AEGON 97/ 1st round: HIARCS lost

34 views
Skip to first unread message

Enrico

unread,
Apr 16, 1997, 3:00:00 AM4/16/97
to

AEGON 97 / 1. Round

The "monster" Hiarcs lost against van Voorthuijsen (Elo 2264)

1. e4 d6 2. d4 Nf6 3. Nc3 g6 4. Nf3 Bg7 5. Bg5 c6 6. Qd2 0-0 7. Bd3 Bg4 8. 0-0 Bxf3
9. gxf3 Nbd7 10. Rad1 Qc7 11. Rfe1 e5 12. Be2 Rfe8 13. d5 c5 14. Nb5 Qb8 15. c4 a6
16. Nc3 Nh5 17. Bf1 f6 18. Be3 Re7 19. Bh3 g5 20. Be6+ Kh8 21. Ne2 Nf8 22. Bg4 Nf4
23. b4 b6 24. bxc5 bxc5 25. Bxf4 exf4 26. Rb1 Rb7 27. Qc2 Raa7 28. Bf5 Qe8 29. Rb3
Ng6 30. Reb1 Rxb3 31. Qxb3 Ne5 32. Qb8 Ra8 33. Qxd6 Nxf3+ 34. Kh1 Nd2 35. Rb7 Qf8 36.
Qxf8+ Bxf8 37. Rxh7+ Kg8 38. h4 Rb8 39. hxg5 fxg5 40. Ng1 Rb1 41. Kg2 g4 42. e5 Re1
43. Rh5 f3+ 44. Kh1 Nxc4 45. e6 Ne5 46. Rg5+ Kh8 47. Bxg4 c4 48. Kh2 Bd6 49. Kh1 Nd3
50. Rh5+ Kg7 51. Rg5+ Kh6 52. Rh5+ Kg6 53. Bxf3 c3 54. Rh4 c2 55. Rc4 0-1

E.

Martin Borriss

unread,
Apr 17, 1997, 3:00:00 AM4/17/97
to

In article <01bc4aa4$f47791a0$98f2...@194.64.64.1.berlin.snafu.de>,

"Enrico" <Enr...@berlin.snafu.de> writes:
>AEGON 97 / 1. Round
>
>The "monster" Hiarcs lost against van Voorthuijsen (Elo 2264)
>

Now look at this! I thought Hiarcs had no weaknesses ?! Also, there is no
such Elo as 2264...

Martin

--
Martin....@inf.tu-dresden.de

Kevin Miller

unread,
Apr 17, 1997, 3:00:00 AM4/17/97
to

On 17 Apr 1997 13:08:02 +0200, bor...@inf.tu-dresden.de (Martin
Borriss) wrote:

If you thought Hiarcs had no weaknesses, then you're not reading the
reviews. Sure, Hiarcs has had a lot of hype recently, but name me any
IM that is undefeated in a career...:)

Kevin Miller

"Jazz is not dead; it just smells funny" -- Frank Zappa

Martin Borriss

unread,
Apr 18, 1997, 3:00:00 AM4/18/97
to

In article <3356471c...@news.tiac.net>,

kmi...@tiac.net (Kevin Miller) writes:
>On 17 Apr 1997 13:08:02 +0200, bor...@inf.tu-dresden.de (Martin
>Borriss) wrote:
>
>>In article <01bc4aa4$f47791a0$98f2...@194.64.64.1.berlin.snafu.de>,
>> "Enrico" <Enr...@berlin.snafu.de> writes:
>>>AEGON 97 / 1. Round
>>>
>>>The "monster" Hiarcs lost against van Voorthuijsen (Elo 2264)
>>>
>>
>>Now look at this! I thought Hiarcs had no weaknesses ?! Also, there is no
>>such Elo as 2264...
>>
>>Martin

>If you thought Hiarcs had no weaknesses, then you're not reading the


>reviews. Sure, Hiarcs has had a lot of hype recently, but name me any
>IM that is undefeated in a career...:)
>
>Kevin Miller

Oh. I see. Thanks for putting this right! I was relying on the reviews by
Komputer Korner. I highly recommend those, he is a real expert. He also
*clearly* states that no IM would stand a chance versus the monster.

