DIR Return Create A Forum - Home
---------------------------------------------------------
The Chess Variant Forum
HTML https://chessvariantforum.createaforum.com
---------------------------------------------------------
*****************************************************
DIR Return to: Correspondence Game Requests
*****************************************************
#Post#: 474--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: Panzerschiff Date: February 16, 2018, 7:42 am
---------------------------------------------------------
It has been a long time and hard to use the word "current" for
this, but Cyber World Championships were held for Grand Chess
through the inventor's (Christian Freeling) web site in 1997,
1999 and 2001. No Cyber tournament took place after 2001 to my
knowledge. Consequently, I guess I would be still be the
current Cyber World Champion (Equivalent to correspondence chess
or similar turn based games conducted through a server) for
Grand Chess. (10x10 variant). Panzerschiff (John Vehre) Wayne
Schmittberger won the first two tournaments.
#Post#: 476--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: Greg Strong Date: February 16, 2018, 11:06 am
---------------------------------------------------------
[quote author=HGMuller link=topic=43.msg221#msg221
date=1516822842]as far as I know there has been no development
of engines that could play Gothic Chess at all, since
then.[/quote]
There's the new ChessV. Say, what ever happened to Spartacus?
#Post#: 480--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: HGMuller Date: February 16, 2018, 12:17 pm
---------------------------------------------------------
True, but I understood it is not WB compatible as an engine yet.
Spartacus is sort of in a sorry state. It seems I inadvertently
forked the development. (I was not using version control at the
time.) So I have two different versions of the executable, one
obviously broken for Spartan Chess (it gives +1.80 in the
initial position), the other for normal Chess. And there aren't
ay source-code files that correspond to their creation date.
Spartacus was supposed to be a varaiant-capable engine based on
an incrementally updated attack map. At one time I had that
working for normal Chess, but then I changed the board size from
32x8 to 24x10 in order to support Grand Chess, and that broke
the incremental update, so that I temporarily disabled it.
I suppose I could fix that, but in the mean time my ideas for
how to incrementally update an attack map have evolved a lot,
and even if I could get Spartacus working as designed, I would
be quite unhappy with it. I applied my new design for the attack
map in my Tenjiku-Shogi engine, and it works like a charm there.
So it seems my time would be better spent to completely rewrite
Spartacus, based on the same principles.
#Post#: 481--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: Greg Strong Date: February 16, 2018, 10:39 pm
---------------------------------------------------------
[quote author=HGMuller link=topic=43.msg480#msg480
date=1518805063]
True, but I understood it is not WB compatible as an engine
yet.[/quote]
Well, you didn't specifically stipulate that ;D
[quote author=HGMuller link=topic=43.msg480#msg480
date=1518805063]Spartacus is sort of in a sorry state. It seems
I inadvertently forked the development. (I was not using version
control at the time.) So I have two different versions of the
executable, one obviously broken for Spartan Chess (it gives
+1.80 in the initial position), the other for normal Chess. And
there aren't ay source-code files that correspond to their
creation date.
Spartacus was supposed to be a varaiant-capable engine based on
an incrementally updated attack map. At one time I had that
working for normal Chess, but then I changed the board size from
32x8 to 24x10 in order to support Grand Chess, and that broke
the incremental update, so that I temporarily disabled it.
I suppose I could fix that, but in the mean time my ideas for
how to incrementally update an attack map have evolved a lot,
and even if I could get Spartacus working as designed, I would
be quite unhappy with it. I applied my new design for the attack
map in my Tenjiku-Shogi engine, and it works like a charm there.
So it seems my time would be better spent to completely rewrite
Spartacus, based on the same principles.
[/quote]
Sorry to hear that. I'm revisiting Quadrox now but it's been so
long and I have so many old hard drives lying around I'm not at
all sure that the code I'm looking at was the latest, so I can
sympathize. Although if you've discovered a better approach in
the meantime, maybe it's not too much of a loss. I hope you get
back to it. We could really use another strong universal chess
engine. Fairy-Max may be a weak engine, and yet it remains one
of the strongest universal engines ...
