Interview
DamItalia
Newsletter (Italie)

In esclusiva per Damitalia
http://www10.brinkster.com/damaitalia/default.asp &
http://it.groups.yahoo.com/group/damitalia/ un'intervista a Nicolas
Guibert sviluppatore di Buggy, il software più forte di dama
internazionale che tenterà di battere il GMI Samb, 8° posto nel
ranking mondiale. Un ringraziamento a Nicolas e al suo staff e un
invito a tutti i damisti ad assistere alle partite live via
internet !
R. Bubbi

Who "builds" Buggy ?

Nicolas Guibert, (that is myself) is the developper of Buggy. I was
France champion in 1999.

I started the project in September 1998. While preparing the first
match Samb-Buggy, I asked GMI Maxime Kouame who lives close to my
home if he wanted to take part to the adventure and help me with the
opening book. He gladly said yes. In 2002, two other very strong
players, Souleymane Keita who is now ranked World #98 (Volmac 2262),
and MI Wieger Wesselink (now World #22 !!) accepted also to become
members of the team. They have a copy of Buggy at home, test the
opening book, play games, give me feedback about what is wrong in
Buggy. They help me a lot in spotting problems.

What are the differences between this match and the first one ?

Well, basically, N'Diaga is even stronger than in 2001. He is now
World #8 whereas he was World #11 in 2001. He is closer and closer to
the World Champion. And who knows ? Maybe some day ?

Buggy is also supposed to be stronger. It won with a large margin the
last computer competition. But beware, competing against humans is
very different from competing against other computers. Humans are
good at exploiting small weaknesses. Computer opponents can hardly do
that.

Remember also that Buggy was only 3 years old for the first match. I
hope it has grown up a bit since then.

What have you learned, what experiences did you get from the first
match, particularly regarding the challenge Man-Machine ?

The first match was really intense. I think that the spectators
enjoyed it a lot. The tie-break of the second set was particularly
dramatic.

In the first match, N'Diaga Samb managed to find 2 or 3 weaknesses of
Buggy. The first and main one was the pressure on the piece 27 (for
white). This came back in the match at least 3 times. The Grand-
Master also exploited some lack of knowledge of Buggy. It is
difficult to explain there here. But, we hope the problems he showed
are now gone.

As I have said above, Man-Machine matches are different from Machine-
Machine competitions. Moreover, a match is something very different
from a normal round robin tournament (or swiss system). You have to
cope with your opponent's particularities. This is not something I
have ever done myself. I never had to play a match against anyone.
This is very special. I regret that little interest is shown for
other matches in our game. Would not it be nice to know the feelings
of Georgiev before he meets Tchizov or even after the match ? I'd
like to read about that.

Beyond that, we have learned that Man-Machine matches are extremely
good for the image of the game. For the first Match, there was an
article on the frontpage of 'Le Monde'. Do you imagine an article
about our game on the frontpage of 'La Stampa' ?

How will the event be advertized ?

We do our best to help journalists. A media pack is available on the
website www.buggy-online.com, a press release also that will be sent
to the press agencies. Banners (in English, French, and Italian) are
available for the webmasters who want to help.

The match will be commented in Cannes by Gilbert Charles, who is very
famous in France for being the organizer of the 'Blitz de Wattrelos',
and also for his big work as a draughts teacher at school.

Lastly, and most importantly, the games will be broadcast live on the
website www.buggy-online.com. I hope many Italian players will come
and see the live interface. There will be a chat accompanying the
games and this promises to be thrilling. For the 1st match, the World
Champion Alexeï Tchizov himself made some comments about the moves.
That was great !

When will there be the first woman-machine match ?

That's indeed something interesting also. The public would like that.

On the one hand, there is of course the Artificial Intelligence
interest (on the machine side of course !). On the other hand, there
is also the possible debate on the strength of women in this game,
and more generally in any game of that kind. Scientists may have
interesting things to say about the difference between men and women
as far as the brain is shaped.

Unfortunately, I am not sure that a fair match between a computer and
a woman would be possible. I fear it is not possible anymore to
resist the computer if you are not among the first 20-30 in the
world. I am not sure that the best woman today is in the top 100. Bad
luck !

What differences between a Draughts program and a Chess program ?

The main algorithm is similar. That is 1) a search algorithm, 2) a
generator of moves, 3) an evaluation function.

However, programming a computer that plays a game is about
optimization. You have to tune everything. Of course you do not tune
a rocket as you tune a car. That's the same for a Chess or a Draughts
program. You do not represent the pieces the same way, you do not
generate the moves the same way (there are no multiple captures in
Chess for example, which makes the draughts move generator more
complex), you do not have the same endgame database, nor the same
opening book, you do not have mates in Draughts, you do not have
forced captures in Chess, etc.

So basically, the ideas are the same, but the way they are
implemented differs significantly from one game to the other.