Martin

--
Martin....@inf.tu-dresden.de

Kevin Miller

unread,
Apr 18, 1997, 3:00:00 AM4/18/97
to

On 18 Apr 1997 10:25:10 +0200, bor...@inf.tu-dresden.de (Martin
Borriss) wrote:

>In article <3356471c...@news.tiac.net>,
> kmi...@tiac.net (Kevin Miller) writes:
>>On 17 Apr 1997 13:08:02 +0200, bor...@inf.tu-dresden.de (Martin
>>Borriss) wrote:
>>
>>>In article <01bc4aa4$f47791a0$98f2...@194.64.64.1.berlin.snafu.de>,
>>> "Enrico" <Enr...@berlin.snafu.de> writes:
>>>>AEGON 97 / 1. Round
>>>>
>>>>The "monster" Hiarcs lost against van Voorthuijsen (Elo 2264)
>>>>
>>>
>>>Now look at this! I thought Hiarcs had no weaknesses ?! Also, there is no
>>>such Elo as 2264...
>>>
>>>Martin
>
>>If you thought Hiarcs had no weaknesses, then you're not reading the
>>reviews. Sure, Hiarcs has had a lot of hype recently, but name me any
>>IM that is undefeated in a career...:)
>>
>>Kevin Miller
>
>Oh. I see. Thanks for putting this right! I was relying on the reviews by
>Komputer Korner. I highly recommend those, he is a real expert. He also
>*clearly* states that no IM would stand a chance versus the monster.
>
>Martin
>
>--
>Martin....@inf.tu-dresden.de

I'm pretty sure that I read a review by Enrique Irazoquoi that
mentioned several of the weaknesses of Haircs 6, including king safety
and suicidal tendencies. And I DID mention that the program has
received quite a bit of hype (read hyperbole) lately.
Jeez dude -- take a chill pill...:)

Komputer Korner

unread,
Apr 19, 1997, 3:00:00 AM4/19/97
to

Martin Borriss wrote:
snipped

>
> >If you thought Hiarcs had no weaknesses, then you're not reading the
> >reviews. Sure, Hiarcs has had a lot of hype recently, but name me any
> >IM that is undefeated in a career...:)
> >
> >Kevin Miller
>
> Oh. I see. Thanks for putting this right! I was relying on the reviews by
> Komputer Korner. I highly recommend those, he is a real expert. He also
> *clearly* states that no IM would stand a chance versus the monster.
>
> Martin
>
> --
> Martin....@inf.tu-dresden.de

Why don't you challenge the monster? No one ever said that Hiarcs had no
weaknesses. I said that Hiarcs could defeat any IM in a match. Of course
it could also lose too. The point being that it is good enough to have a
chance of winning a match against an IM. The time controls in Aegon are
not 40/2. They amount to 60/2 and 180/3, because they are Fischer
controls
of 90 30. Hiarcs, because it is a slow processor needs long time
controls
to really shine. Even so it lost a game in Aegon because it misvalues
passed pawns. The trick however is to force a passed pawn on it. Not so
easy to do. There are other computer programs in Aegon that are taking
scalps off of GM's. Your turn is next Mr. Borriss. Deen in his match
with Hiarcs is mounting a comeback and may well tie the match on
Saturday, but Deen will attest that it has been a harrowing experience.
Titled players are scared of the computers progress and why not, it is
only a matter of time before the programs are better than most. Hiarcs
impresses me because of its positional play. It may turn out that the
present version's weaknesses will be known enough that it will have a
hard time in actual games, but that isn't the point. The real point is
that as a tool for improving ones play and as a playing partner it is
ideal. Of course, Rebel 9, M-Chess Pro 7, Genius 6, Nimzo 4, Virtual
Chess 2 and The King 3 also will be monsters. The point is that
computer chess has come of age.
--
Komputer Korner

The inkompetent komputer.

Martin Borriss

unread,
Apr 21, 1997, 3:00:00 AM4/21/97
to

In article <335847...@netcom.ca>,

Komputer Korner <kor...@netcom.ca> writes:
>
>Why don't you challenge the monster? No one ever said that Hiarcs had no
>weaknesses. I said that Hiarcs could defeat any IM in a match. Of course
>it could also lose too.

This is a good one. Of course, if someone writes as much as KK one easily
forgets what one writes. To refresh your memory, you said:

1) Hiarcs 6 *has no* weaknesses.
2) Hiarcs *will* defeat any IM in a match.

