Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 12th WCCC, Bar-Ilan University: why not to go.

Author: Omid David Tabibi

Date: 13:08:23 12/11/03

Go up one level in this thread


On December 11, 2003 at 06:20:48, Amir Ban wrote:

>On December 10, 2003 at 19:04:42, Robert Hyatt wrote:
>
>>On December 10, 2003 at 18:34:54, Amir Ban wrote:
>>
>>>On December 10, 2003 at 18:12:39, Robert Hyatt wrote:
>>>
>>>>On December 10, 2003 at 17:27:04, Amir Ban wrote:
>>>>
>>>>>On December 10, 2003 at 15:29:44, Robert Hyatt wrote:
>>>>>
>>>>>>On December 10, 2003 at 13:33:46, Matthew Hull wrote:
>>>>>>
>>>>>>>On December 10, 2003 at 13:19:49, Gian-Carlo Pascutto wrote:
>>>>>>>
>>>>>>>>On December 10, 2003 at 10:48:21, Matthew Hull wrote:
>>>>>>>>
>>>>>>>>>You can still do this with a local FICS server, authors present.
>>>>>>>>
>>>>>>>>But what's the advantage of autoplay then?
>>>>>>>
>>>>>>>- No bungling of repetition draws.
>>>>>>>- No issues of operator errors and/or fatigue affecting the games.
>>>>>>>- No time lost in move transmission.
>>>>>>>- Programmers/operators are more free to move about and relax without fear of
>>>>>>>missing a move and losing time.
>>>>>>>- Programs make all the decisions about draw offers, draw acceptance, resigning,
>>>>>>>etc.  The human element is removed and it's true CC.
>>>>>>>- More rounds can be played in the time specified, reducing costs to
>>>>>>>participants and organizers alike.
>>>>>>>
>>>>>>>
>>>>>>>It's all advantages and no disadvantages.
>>>>>>>
>>>>>>>Matt
>>>>>>>
>>>>>>
>>>>>>I can't even believe that question was asked, much less answered. It
>>>>>>seems intuitively obvious to the casual observer, for those that play
>>>>>>hundreds of thousands of games on chess servers...
>>>>>>
>>>>>
>>>>>Well, perhaps I don't qualify as a casual observer, but you and Matt are making
>>>>>no case at all, and shouting "it's so obvious !" doesn't make your case.
>>>>>
>>>>>Perhaps the solution would be for you to observe a WCCC and see what little
>>>>>difference this would make.
>>>>>
>>>>>Amir
>>>>
>>>>I _have_ observed them.  I have seen operator interference.  I have seen games
>>>>that had to be backed up.  I have seen time lost because an operator was away
>>>>from the board for restroom break or whatever.
>>>>
>>>
>>>Operators lose time all the time, going to the bathroom, have coffee, wandering
>>>to other tables, or even reading CCC. Everybody allows 10 minutes or so for lost
>>>time, or adjust the clock later, and it makes so little difference to the game
>>>that I don't know why it should even be discussed.
>>>
>>>I've only been attending since 1995, and I don't remember any problem caused by
>>>operators.
>>>
>>>
>>>>How could you _not_ think that an automatic event would be better?  IE on the
>>>>CCT events we can chat in the games via kibitz.  We can chat in channels to a
>>>>wider audience.  This would happen locally where we could gather around
>>>>interesting games and discuss them without worrying about missing something
>>>>on our own games.  We could use automatic pairings, so that round 2 fires up
>>>>right after round 1 ends, compressing the event and allowing even more rounds
>>>>(or fewer days).  Etc.  I don't see _any_ down-side to getting the humans
>>>>out of the loop, since the computers are supposedly playing the games anyway.
>>>>
>>>
>>>Sorry, I don't understand this at all.
>>>
>>>Pairings are done at the end of a round, by a pairing program. Except for the
>>>handshake, how do operators delay the next round ?
>>
>>I can't ever recall an event where the starting time was announced as 7pm
>>and at 7pm all N games started.  With automatic software this would not
>>happen.  Even more important, start times can be moved up when all games
>>finish early.  But not if we depend on operators as they can go to eat,
>>knowing the next round doesn't start until 7pm.  When it might actually start
>>at 6pm if everyone were ready.  Even more importantly, NO operator time would
>>be needed or lost.  It would be computer vs computer, with no time lost in
>>relaying moves, whatsoever.
>>
>
>Rounds start at the scheduled times, not because the operators go shopping. If
>the TD wants to compress schedules, he just says this to the operators, and it
>happens.
>
>It's not the operators, but the spectators and the press who want it to start on
>scheduled time. Turn on to ICC to view the game of the century on 1100 EST.
>Oops! It's already over.
>
>I have a solution to this: Why not make the *spectators* automatic too :-)
>Scrappy(C) can watch the game for you, and you'll be free to do more classes.

Let's take this idea to its logical conclusion: why not make the *programmers*
automatic too? Actually, why not automate everything? You ask what the humans
will do then; no worries, Terminator(C) will take care of that problem :)



>
>This suggestion mean unmanned, unwatched, un-TD'ed events (i.e. SSDF). Do you
>think I want to fly to Austria, have breakfast at the hotel, come to the
>tournament hall, and find out that the championship ended overnight ?
>
>
>>
>>
>>>
>>>
>>>>As far as observing goes, I observed at the first WCCC event.  I participated
>>>>in the next one in 1977.  I missed 80 as it was out of the US and too expensive
>>>>to attend.  I played in 1983, 1986 and 1989.  That's six I personally attended.
>>>>Not to mention 20 ACM events that had just as many entries and just as many
>>>>problems to handle.  The last CCT event went smoothly and quietly...
>>>
>>>Well, either things have changed or your memory needs refreshing.
>>
>>I have seen bad decisions made.  IE with an automatic interface, Fritz
>>would be the current WCCC champion.  The draw claim would have been
>>handled correctly, the correct result recorded, and a 500-post thread
>>would never have happened...
>>
>
>It should have been a draw anyway. But just wait to see the length of the thread
>when the championship is decided by two programs misunderstanding each other
>while the TD was out shopping.
>
>Amir
>
>
>>
>>>
>>>I'm not saying automatic is bad, though it would take some time to stabilize:
>>>How to let the TD have the control he needs and what to do when the damn things
>>>don't talk to each other, etc. It will work in the end, but since there's no
>>>clear benefit, why bother ?
>>
>>It is just the right way to do things.  Humans influence the games,
>>intentionally and unintentionally.  With no human hands in the mix, the
>>games would be computer vs computer as intended.
>>
>>If a program won't play, it loses.  The author fixes it before the next
>>round.  But with plenty of free chess servers around, there is really no
>>excuse in showing up with a program that won't connect and play normally.
>>
>>>
>>>Amir



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.