Incrementally updated attack maps are an interesting idea that
I've thought about from time to time for at least a decade. For
large board variants at least I think the idea has great
promise. But when I think deeply about it and try to nail down
all the details, I see it's pretty complicated and I'm not
confident I can get it working correctly without spending a
*TON* of time developing and debugging, so I keep putting it
off. I've been able to create some decent engines, but I'm not
the bad-ass programmer you are. (not being sarcastic.)
#Post#: 482--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: HGMuller Date: February 17, 2018, 4:10 am
---------------------------------------------------------
True, the incremental update was hard, and especally hard in the
case of Tenjiku Shogi, where you have to deal with jumping
generals, Fire-Demon burning, area moves... I did make some
allowences there. E.g. the attack map only indicates which
victims are in area-move range, and whether the capture is
actualy possible is the verified from scratch during move
generation. And Demon burns are generated from scratch in every
node, assisted by an incremetally updated set of white/black
'bitstrips' (partial bitboards covering 3 files). I am still
puzzling on how to incremetally do hook movers, so that the
engine could do the larger variants as well.
I just started with a routine to generate the attack map from
scratch, and compared that map with the incrementally updated
one in every node of a perft. But I was surprised how quickly I
got it working. Just a few cases of non-terminating loops,
because I forgot to increment the index, and such. And throwing
out all the stuff specific for Tenjiku would be quite easy.
The problem is that the number of Chess-programming projects I
have is increasing, and especially the number of 'nearly
finished' projects is increasing. 'Inferno' (the Tenjiku engine)
has not been released at all, and is waiting for addition of an
evaluation term that is crucial in the opening phase. And
'CrazyWa' (my engine for variants with drops on boards up to
11x11) is still waiting for a release of a cleaned-up source
code, and addition of some variants of interest (mainly Kyoto
Shogi). And I just started on writing a Chess program in
assembler for the 'Gigatron' retro computer.
I also would like to equip the 'Interactive Diagram' (written in
JavaScript) with an AI. And develop (and in particular tune) an
engine for Peace Chess (Paco Shako), where no clue exists for
what is optimal strategy, so that it would have to be some
self-learning thing. Perhaps Alpha Zero can provide some
inspiration for that. It would be cool to write a program that
learns to play just from the rules anyway.
Working on a program for 'ordinary chess variants' has become a
somewhat boring task in comparison...
#Post#: 483--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: GothicChessInventor Date: February 17, 2018, 10:18 am
---------------------------------------------------------
[quote author=chilipepper link=topic=43.msg228#msg228
date=1516836018]
Here is an update on some of the star players based on
information posted on this thread. I didn't yet add the
information from ebinola's post, but plan to do so soon (within
a few days). :o
[/quote]
There are many Gothic Chess programs that are not on the list.
The oldest = CapaGNU Modified, which was a DOS program that
played Capablanca Chess which could be modified to play Gothic
Chess.
Zillions of Games has a Gothic Chess implementation.
Gothic Vortex was the creation of a "Dream Team" of programmers
where I was the lead developer. Derived from Robert Hyatt's
Crafty 19.something, with programmers Gil Dodgen, Steven Dodgen,
Ed Trice, Eugene Nalimov, Marc Bourzutschky, and Andrew Kadatch.
TSCP Gothic was a modified version of Tom's Simple Chess Program
(by Tom Kerrigan) made by Dutch programmer Michel Langeveld.
S.M.I.R.F. ("Strategiespielprogramm Mit Intelligent
Rückkoppelnden Funktionen" which is Strategic Game Program With
Intelligent Recursing Facilities in German) by Reinhard
Scharnagl.
Joker80 by H.G. Muller
FairyMax by H.G. Muller
Variant Pulverizer, originally named "Variant Shredder" but a
strongly worded communication from Stefan Meyer-Kahlen (the
programmer of the "Shredder" chess engine) quickly disavowed
authorship and requested me to contact that program's author to
insist he change the name. (He did so, then withdrew his program
from all future events after a strong showing.)