In addition, I don't feel obliged to follow your recommendations whom to
challenge and whom not. Feel free to challenge my toaster.

--
Martin....@inf.tu-dresden.de

Komputer Korner

unread,
Apr 22, 1997, 3:00:00 AM4/22/97
to

Martin if you can find where I said that Hiarcs 6 had no weaknesses, I
would appreciate it. The best that I can find after an exhaustive search
on Deja News is the following:

"I repeat my prediction. Any IM,
any time, any where, in a 6 game match 40/2 against Hiarcs 6 will lose
the match."


"Of course you are right about the difference between strength in
IM's but I still say that Hiarcs 6 on a Pentium Pro will defeat
them all in a 6 game match. The reason is that it is not long
enough for the IM/GM to gain an understanding of its weaknesses.
Of course if the IM practices with it, perhaps he will discover
its weakness but based on the latest training matches that I have
seen, it is difficult to imagine that Hiarcs 6 has weaknesses
except maybe a long term sacrifice against its Kingside. Even so
it is a book learner so after one of these losses, it won't play the
same line again. Hiarcs 6 will definitely take some scalps even at
40/2. We just need the sponsors to pay the IMs for these matches."


Notice that in the 1st post, I implied a random IM. Of course if Hiarcs
6
was to play enough matches against an IM, it would lose some. The point
being that I believe that Hiarcs 6 is better than an average IM at 40/2
chess. It has beaten 1 IM at this time control. Are there others who
would like to challenge it? Of course, based on Aegon's results it looks
like there are more computer programs that could defeat IM's in matches.
Bob Hyatt has a point that eventually the titled players find a
weakness in each program that they hammer home to gain superiority over
it. This is becoming harder and harder to do for 2 reasons. 1st, the
possible weaknesses that the titled players find are becoming more and
more obscure so that it takes much greater technique today to exploit
the weaknesses found. 2nd, the programs are being implemented with book
learning and middlegame learning algorithms. These together with the
near
perfect allocation of time algorithms, makes these monsters much harder
to beat in the long run, especially when the titled player gets into his
own time troubles as Deen got into in his match. In any case, once you
find a fatal weakness in your program that you have purchased, it would
be folly to constantly defeat the machine because of this weakness. Just
play your normal game of chess Mr. Borriss, and even you will learn from
programs like Hiarcs 6.

My 2nd post above did not say that Hiarcs 6 had no weaknesses. Please
reread it.

Robert Hyatt

unread,
Apr 23, 1997, 3:00:00 AM4/23/97
to

Komputer Korner (kor...@netcom.ca) wrote:

I disagree. You take *any* program, put it on ICC using an automatic
interface, and you will find humans that will take it's scalp over and
over. Book learning be damned. How will Hiarcs defend against this:

1. e4 a6

it will fall out of book, and probably win. But it will play the same
moves the next time, and the next, and the next, and the IM *will* figure
this out. The position learning I do in Crafty partially helps, but not
nearly enough. The only reason you haven't seen this is that programs like
Hiarcs simply don't automatically play. Their operators are selective in
who they play on the servers, and won't accept 100 games in a row like the
auto-programs do.

Until you've been there, it may be difficult to understand just how serious
this is. Granted it is not something that will happen in a tournament. But
book learning won't help on the servers. It needs more. *much* more. And
it won't be *just* IM's that beat it either. Just ask Bruce, Stanback,
Kittinger, me, and the gnuchess/crafty clone operators...


: perfect allocation of time algorithms, makes these monsters much harder


: to beat in the long run, especially when the titled player gets into his
: own time troubles as Deen got into in his match. In any case, once you
: find a fatal weakness in your program that you have purchased, it would
: be folly to constantly defeat the machine because of this weakness. Just
: play your normal game of chess Mr. Borriss, and even you will learn from
: programs like Hiarcs 6.

Also, don't forget, Deen had two won games and he lost. What would a
Borriss hae done? Maybe one day we'll see. But I would not say that any
program that gets into horrible binds will "never be beaten by an IM"..
we just haven't seen the right IM play it yet... The IM's will pick up
on most of the tactics, once they figure out that they have to be more
careful. And they still have one heck of a lot of positional skill that
none of the programs of today exhibit...

