s. What I am comparing is not
my expertise against programs, but the perceieved strength I recall from playing
against really strong human players and against the nominally stronger programs.
My perception is that the very strong players were stronger than me in every
aspect of the game, while programs are stronger only in some aspects. For me the
chain is as strong as its weakest link, so the IM strength human player rated at
2450 is for me stronger than the program showing 2600 (self-)rating.


> I would be interested in playing you a short match on
> ICC running chessmaster 6000, I would love for you to demonstrate to me
> your computer chess prowness!!!

I had CM6k on my computer (for over a year), although I probably played fewer
than couple dozen games against it in all this time. While my kids (the oldest 3
kids, ages 7, 9, 12; the younger two are ages 1 and 3, too young to be taught
chess) play with it often and love it. I find it one of more boring programs
(with the oddest, most computer-like moves) out of about half a dozen of most
recent ones. If its UI weren't as annoying I might have played it a bit more. In
any case, what exactly would I have to prove in such match? That I can win 2
points out of 10? Or outplay it positionally (by whose judgment) in at least 5
out of 10 games? Why would I bother? I have roughly 30 wins and 60 draws, as
saved games, mostly against Fritz & Hiarcs (since they save it automatically, I
wouldn't have bothered otherwise). That's from about the last month and half I
had the latest programs (with the first week of AutSave.cbh erased during one of
the fresh re-installs of Fritz when it started playing the same opening line
over and over). I also delete all "uninteresting" games; all losses to the
programs are on the cheap shots and generally "uninteresting," of course:). I
can email you that saved stuff if you wish to see, but going personally into
public competitions, sorry, but not interested at this time (twenty years ago, I
might have taken it up). Maybe my program some day will do that, if I can get
something similar to the Botvinnik's approach to work (which still hasn't gone
beyond separate evaluation functions).




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.