There are others that I am forgetting, and H.G. can certainly
add the ones from his own "Battle of the Goths" events he
hosted.
As for "World Champions" of Gothic Chess, or other variants, I
am hesitant to use any such title because it really pales in
comparison to how the FIDE title is organized. I would say, in
the absence of a World Governing Body where the variant is
played in live, over-the-board tournaments in multiple
countries, with certified tournament directors and a
diligently-maintained ratings list, we should avoid such a
title.
Instead, I propose using the label: Recognized Strongest Player
From my involvement with Gothic Chess, here is my list of known
"RSP" based on their years of activity
2000 Ed Trice
2001 Larry Kaufmann (I played over-the-board with Larry. He
consistently swapped into winning chess positions and killed
me).
2002 Michael De La Maza (scored 7 wins and 2 draws in the under
2000 section of the 8x8 chess World Open in the year 2001)
2003 Ed Trice
2004 John Vehre
2005 Bobby Fischer (played online at GothicChessLive.com under
the pseudonym of "Umbra")
2006 Susan Polgar (Fischer came online less and less frequently,
and Susan beat me pretty bad)
2007 Bobby Fischer
2008 Anatoly Karpov (was preparing to face Fischer, safe to say
he was the best)
2009-2015 No activity while I built a Supercomputer Design Firm
2016 Ed Trice
2017 Ed Trice
It should be further noted that this is no way implies a "line
of succession." In 2000 and 2001, we were still a fledgling
variant, and even 1 game with 1 person showcasing dominance is
clearly enough to convince me who belongs on the list. My 2
games with Larry Kaufman in 2001 were enough to convince me of
that, and watching Michael De La Maza manhandle some players in
the earliest version of our online playing site in 2002 led me
to believe I would not outlast him in a set match. Both of these
players were not active beyond this time, so players coming
after had not crossed swords with either of them. John Vehre
played a correspondence game with me, and I learned very quickly
you can't rely on e4 in Gothic Chess the way you can in 8x8
chess. Fischer was destroying everyone online, giving them 10
minute vs. 1 minute odds and mopping them up. He died 6 months
before the match between he and Karpov was slated to take place.
In Fischer's absence, Karpov was clearly at the top of the list.
It was also the start of a period of Gothic Chess decline,
because I owed the match financiers over $800,000. They
stipulated that should Fischer not appear and play, for whatever
reason, I had to pay them back TWICE the interest that accrued
on their portion of the $15,000,000 that was raised for the
match. Well, over a 21-month period of time, this interest
totalled $400,000, so double that, and that is what I paid back
(painfully!) So you can see why I was not interested in playing
much for a long, long time. In between that, Susan Polgar played
me in 2006, and won in convincing style. She would be awesome in
just about any variant I would imagine. In the absence of the
Titans, and going over all of the 500+ games I have played, I
improved considerably. Without a doubt, I am also at a plateau,
and the next strong chess player who comes along could probably
knock me off the perch.
#Post#: 490--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: GothicChessInventor Date: February 18, 2018, 11:42 pm
---------------------------------------------------------
[quote author=HGMuller link=topic=43.msg221#msg221
date=1516822842]
Indeed, the Gothic tournaments were great. The last Battle of
the Goths event is already quite some time ago, but as far as I
know there has been no development of engines that could play
Gothic Chess at all, since then. Bihasa won the last tourney
convincing. (But of course only WinBoard-compatible engines
could compete, so Gothic Vortex wasn't there.)
[/quote]
I recently heard from someone named "Fernando" who said he is
the author of Bihasa. He contacted me through the TalkChess
Forum.
Over the years, I have heard from other engine developers, or I
should say, programmers who started developing Gothic Chess
engines to various degrees. Usually they would ask about the
piece values for the new pieces, and/or how to adjust the values
for the chess pieces on the 8x8 board to the 10x8 configuration.