I watched two good commercial programs go down in flames at Aegon today,
in positions that should have been drawn, simply because they didn't
understand and the GM's did...


: My 2nd post above did not say that Hiarcs 6 had no weaknesses. Please

Martin Borriss

unread,
Apr 23, 1997, 3:00:00 AM4/23/97
to

In article <335D3D...@netcom.ca>,

Komputer Korner <kor...@netcom.ca> writes:
>
>Martin if you can find where I said that Hiarcs 6 had no weaknesses, I
>would appreciate it. The best that I can find after an exhaustive search
>on Deja News is the following:
>
>"I repeat my prediction. Any IM,
>any time, any where, in a 6 game match 40/2 against Hiarcs 6 will lose
>the match."
>
>
>"Of course you are right about the difference between strength in
>IM's but I still say that Hiarcs 6 on a Pentium Pro will defeat
>them all in a 6 game match. The reason is that it is not long
>enough for the IM/GM to gain an understanding of its weaknesses.
>Of course if the IM practices with it, perhaps he will discover
>its weakness but based on the latest training matches that I have
>seen, it is difficult to imagine that Hiarcs 6 has weaknesses
>except maybe a long term sacrifice against its Kingside. Even so
>it is a book learner so after one of these losses, it won't play the
>same line again. Hiarcs 6 will definitely take some scalps even at
>40/2. We just need the sponsors to pay the IMs for these matches."
>
>My 2nd post above did not say that Hiarcs 6 had no weaknesses. Please
>reread it.

Ok, this was also the article I was referring to.

Subject: Re: Deep Blue vs Micros
From: Komputer Korner <kor...@netcom.ca>
Date: 1997/02/28
Message-Id: <331674...@netcom.ca>
Newsgroups: rec.games.chess.computer

"it is difficult to imagine that Hiarcs 6 has weaknesses except maybe..."

sounded to me like you were trying to say that Hiarcs has no weaknesses.

As for the first point ("but I still say that Hiarcs 6 on a Pentium Pro will
defeat them all...") I gather that you are perhaps not so sure anymore.

No big deal, I was just hoping that you would not make such huge predictions
any time a new release appears... If I let the old Genius 2 run on current
hardware it is not bad either.

Martin

--
Martin....@inf.tu-dresden.de

Komputer Korner

unread,
Apr 23, 1997, 3:00:00 AM4/23/97
to

Robert Hyatt wrote:

> I disagree. You take *any* program, put it on ICC using an automatic
> interface, and you will find humans that will take it's scalp over and
> over. Book learning be damned. How will Hiarcs defend against this:
>
> 1. e4 a6
>
> it will fall out of book, and probably win. But it will play the same
> moves the next time, and the next, and the next, and the IM *will* figure
> this out. The position learning I do in Crafty partially helps, but not
> nearly enough. The only reason you haven't seen this is that programs like
> Hiarcs simply don't automatically play. Their operators are selective in
> who they play on the servers, and won't accept 100 games in a row like the
> auto-programs do.

snipped


> Also, don't forget, Deen had two won games and he lost. What would a
> Borriss hae done? Maybe one day we'll see. But I would not say that any
> program that gets into horrible binds will "never be beaten by an IM"..
> we just haven't seen the right IM play it yet... The IM's will pick up
> on most of the tactics, once they figure out that they have to be more
> careful. And they still have one heck of a lot of positional skill that
> none of the programs of today exhibit...
>
> I watched two good commercial programs go down in flames at Aegon today,
> in positions that should have been drawn, simply because they didn't
> understand and the GM's did...

5 minute chess is a long way from 40/2 even for computers. Even if
it turns out that in the long run, the human figures out enough
weaknesses in the machine and hammers this home at 40/2, what is the
point? Computers are and always will be a tool to help us play chess
better. Programs like Hiarcs play good positional chess as seen in
the Hiarcs-Hergott match. We should celebrate this instead of
knocking it.

