Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Tiger wins 12 games in a row, WOW!

Author: Christophe Theron

Date: 21:00:57 11/20/99

Go up one level in this thread


On November 20, 1999 at 18:47:28, Tina Long wrote:

>On November 19, 1999 at 04:08:22, Jouni Uski wrote:
>
>>From SSDF:
>>
>>1½01110011½1½111111111111½1½½10½11½11111 32.0/40  Tiger 12 amd-450
>>0½10001100½0½000000000000½0½½01½00½00000  8.0/40  Nimzo 99 200mmx
>>
>>12 wins in row against 2567 program! I wonder if Garry Kasparov can repeat this?
>>Actually I can bet $100 that he can't.
>>
>>Jouni
>
>I think Kasparov could easily reach the 32.0/8.0 (do you have to extract repeat
>games :} ), but the 12/0 is a bit tricky.
>
>Christopher,
>I think we accept that Tiger is very very very good in Computer v Computer.
>Are there any plans to formally test it against Humans, in the style of the
>Rebel Challenge?


We have no such plans sofar, but it does not mean we are not going to do it.



>Ed said on his pages as much as a year ago that Tiger was better at Com. v Com.
>than Rebel.  I felt the implication was that Tiger may not be so strong in Human
>games.


But this is not what is written, I think you'll agree. It just happens that
Tiger is very strong against computers, without having specifically been
designed to do so.

I don't have many tournamanent games against humans.

Tiger 12.0 won a simultaneous game against Joel Lautier (2638 FIDE) recently.

Apart from that the rest is maybe not very significant:

Tiger 11.8 won a tournament here in April. The players were not very strong, but
Tiger did a 2200 elo performance against humans on a 386sx 20MHz (this point is
important I think).

Also last year Tiger 11.5 played a 21 games match against a player that is rated
about 2000 elo FIDE. The computer used was a K5-100MHz (equivalent to P200), and
time controls were game in 30. Tiger 11.5 won 18.5-1.5.



>I feel the opposite about Hiarcs (but it is only a feeling), that it would
>outperform some top programs in matches against humans, where it (more or less)
>equals them in com v com.
>
>What Human testing have done, & do you intend.
>
>Also,
>Rebel-Tiger is a stand alone product.  Will it also be an engine for
>Rebel-Windows?


I don't know, but I guess it will be. Rebel for Windows and Rebel-Tiger will use
the same user interface, it will be easy to add the Tiger engine to Rebel for
Windows.



>I think I've said it before, but Congratulations on the culmination of your many
>years of work.  I hope you get lots of money (including a small lump of mine)
>from this to make life easy for you.


I don't get a lot of money, so your personal contribution will be welcome. Just
to give you an idea, I have two other jobs at this time.

But if people like you contribute, and if the number is enough, I'll maybe be
able to drop one of my two additional jobs and spend more time on Chess Tiger.

So far, what helps me the most are your kind words. I want to thank you for
this.



    Christophe



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.