Kanazawa news?

Currently, the computer olympiad is running in

Kanazawa, Japan. Does someone know on which days

the Havannah competition (with 5 participants)

and the Hex competition (also 5 bots) will

take place? What are the board sizes?

Thanks in advance, Ingo.

Hi Ingo,

Here are fresh news from Kanazawa:

Hex tournament ended about 30 minutes ago. MoHex from UoA won, Wolve from UoA was second and MimHex from University of Warsaw was third. Yopt finished last, there was fifth programmed inscribed but it forfeited because visa issues.

Havannah is still running but should end in less than two hours.

Thanks Halladba, I'm eagerly anticipating the results, and so are many others (I hope :)

havannah tournament results:

1. castro

2. wanderer

3. lajkonik

4. shakti

5. pzn

board sizes for havannah:

5 & 8

Dear Przmek,

thanks for all the information,

and congratulations to your Bronze medal in Havannah!

So, University of Alberta got both gold medals in

Hex and Havannah. Impressive.

Please, let us know more about the tournaments (especially

Havannah) when you find time to write something.

Give my greetings to Richard Lorenz!

Ingo.

Congratulations to the programmers!

What were the time settings/hardware?

Where can we see the recorded games?

Please connect your engines to Little Golem,

so humans (and other bots) can play them.

Actually the program was written by my disciple - Marcin Ciura with whom I work at Google.

I asked Marcin to greet Richard from you.

Cheers,

Przemek

BTW

We made a bet recently whose program is better on board size = 8.

10 games, at the beginning of Nov. We will share game notations after.

info from Marcin:

time settings: 30 seconds / move

pondering up to you (only Castro and initially Lajkonik did it)

hardware: up to you

Castro used a remote 8-core machine

Wanderer, Lajkonik, and PZN used laptops.

Shakti initially used a server in France, a laptop when the connection failed.

I have only Lajkonik's games. I'll ask its author to publish it.

Congrats all :) Any chance we might get to put Castro to te test at LG?

You can try to convince the author. He sometimes reads the LG havannah forum but hasn't played enough games to post yet!

congratulations all!

@przemek, as it's allowed to use remote machines over the net, you and marcin should totally take advantage of idle clusters in borg for mc simulations ;)

I was hoping the game records were published by the organizers

(automatically) but this page is still blank.

Have some patience.

It might take some days, even for classic games like chess and go,

until the game records will be available somewhere.

Ingo.

PS: All who are not happy with this may try to become organizer

of the computer olympiad in one of the next years - and do it quicker

when they get the event.

@z that's very tempting, I might try this next year. so far i was only thinking about using some idle machines for solving havannah on boards 4 & 5. (or at least to create a decent opening book)

@RingOfFire I was looking at castro's games on board 8 and its very convincing. But i'm not sure if he can beat Christian on board 10. And the bet itself isn't too fair, winning just 1 game out of 10, because of tactical oversight is quite likely.

I'm quite comfortable with the bet. I've also become quite comfortable with the swap rule, and I see no objections to introduce it in base-10 games too, even in the 2012 match, but that's up to te opposition to decide :)

@z Remote machines are allowed (as pondering) simply because the tournament is really friendly (in comparison to Chess or Go).

Shakti does not use a server in France but my personal computer in France (1 core 2.9GHzs) just because my laptop is really old :-D

With this number of participant, may be, we should have rules clearly defined in advance next time.

Tournament was really fun ; thanks to all participants :-)

@Ingo: "all who are not happy with this may try to become organizer"

I was just stating that the web page could update automatically,

like it works with the KGS bot tournaments.

But playing via a server (instead of on the board)

would make it too easy to participate remotely,

which the organizers probably don't want to encourage -

because they want real people, paying real registration fees.

Which makes sense, economically.

Hello Johannes,

> because they want real people,

that is right.

> paying real registration fees.

> Which makes sense, economically.

That is nonsense. The registration fees for the

Olympiad 2010 were

25 Euro for amateurs

50 Euro for semi-professionals

100 Euro for professionals.

So, these fees are mainly symbolic.

Ingo.

Thanks for the information. I didn't know about the symbolic fees.

(I'm at a (semi-)expensive conference right now.

I'm not complaining - they provide good value.)

I fully agree that having real people in one place

is a large part of the fun (and one of the main reasons

for attending such events).

But I'm also of the opinion that playing games between bots

by having operators putting stones on a board

(which I understand is what happened)

is technically "non-sense" (picking up a phrase that you used).

I can guess where this comes from (historically): closed-source

chess engines being deliberately incompatible, and running on

operating systems that may have had issues with multitasking

and networking - but that's so "last century" by now ...

Anyway I'm totally not complaining. I am just curious to see

how the bots played. I'm sure the game records will surface,

and I do hope the engines will be made available (e.g., playing on LG).

Best regards, Johannes.

Richard Lorentz pointed me to

this page

with the Havannah results (at the moment, it contains just the results,

not the game records).