Robert Hyatt

unread,
Apr 23, 1997, 3:00:00 AM4/23/97
to

Komputer Korner (kor...@netcom.ca) wrote:
: Robert Hyatt wrote:

: > I disagree. You take *any* program, put it on ICC using an automatic


: > interface, and you will find humans that will take it's scalp over and
: > over. Book learning be damned. How will Hiarcs defend against this:
: >
: > 1. e4 a6
: >
: > it will fall out of book, and probably win. But it will play the same
: > moves the next time, and the next, and the next, and the IM *will* figure
: > this out. The position learning I do in Crafty partially helps, but not
: > nearly enough. The only reason you haven't seen this is that programs like
: > Hiarcs simply don't automatically play. Their operators are selective in
: > who they play on the servers, and won't accept 100 games in a row like the
: > auto-programs do.

: snipped
: > Also, don't forget, Deen had two won games and he lost. What would a


: > Borriss hae done? Maybe one day we'll see. But I would not say that any
: > program that gets into horrible binds will "never be beaten by an IM"..
: > we just haven't seen the right IM play it yet... The IM's will pick up
: > on most of the tactics, once they figure out that they have to be more
: > careful. And they still have one heck of a lot of positional skill that
: > none of the programs of today exhibit...
: >
: > I watched two good commercial programs go down in flames at Aegon today,
: > in positions that should have been drawn, simply because they didn't
: > understand and the GM's did...

: 5 minute chess is a long way from 40/2 even for computers.

This is a red herring. Log on to chess.net, and see how many *long*
games crafty has played. It (today) has played 12 games of 30
minutes per side... which is not 40/2, but it's a long way from
5 minute chess.

5 minute chess + 5 second increment games can be reasonably long
games as well...


: Even if

: it turns out that in the long run, the human figures out enough
: weaknesses in the machine and hammers this home at 40/2, what is the
: point? Computers are and always will be a tool to help us play chess
: better. Programs like Hiarcs play good positional chess as seen in
: the Hiarcs-Hergott match. We should celebrate this instead of
: knocking it.

I'm not knocking hiarcs at all. I'm defending the IM/GM players
against claims that logic says are simply incorrect. IE, GM players
at Aegon are outplaying the machines pretty handily. IM class players
are also doing well. Lower rated humans are cannon fodder, I agree,
but the micros are simply not IM-class players yet. They are certainly
IM players tactically, or even better, but positionally, they still lack
a long list of concepts. Crafty not excluded either... They can all
whup a GM on occasion. But would you think any micro would really do well
against a single GM in an extended match? I wouldn't...

This is *not* a Hiarcs issue at all... Didn't mean for it to sound like
one. This is a computer chess in general issue... Many have made progress.
But to jump 'em to IM and GM status is quite a stretch... And I'd *still*
like to see Martin play one of the micros. :)


Komputer Korner

unread,
Apr 23, 1997, 3:00:00 AM4/23/97
to

Martin Borriss wrote:

> From: Komputer Korner <kor...@netcom.ca>
> Date: 1997/02/28
> Message-Id: <331674...@netcom.ca>
> Newsgroups: rec.games.chess.computer
>
> "it is difficult to imagine that Hiarcs 6 has weaknesses except maybe..."
>
> sounded to me like you were trying to say that Hiarcs has no weaknesses.
>
> As for the first point ("but I still say that Hiarcs 6 on a Pentium Pro will
> defeat them all...") I gather that you are perhaps not so sure anymore.
>
> No big deal, I was just hoping that you would not make such huge predictions
> any time a new release appears... If I let the old Genius 2 run on current
> hardware it is not bad either.
>
> Martin
>
> --
> Martin....@inf.tu-dresden.de

Well, based on the latest Aegon results, it looks like only the GM's
are safe for now. However what we need is either 40/2 tournaments
or matches with micros/computers vs IM's. These are extremely
interesting
as proved by the latest Hiarcs-Hergott match. Aegon was also
extremely interesting but unfortunately is a faster time control. I will
try to control my enthusiasm in the future, but when Rebel 9, CM 6000 or
Kallisto hit the market, it will be difficult to bridle my admiration
for these beasts! We are truly in exciting times! Congratulations to
Rebel X , Kallisto X and CM X for a top 10 finish at AEGON!!!!!.

Ed Schroder

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

From: Komputer Korner <kor...@netcom.ca>

: Well, based on the latest Aegon results, it looks like only the GM's


: are safe for now. However what we need is either 40/2 tournaments
: or matches with micros/computers vs IM's. These are extremely
: interesting
: as proved by the latest Hiarcs-Hergott match. Aegon was also
: extremely interesting but unfortunately is a faster time control.