Of the dozen queries only 3 made it across the finish line. I
would inevitably get a request to play a game with the engine
over the phone, which I would accept. Incidentally, nobody ever
gave me the white pieces :)
In retrospect, I probably should have thrown a game or two along
the way. After blanking their digital offspring a few times, I
would no longer hear back from them.
I understand their plight. I can't count the number of programs
I have started and never finished. Perhaps it was just an
interesting distraction for them, knowing full well how hard it
would be to create a strong chess engine with so much
competition that is out there right now. Anyway, who knows what
lies beneath the surface. Another great engine might be lurking
out there somewhere.
#Post#: 493--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: HGMuller Date: February 19, 2018, 5:14 am
---------------------------------------------------------
Actually his name is Ferdinand (or Ferdy) Mosca, from the
Philippines. Fernando is another guy, who frequently posts in
the CTF sub-forum. Ferdinand is a very talented chess
programmer, and a real asset for the community of chess
variants. He has very many engines to his name, all dedicated to
a different chess variant. (Presumably they are all derived from
the same code base.) And he offers free download of their
executables, see:
HTML https://sites.google.com/site/deuteriumengine/files
And his engines are quite strong, not at all comparable with
simple, minimalistic general-purpose codes like Fairy-Max.
Although the state of the art for chess variants is not nearly
as advanced as that in orthodox chess.
Recently there is also a project to convert Stockfish, the
absolute top engine for orthodox Chess, to play Chess variants.
With amazing success: in Crazyhouse it blows away all
competition, even Ferdy's engine 'Immortal'. The Stockfish code
depends very heavily on the board being 8x8, however. So it is
unlikely it will ever be able to do Gothic Chess. They are
working on Seirawan Chess, however, which also supplements the
FIDE piece set with an ArchBishop and Chancellor, but on 8x8.
(Too crowded, for my taste.)
#Post#: 494--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: GothicChessInventor Date: February 19, 2018, 9:03 am
---------------------------------------------------------
[quote author=HGMuller link=topic=43.msg493#msg493
date=1519038847]
The Stockfish code depends very heavily on the board being 8x8,
however. So it is unlikely it will ever be able to do Gothic
Chess.
[/quote]
That's what they told me about Crafty, but Gil Dodgen and Steve
Dodgen had a clever idea: Use "operator overloading" to convert
the 8x8 operations to 10x8. All of the 64-bit variables were
converted to structs with a 16-bit component and a 64-bit
component. The flow from one field to the other was done in the
operator overloading definitions. The hard part was changing the
indices of all of the crafty "rotated bitboards" to their
80-square counterparts. I put this in the multi-screen About Box
to Gothic Vortex...
HTML https://i.imgur.com/nO4m2Dc.jpg
#Post#: 496--------------------------------------------------
Re: Who are Today's World Champions in Variant Chess?
By: chilipepper Date: February 19, 2018, 12:29 pm
---------------------------------------------------------
[quote author=Panzerschiff link=topic=43.msg474#msg474
date=1518788571]
It has been a long time and hard to use the word "current" for
this, but Cyber World Championships were held for Grand Chess
through the inventor's (Christian Freeling) web site in 1997,
1999 and 2001. No Cyber tournament took place after 2001 to my
knowledge. Consequently, I guess I would be still be the
current Cyber World Champion (Equivalent to correspondence chess
or similar turn based games conducted through a server) for
Grand Chess. (10x10 variant). Panzerschiff (John Vehre) Wayne
Schmittberger won the first two tournaments.
[/quote]
What was the format for the Grand Chess Cyber World
Championship? Is that a variant chess-engine tournament, or
engine-assisted play?
The listing that I've produced is intended to show the World
Champions for strictly human-play. But engine vs. engine play is
interesting and I enjoy reading about it. Not sure if anyone has
started a thread specifically for that topic yet. It's awesome
to see how engines have improved and see who has developed the
best code, particularly for individual games. :)
*****************************************************
DIR Previous Page
DIR Next Page