Vlasov and Atomix(C)

For many people, when they search for information on atomic chess they’re bound to come upon Vlasov’s ATOMIC-HOMEPAGE.  The crowning achievement of this webpage is the innumerable people that have been introduced to atomic by reading that webpage.  But it’s also a bit empty and one is left wondering… Who exactly was Vlasov?

Vlasov was an old player at GICS (German Internet Chess Server).  GICS is where atomic was first introduced in 1995 and Vlasov was one of the very first to play atomic chess.  Obviously, he had a passion for it.  I don’t think anybody can deny that.  After all, he created the best known page about atomic chess up to when I began my atomic chess website.

I don’t know what happened to Vlasov.  But I owe him some thanks for creating his page.

Statistics for Vlasov      (Last disconnected Fri Nov 20 1998, 14:26 CET):

         rating     RD     win   loss   draw  total   best
Blitz     2010     52.7    142    210     57    409   2116 (01-Mar-96)
Standard  2050    121.7     12      2      2     16
Lightning 1743     48.2     70     76      9    155   1833 (18-Jul-96)
Wild      2122    107.5     22      2      1     25   2122 (05-Sep-96)
Bughouse  2404    235.5     26      0      0     26   2404 (28-Mar-96)
Atomic    2081     46.7     83     31      8    122   2092 (20-Aug-96)
Suicide   1544    215.5      0      3      0      3

 1: Vlasov ist nach ueber 2 Jahren Abstinenz wieder auferstanden.
 2: Danke an OliB
 3: Wer kennt mich noch von frueher ?

The focal point of Vlasov’s page is the Atomix opening book.  This is what most people looked at when they went to Vlasov’s page, the opening theory for atomic.  But that’s not what I like about it.  The theory itself is bad for the most part but wonderful for a historical perspective on who played atomic when it was first begun.

But first, who or what was Atomix?  Atomix was the first atomic chess playing chess program.  It was written by Peter “Nightrider” Schaefer, originally as an experimental program.  The opening book that Atomix used to decide its opening moves was submitted verbatim to Vlasov who posted it on his page.  This opening book was a remarkable accomplishment for its time.  You have to bear in mind that all that was done in 1996 and 1997 — they were basically writing theory as they went.

People have complained that the Atomix book was full of bad lines.  I have to gently chide those people here.  To better understand why the book is “bad”, we have to look at how the Atomix engine itself was programmed.  It used the book both as white and black and quite simply, the bad lines are in there so if the player played into one of them, Atomix would know how to win.  Of course the downfall was that once in a while, Atomix would play those bad lines itself.  But for the pioneering engine and the very first opening book, I don’t think they did a bad job at all.  I mean, something bad is worth knowing so you can take advantage of it if it’s played against you and also so you won’t play it.  The book isn’t really that bad as a learning tool… but bad to memorize for your openings.

Reading over the Atomix book, one might wonder how they came up with the names for certain openings.  That’s simple to answer.  Those players were the ones best known or the ones that used those lines the most.  Or they were the actual pioneer of said line.  There’s a lot to be said for being some of the first to ever play a new chess variant.  Or at least being the first to record it all.

Statistics for Atomix      (Last disconnected Wed Jun 11 1997,  1:56 CEST):

         rating     RD     win   loss   draw  total   best
Blitz     ----    350.0      0      0      0      0
Standard  ----    350.0      0      0      0      0
Lightning ----    350.0      0      0      0      0
Wild      ----    350.0      0      0      0      0
Bughouse  ----    350.0      0      0      0      0
Atomic    1603     44.2   6035   4526    225  10786   2011 (07-Nov-96)
Suicide   ----    350.0      0      0      0      0

 1: "atomix"
 2: ein experimentelles ATOMIC-Programm von "Nightrider" Peter Schaefer
 3: Eroeffnungsbuch von "JoelH" Jonas Hoffmann
 4: Danke an "Pfiffigunde" fuer viele Eroeffnungstips
 5: Tips und Bug reports, sowie Beitraege zur ATOMIC Eroeffnungstheorie sind 
\   immer willkommen; ausserdem such ich noch Sprueche (schlaue, dumme, 
\   witzige, geschmacklose...)
 6: man kann mit mir reden; wer eine sinnvolle Antwort haben will, muss mir 
\   eine message schicken.
 8: Viel Spass !
 9: Danke auch an : Grobi , EinsNull

We can see that Atomix certainly got quite some playing time.  It didn’t connect again after June 11, 1997.  The last revision of the Atomix opening book was January 22, 1997.  10,000 games between the time atomic was introduced and the last time it connected is nothing to sneeze at.  Especially when you consider that they basically had to write a chess program to play a chess variant that was just released.  This was the GICS (and original version) of Atomix.  There was another copy (two actually) playing at MEWIS, which I might address later.

These words above were written 22 March 2002, and updated again on 05 April 2005. Since these words were written, I’ve heard from JoelH. He sent an email in January of 2006:

From : HOFFMANN JONAS <xxxxxxxx@xxxxx.de>
Sent : Friday, January 6, 2006 3:59 AM
To : <nick.long@hotmail.com>
Subject : Your Atomic Page

Hi Nick,

I just found your Atomic-Page in the Net, and I am glad to see my User-Handle all around. You have to know, I’m “JoelH”, one of the Guys who wrote this first Atomic Openingbook. It’s really a long time since we made this stuff, but some lines are still played 🙂 It is interesting, that 1.Nh3, which was introduced on GICS by a the Computerprogramm TrojanKnight, is still unrefused and complicated for black. Some weeks ago I reactivated my Handle on FICS and played a couple Atomicgames, but its really hard, the players learned a lot, as well as I forgotten even the mainlines 🙁

Some Words to Atomix, a lot of the lost games came because of time and lag problems, the program had no timeseal.

Jonas ‘JoelH’ Hoffmann

Much thanks to JoelH for writing in to explain some of the commentary that I originally provided. In the years since, he’s even revived the Atomix program itself and started a blog of his own regarding the programming and updating of the engine. It was mainly done between 2014-2015, but much like everything else, it’s massive productivity in a short spurt of time. If you’re at all interested in general atomic chess concepts (obviously, since you’re reading this page) or variant chess engine writing by a hobbyist, give his site a read. The link is for the English translation, but if you can read German, you might prefer to read the native language that he posts in.

Back to the Atomix opening book “issues”. I’ve heard from a number of people in the late 2010s complaining about “bad lines” and the like. Remember that Atomix needed the book to know what was a bad line in the first place. It is not a cherry-picked opening tree full of solid lines. The bad lines are there so that Atomix could win, as the processing power back in the mid-90s was vastly inferior to the capabilities of today’s machines. And yes some lines from the mid-90s can be refuted by today’s players. This was an opening book made in 1995. By a limited number of players. With limited computing capabilities and lacking hindsight. Come on.

The Atomix book is a historical artifact. It’s a great way to study some concepts of the game of atomic itself. To see the development of theory. To see what has been tried in the past. To better understand the concepts at play within the opening. But by no means is it a modern opening book. And even a modern opening book would need to be intentionally limited simply not to ruin the game itself.

Other Posts