: I will try to control my enthusiasm in the future, but when Rebel 9,
: CM 6000 or Kallisto hit the market, it will be difficult to bridle
: my admiration for these beasts!

I have saved your posting Alan as evidence, be veeeeeeeeery careful!
Perhaps I should sent you a reminder at the time?

- Ed Schroder -


: We are truly in exciting times! Congratulations to

Ed Schroder

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

From: hy...@crafty.cis.uab.edu (Robert Hyatt)

[ snip ]

: : 5 minute chess is a long way from 40/2 even for computers.

: This is a red herring. Log on to chess.net, and see how many *long*
: games crafty has played. It (today) has played 12 games of 30
: minutes per side... which is not 40/2, but it's a long way from
: 5 minute chess.

: 5 minute chess + 5 second increment games can be reasonably long
: games as well...

Even if
: : it turns out that in the long run, the human figures out enough
: : weaknesses in the machine and hammers this home at 40/2, what is the
: : point? Computers are and always will be a tool to help us play chess
: : better. Programs like Hiarcs play good positional chess as seen in
: : the Hiarcs-Hergott match. We should celebrate this instead of
: : knocking it.

: I'm not knocking hiarcs at all. I'm defending the IM/GM players
: against claims that logic says are simply incorrect. IE, GM players
: at Aegon are outplaying the machines pretty handily. IM class players
: are also doing well. Lower rated humans are cannon fodder, I agree,
: but the micros are simply not IM-class players yet. They are certainly
: IM players tactically, or even better, but positionally, they still lack
: a long list of concepts. Crafty not excluded either... They can all
: whup a GM on occasion. But would you think any micro would really do
: well against a single GM in an extended match? I wouldn't...

I agree on the GM part.
I disagree on the IM part.

Todays top Pc programs (on a fast Pc) do play on IM level. There is
enough evidence for that.
- Hiarcs-Hergott
- Long list of Fritz results (posting of Enrique)
- Results Aegon tournaments last 2-3 years.

For Rebel...
Aegon 94: elo 2470 (Rebel6)
Aegon 95: elo 2473 (Rebel7)
Aegon 96: elo 2530 (Rebel8)
Aegon 97: elo 2619 (TV97.1)

On Aegon 97 Rebel played:
- 2 x GM (50%)
- 4 x IM (87.5%)

Bob, better accept the facts... :)

: This is *not* a Hiarcs issue at all... Didn't mean for it to sound

: like one. This is a computer chess in general issue... Many have
: made progress. But to jump 'em to IM and GM status is quite a stretch...
: And I'd *still* like to see Martin play one of the micros. :)

Ok, here is some interesting data. Beginning June Rebel in Italy will
play a match against GM Arthur Yusupov (elo 2660).

The schedule looks as follows:
1st day 10 blitz game 5 minutes
2nd day 5 semi blitz 15 minutes
3rd day 2 active chess 30 minutes (Black and white)
4th day 1 1hour game (white for Rebel)

Official announcement will follow but maybe I can persuade Bob for a
prediction?

- Ed Schroder -

Robert Hyatt

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

Ed Schroder (rebc...@xs4all.nl) wrote:
: From: hy...@crafty.cis.uab.edu (Robert Hyatt)

: [ snip ]

: - Ed Schroder -

I'm game. :)

1st day 6-4 rebel, assuming that rebel plays "normal" chess, where the
operator has to move pieces on the board and press the clock. If the
human has to play using the computer and mouse, then I'd change this to
8-1 or 9-1 because the machine is at a time advantage then.

2nd day 3-2 rebel. Again assuming the game is played on a real board
as it should be, and not on the pc running rebel.

3rd day .5-1.5

4th day either .5-.5 or 0-1

all games are short enough that the computer should win some on time.
However, at blitz, it should be a Masscre. I've had the opportunity to
play a top-3 US GM quite a few games. At blitz, he has not just trouble,
but *big* trouble playing Crafty. At more reasonable time controls, his
skill becomes more obvious. This GM was at Aegon and did quite well...
Probably using Crafty as a practice vehicle no doubt... sorry guys. :)

Herbert Groot Jebbink

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

Hello,

Bob> the micros are simply not IM-class players yet.
Ed> I disagree on the IM part.

