Monday, July 07, 2014

Cryptochessanalysis

"If you want new ideas, read old books..." -- Ivan Pavlov

I have been unconsciously devoted to cryptochessanalysis for many years.  Only recently did I recognize the depths of my delusion.  Like cryptanalysis, the goal of cryptochessanalysis is to crack a code to reveal hidden knowledge -- only instead of revealing secret information, it is directed specifically at understanding the secrets of chess openings.  Crytochessanalysis often begins by trying to decipher the hidden intentions and ideas of GMs and other strong players, but its true motive is more quixotic: to crack the code of chess itself by examining hints from those who have most deeply plumbed its mysteries -- those imagined to be, like the Priory of Sion, the keepers of the Grail.  I admit it is a crazy notion, at least once you carefully examine it, knowing full well that even GMs can fumble their way through theory, or construct elaborate magic acts to perform at the board which are revealed as hollow once you know the trick.  But I would not be surprised if there were others like me who have gotten caught up in a similar quest for truth in the opening.  

Cryptochessanalysis is unconsciously motivated by a fancifully Gothic story about chess, one that imagines our immortal game to be like The Da Vinci Code, where "the truth" about a particular opening line might be discovered through the close observation and study of games by its high priests.  Or, in true Gothic fashion, "the truth" is uncovered by studying games or analysis unearthed from the crypt of forgotten chess ideas, which sometimes come back to life like the Mummy released from its tomb.


Torre - Santasiere, Dimock Theme 1924
White to play after 11...O-O
I think it was my study of the Urusov Gambit that started me down the cryptochessanalytic path.   Tracking down the games of the Dimock Theme Tournament of 1924 in old newspapers, I felt like I was on the trail of true knowledge about the opening.  For example, when I first played over the game Torre-Santasiere (Brooklyn Daily Eagle, October 9, 1924), with its exciting Exchange sac 12.Rxe7, I knew this was a key tactic in the position. Torre's exact move order did not quite hold up to silicon-supported scrutiny, but he was definitely onto something.  One possible improvement on the line begins with 12.Qh4!? to be followed by Rxe7. Later, correspondence player Max Burkett tried 12.h4!? -- one idea being 12...h6?! 13.Rxe7! hxg5?! 14.hxg5!! with a winning attack for White.  I was amused when Tim Harding, in reviewing my analysis, wondered why Gabriel Velasco had overlooked this fascinating game in his book on Torre.  But the game had been completely forgotten until I dug it up.  And it was this game from the crypt, previously unknown to modern chess analysts, that pointed the way for future exploration.

Similar revelations came from an old piece of analysis by Al Horowitz, which led to my article on The Modern Horowitz Variation of the Max Lange Attack (blog/games/pgn).



This piece attracted significant attention, getting mentioned in Kaissiber #22 and in Dangerous Weapons 1.e4 e5.  But all I did was use Fritz to take a close look at an old article, and provide the world with a PGN file so they could carry the analysis further themselves.  Other adventures spelunking through the crypt of chess history have led me to explore The Panther and The Big Clamp.  In every case, it was my unconscious belief that new ideas were to be found in old books and old games that motivated my inquiries.

Not all cryptochessanalysis seeks its ideas in the past.  Sometimes my explorations begin with a puzzling game by a strong player, whom I suspect has done significant analysis before revealing his novel idea.  Back in 2006, I stumbled upon a fascinating game in the Advance Variation of the Caro-Kann that began 1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5 Qb6 6.Bd3! Bxd3 7.Qxd3 Qxb2 8.e6!!


Position after 8.e6!!
I knew instantly that I had found my latest subject for cryptochessanalysis. After some study, I finally wrote about the game in early 2007 in an article titled The Caveman Caro-Kann: Advance Variation with 4.h4 (blog/game/pgn). I played and analyzed the line for several years after that, collecting every published game with it.  There were clearly other cryptochessanalysts who had been drawn to the idea.  Eventually I put together all of my analysis in The Complete Caveman Caro-Kann (blog/game/pgn).  Even as I published that article, however, I wondered if I had not brought an end to the line in doing so.  Sometimes it is the complete mystery of an opening that makes it most interesting for new explorers.

One of my most elaborate adventures in cryptochessanalysis was in writing about what I called the "Guseinov Gambit," following the games of Gad Guseinov in two articles: Guseinov's Anti-Paulsen (blog/game/pgn) and Guseinov's Gambit Refuted? (blog/game/pgn).  These explorations were inspired by Guseinov's discussion of his ideas in an interview with Misha Savinov at ChessCafe.  After many hours of deep analysis, it was I who most doubted the line in the end, especially after I tracked down GM Guseinov's games on ICC (played as "GGuseinov") and found that he did not have a good answer to opponents playing the remedy suggested by Richard Palliser in Fighting the Anti-Sicilians (Everyman 2007).  The veil had begun to slip.  However, even though I came to a rather negative conclusion in the second article, I continued to play the line with great success and have yet to have an opponent play the critical line revealed by my analysis. 

If I ever do, I have something prepared....

Other cryptochessanalytic explorations inspired by mysterious novelties led to The Left Hook Grand Prix and The Werewolf (blog/game/pgn), both of which began with mere curiosity as to why strong players were playing strange moves.  It often turns out that these lines do not completely hold water -- there is always a leak somewhere.  But that does not mean they cannot lead to interesting chess!

In 2008, I analyzed the game Weeramantry--Bisguier, US Amateur Team 2008, which showed me that it was possible to play a fianchetto system in the open games. After all, if Bisguier could get away with it against strong opposition, then I certainly could against weaker players. This led to articles on A Black Fianchetto System in the Open Games, Part I and Part II.  It remains an unfinished project.  But now I know that the most interesting ciphers are the ones we have not solved.

2 comments:

Prof Camenares said...

A great article, nicely framing some great analysis over the year. The search for truth in chess has some similarities to the searches conducted by scientists, albeit in different arenas.

There too, careful examination of past results and new experiments can yield insights. In fact, I often wondered there can be a lot in common, and that chess can be studied in a systematic and scientific way.

Michael Goeller said...

I definitely think the systematic analysis of openings was inspired by science. And there are a lot of similarities between chess and all fields of study and discovery. That's why chess can be such a great training ground for academics -- as your own career seems to be demonstrating! Good to hear from you, Dr. Camenares.