Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: About tie-breaks and RUFFIAN in special

Author: Daniel Clausen

Date: 07:57:39 01/20/03

Go up one level in this thread


On January 20, 2003 at 10:21:29, Rolf Tueschen wrote:

>On January 20, 2003 at 09:47:07, Daniel Clausen wrote:
>
>>On January 20, 2003 at 09:37:51, Rolf Tueschen wrote:
>>
>>[snip]
>>
>>>That shouldn't be the motto of CCC for sure. But you are right with the
>>>conditions, I didn't know that until just a few minutes ago when I read it and I
>>>already excused myself for that one. I fear that Bob or Bruce didn't even think
>>>that they could be that strong and so they didn't spend too much thought on that
>>>matter, the tie-break. BTW we had the same procedure in Masstricht. I am sure
>>>that SHREDDER would be Wch and not JUNIOR if it wouldn't have been in Blitz at
>>>the end.
>>>
>>>Rolf Tueschen
>>
>>How many tournaments (not necessarily chess) do you know where ties for the
>>first place are handled much different? The only thing I know is the FA soccer
>>cup (I think) in England, where they replay a whole game when the first game is
>>drawn. (not sure what they do if the 2nd game is draw too though - I'm sure
>>Andrew (CTID) knows =)
>>
>>In most other soccer tourneys they have overtime and afterwards penalties. Or
>>sudden death in ice hockey.
>>
>>The main reason for this is to:
>>
>>-make sure the game/tournament ends at that day (otherwise the commitment is too
>>big)
>>-entertain the spectators (I think the tie-break was a blast)
>>
>>
>>If I'd be in charge of the rules, I would skip the tie-break (although it's fun
>>to watch! :) and just call the engines with the most points the winners of the
>>tournament.
>>
>>The reason for this is that - although you can tell people 100 times that this
>>blitz tiebreak is a lot of luck - people will forget about it and the only thing
>>remember will be that Ruffian won the CCT5. (I have nothing against Ruffian!
>>It's just that Yace and Crafty had the same #points :)
>>
>>Sargon (aka ruffian_groupie ;)
>
>Thanks for your message; let me just add two things:
>
>- soccer is different because you have real human beings and a complete staff on
>the run, a tie could never be solved irectly afterwards; of course sudden death
>is luck but the logic is mainly that BEFORE you had time enough to win; chess is
>different; compared with soccer, Crafty and Yace would have a clear advantage
>over Ruffian, so that is what makes it a bit sad to see; perhaps your comparison
>with soccer would be better if we assumed that the decision would then played
>between two teams a 5 men... :)

I agree that the comparison between soccer and chess doesn't hold everywhere (I
would be crazy to suggest that on a chess (or a soccer) forum ;) Basically, what
I meant was that whatever solution you come up, there are always drawbacks as
well.

If you skip the tie-break, then you don't have one winner. If you play "45 10"
until you have a winner (which sounds fairest) then you don't know how long the
tournament takes. If you play blitz instead it's similar to penalties in soccer.
(it's just a different game) Disadvantages everywhere.

In order to choose one solution, we should first agree what is more important
for us: to have the "real winner" of a "45 10" tournament. (which rules out a
blitz tiebreak) Or to have lots of spectators. (which rules out Uri's suggestion
of playing one game a month) We simply can't have everything.


>- I want to add this one that is completely overseen in our case: you said that
>it should have been done in a day; what day are you talking about?

I didn't mean the "end at that day" literally. (bad formulation, sorry) I meant
that:

(a) people (programmers as well as spectators) have an idea when the tournament
is finished approximately
(b) people know before the tournament, what days they have to reserve in order
to participate (if the tie-break with "45 10" games would be next weekend,
everyone in the tournament would have to reserve that weekend too, because
everyone can be a potential tie-breaker)

Hope this is clearer now. :)


>Hey, I was too stupid to continue to
>observe because I didn't know the rule. And I thoought that it was either
>finished or done on another day... So I missed the whole high-light - sh___!

You mean you missed the tie-break games? Ack! That's unfortunate! :(


>Anyway, the whole event was fantastic.

Agreed 100%! (although I only could see the last 2 rounds)


>Let me end with a word about Ruffian since you are a real fan of.

Uh, let's hope we don't start a flame war now. :)


>The whole excitement about that program is somewhat hypocrisy. Let me explain.
>For certain nobody on the Earth could program awinner prog just out of nothing.
>Without tuning against known progs etc. Also as a programmer you can't enter
>into CC and start last week and next month you have a killer. That is
>impossible. So - it is clear that if you want to surprise the world that you
>must keep secrecy over months and years. I don't see the value of such an
>attitude.

Well, AFAIK he never claimed that he made this engine in a few weeks. In this
interview he gave he mentioned when he started with the engine. (I don't
remember the date/year now though)

I don't see the "must keep secrecy over month and years" as you see it. People
are different. He just chose to code "at home" (and surely read about CC on the
internet) and work on his engine w/o participating in this community. There's
enough information available on the net to produce an engine which comes close
to Craftys strength. (I'm not implying it's easy though)


>That is completely against all known aspects of modern openess,
>democracy and science.

I would also prefer if he would for example participate on this board, but he
chooses not to. And he even gave a reason for his decision. (boards like these
tend to be repetitive etc) And you can't deny that. :) [just think about
DB-posts :)]

And maybe, he will write a paper about forward-pruning one day! You never know!
:)


>You must accept that there is no hidden mysth that would
>allow you to invent such a good prog out of nothing.

I don't. I'm a (chess-) programmer myself and have a pretty good idea what it
takes to produce such a strong engine.


>So then you hide yourself
>for years and then make even a mysth about your name at the beginning.

Not sure what you mean by that. If I don't shout out loud on a CC forum, I'm
hiding? :) Also, I didn't know that he tried to not reveal his name at the
beginning. But I'm not always reading this board, so it's very well I missed
some posts. ;)

I'm not a fan of Ruffian (although I even used the word groupie, hehe). I just
think it was very refreshing to see such a strong new engine in CC. That and the
fact that there's still a slim hope that the engine will make it to the Linux
(or even OSX) platform adds to it, I confess. (I'm sorry if I couldn't care less
about Fritz 8.0.0.8.3.56.3.3.45 beta - it just doesn't support my platform(s))

Sargon



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.