Bob, Ed, you are both right !!!

Micros do not play like IM players.

Not any IM would play like Rebel did in the moves 70 - 120 in the
game against S. Polgar

Micros do get results like IM players.

A 75 % score against the opponements that Rebel played at Aegon is a
very good IM result.

So, now you are both right, do what you are best in.
Write chess programs ! We are wating for Rebel 9 and Crafty 11.69 !

Greetings, Herbert


---
The Trans-Siberian Railroad Page, http://www.xs4all.nl/~hgj/

Komputer Korner

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

Robert Hyatt wrote:

> This is *not* a Hiarcs issue at all... Didn't mean for it to sound like
> one. This is a computer chess in general issue... Many have made progress.
> But to jump 'em to IM and GM status is quite a stretch... And I'd *still*
> like to see Martin play one of the micros. :)

No one said that micros are ready for GM's, but looking at the results
at Aegon, it seems that most IM's are in danger. We just need some more
matches to prove it. Rebel X, Kallisto X, CM X, and Hiarcs 6 to name a
few
do not need to fear IM's any more.

Robert Hyatt

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

Komputer Korner (kor...@netcom.ca) wrote:

: Robert Hyatt wrote:

That's where you are *dead* wrong. put any program of your choice on a
server. Agree to play any number of games against a good IM, at any
time control he agrees, and watch what happens. You *have* to do the
following:

1. book learning. Without it, you get killed. period. Else you
have to do manual book learning and fix every hole the IM will find,
and he'll keep you plenty busy.

2. handle the a3 type of case, where the IM plays a move to take you
out of book, and then because the program will repeat the same moves
over and over, once out of book, he'll find a thread to work on and
will end up at >3200, with the computer at <1000... :) Ask *any*
of us that play automatically... Kittinger... Ferret... Crafty...
GnuchessX... Rajah... and so forth.

To not fear an IM is to not fear a truly dangerous chess player. To
take one lightly, because of one match, is not only dangerous, but
foolish. To write IM's off as a waste of time is even worse... I don't
want to see automatic commercial programs, because that's my biggest
hole-card at present... getting to play so many games against high-
quality opponents. But in doing so, it is apparent just how inferior a
computer is to these guys. Tactically? No. Positionally? in many cases,
no. But in too many cases, tragically worse. It's improving. But it is
*not* solved yet.

I'd be willing to play Martin Borriss in a blitz match and expect to win.
But as the time control extends, my probability for winning decreases. It
doesn't take 40/2 to make me expect to lose against him at present. I'm
working to solve this, but it's a long-term problem/solution, not a quick
fix. He's dangerous, and he's not the only one. There are *many*...


Robert Hyatt

unread,
Apr 24, 1997, 3:00:00 AM4/24/97
to

Herbert Groot Jebbink (h...@xs4all.nl) wrote:
: Hello,

: Bob> the micros are simply not IM-class players yet.
: Ed> I disagree on the IM part.

: Bob, Ed, you are both right !!!

: Micros do not play like IM players.

: Not any IM would play like Rebel did in the moves 70 - 120 in the
: game against S. Polgar

: Micros do get results like IM players.

: A 75 % score against the opponements that Rebel played at Aegon is a
: very good IM result.

: So, now you are both right, do what you are best in.
: Write chess programs ! We are wating for Rebel 9 and Crafty 11.69 !

hmmm... we need to talk about that crafty version. :) sounds somewhat
"odd"... :)


: Greetings, Herbert

Ronald de Man

unread,
Apr 25, 1997, 3:00:00 AM4/25/97
to

hy...@crafty.cis.uab.edu (Robert Hyatt) writes:

>2. handle the a3 type of case, where the IM plays a move to take you
>out of book, and then because the program will repeat the same moves
>over and over, once out of book, he'll find a thread to work on and
>will end up at >3200, with the computer at <1000... :) Ask *any*
>of us that play automatically... Kittinger... Ferret... Crafty...
>GnuchessX... Rajah... and so forth.

I guess a solution to this problem would be randomizing the first moves
out of book. This solves the problem for my program (TrojanKnight,
playing automatically on fics).

Ronald de Man


William B. Wright

unread,
Apr 25, 1997, 3:00:00 AM4/25/97
to

> Robert Hyatt wrote:
>
> > This is *not* a Hiarcs issue at all... Didn't mean for it to sound like
> > one. This is a computer chess in general issue... Many have made progress.
> > But to jump 'em to IM and GM status is quite a stretch... And I'd *still*
> > like to see Martin play one of the micros. :)
>
> No one said that micros are ready for GM's, but looking at the results
> at Aegon, it seems that most IM's are in danger. We just need some more
> matches to prove it. Rebel X, Kallisto X, CM X, and Hiarcs 6 to name a
> few
> do not need to fear IM's any more.

> --
> Komputer Korner
>
> The inkompetent komputer.

I didn't realize anyone had written a program that could feel emotion (be
afraid) :)

--


Tom C. Kerrigan

unread,
Apr 25, 1997, 3:00:00 AM4/25/97
to

Komputer Korner (kor...@netcom.ca) wrote:

> weaknesses in the machine and hammers this home at 40/2, what is the
> point? Computers are and always will be a tool to help us play chess

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


> better. Programs like Hiarcs play good positional chess as seen in

^^^^^^^

This sentence doesn't seem to be connected to anything, and the tone is
contradictory, so I infer that the only purpose of chess programs is to
improve our play. Stupid. Personally, I enjoy writing my program and don't
give a first thought to improving my own play. I can name a dozen uses of
chess programs (well, at least a handful) that don't have anything to do
with improving play.

Aren't you some sort of professional writer? Computer chess reviews or
something? I'm glad I haven't read any. I get the feeling they would just
give me headaches.

Cheers,
Tom

mclane

unread,
Apr 26, 1997, 3:00:00 AM4/26/97
to

Komputer Korner <kor...@netcom.ca> wrote:
>"I repeat my prediction. Any IM,
>any time, any where, in a 6 game match 40/2 against Hiarcs 6 will lose
>the match."

This is nonsense. I can show you IM's who beat Hiarcs6.

Maybe you should ask IM Bernd Kohlweyer or IM Markus Schaefer, if you
don't believe it.

mclane

unread,
Apr 26, 1997, 3:00:00 AM4/26/97
to

Ed Schroder <rebc...@xs4all.nl> wrote:


>: We are truly in exciting times! Congratulations to
>: Rebel X , Kallisto X and CM X for a top 10 finish at AEGON!!!!!.

Please do not forget that some participants start with ONE program
meanwhile other companies start with 4 clones of one engine.

IMO this increases the chances in a suisse-tournament.

>: --
>: Komputer Korner

>: The inkompetent komputer.

Komputer Korner

unread,
Apr 28, 1997, 3:00:00 AM4/28/97
to

mclane wrote:
>
> Komputer Korner <kor...@netcom.ca> wrote:
> >"I repeat my prediction. Any IM,
> >any time, any where, in a 6 game match 40/2 against Hiarcs 6 will lose
> >the match."
>
> This is nonsense. I can show you IM's who beat Hiarcs6.
>
> Maybe you should ask IM Bernd Kohlweyer or IM Markus Schaefer, if you
> don't believe it.
>
> >Komputer Korner
>
> >The inkompetent komputer.

Did they play a 40/2 match of at least 6 games with the program
running on a Pentium 200 MMX or Pentium Pro?

Serge Desmarais

unread,
May 1, 1997, 3:00:00 AM5/1/97
to

mclane wrote:
>
> Komputer Korner <kor...@netcom.ca> wrote:
> >"I repeat my prediction. Any IM,
> >any time, any where, in a 6 game match 40/2 against Hiarcs 6 will lose
> >the match."
> Nope. Some IM could/would lose and some would win!

Komputer Korner

unread,
May 3, 1997, 3:00:00 AM5/3/97
to

My point is that the computers never play below their level. They are
100% consistent except for the odd hash collision. If an IM plays
above his level he will defeat Hiarcs 6 and Rebel X, but if he does
play his usual game, the IM will lose. That was all I was trying to say.
There are some IM's who are actually of GM strength. I wasn't counting
those. I will try to be less enthusiastic about these programs but
folks we are truly in exciting times. In 2 years time the level of
play should be on average of GM strngth assuming Intel can give us more
speed. I am the first one and only one to predict this, but based
on the last couple of years, it is not so wild a prediction. We will
see.

Reply all
Reply to author
Forward
0 new messages