26 February 2010

Harbin Ice and Snow Sculpture

Harbin Snow and Ice World © Flickr user SlytherinPrefect under
Creative Commons.

'Chess board: Pictures of the snow and ice sculpture park just a little north of downtown Harbin.' • Wikipedia: Harbin International Ice and Snow Sculpture Festival.

25 February 2010

Knockout Champions

While working on my post titled Calculating Collusion, a critique of Did the Soviets Collude? A Statistical Analysis of Championship Chess 1940-64 by Moul and Nye, a subject unrelated to collusion caught my attention.

More recently, Ryvkin (2005) has focused attention on the problem of how tournament design not only motivates effort but can also be structured to reveal the relative strengths of various players. In this light, the purpose of good tournament design is to maximize the probability that the result of the tournament will produce outcomes that conform to the "true," unobserved relative strengths of the participants. Many discussions of tournament structure have instead focused on "competitiveness." The latter goal actually values a large number of upsets or unlikely results -- such as seem likely to emerge from multi-round knockout tournaments which are common in tennis or collegiate sports.

In contrast, the chess community seems to put a very high value on competitions that accurately "reveal" the true strengths of players. Judging by the history of the world championship and its biases in favor of the incumbent world champion, there seems to be a revealed preference for a system that anoints the strongest player as World Champion. Periods when the official world champion was clearly not the strongest player -- as occurred when Garry Kasparov split from FIDE in the 1990s and created a parallel system for the championship -- have seen a loss of legitimacy for the systems that produced champions not widely viewed as the dominant players.

[Reference: Ryvkin, Dmitry (2005). "Tournaments: A Review." Unpublished working paper.]

The observation that, among chess players (is this unique to chess?), 'there seems to be a revealed preference for a system that anoints the strongest player as World Champion' is certainly true. The 'loss of legitimacy for the systems that produced champions not widely viewed as the dominant players' refers, of course, to the FIDE knockout tournaments, won successively by Khalifman (1999), Anand (2000), Ponomariov (2001), and Kasimdzhanov (2004). Excluding perhaps the event won by Anand, all of these knockout tournaments were criticized for producing a less-than-worthy World Champion.

As for the knockout event that produced a challenger -- Anand (1997) -- or the events that qualified players to the next stage of the championship -- Aronian (2005), Kamsky (2007), and Gelfand (2009) -- there was no criticism whatsoever. The chess world accepts the knockout format for identifying a challenger, but not for choosing a champion. Why should this be?

23 February 2010

Operators and Things (*)

My most recent chess960 post, A Few Novel Ideas, got me thinking about a subject that is not limited to chess960: What factors make one player more successful at correspondence chess than another player. I came up with the following list.

  • Chess Knowledge
  • Engine Use
  • Work Capacity

I could have also mentioned Talent, but some people would say that it plays no role whatsoever, so I left it out. Next I came up with four categories of player and assigned each category a value for each success factor, according to the following table.

For example, a titled player has a higher capacity for work than a Master/Expert, while a Master/Expert has greater Chess Knowledge than a player in the next tier, an Average Player. My assignments are all subjective, and someone else might quibble with them, but they work for me.

Finally, I added another category called Engine Operator. These are players who simply transmit the best move calculated by a chess engine after letting it run for some time. I've encountered a few of these players in my own games and am certain that they exist. Where do these players fit in the hierarchy?

I believe that operators currently fall in the band between Master/Expert and Average Player, overlapping both. This means that it is possible to defeat them with good chess knowledge and some concentrated work on that specific game. Some positions are more amenable to knowledge and work than are other positions, but that isn't relevant here.

As stronger chess engines become commercially available, the playing strength of operators is constantly on the rise. Will they eventually occupy the highest tier? This would mean that chess knowledge and work capacity will ultimately count for little to nothing.


(*) Operators and Things: The Inner Life of a Schizophrenic by Barbara O'Brien.

22 February 2010

Specialization and Secrecy

In my last post, 'An Enormous Investment of Effort and Time', I quoted from Bareev and Levitov's 'From London to Elista', regarding the pre-match effort involved in World Championship Opening Preparation. Bareev, who served as Kramnik's second in 2000 and in 2004, continued his explanation of pre-match preparation with a discussion of the team of assistants who helped Kramnik prepare his openings for the 2000 match with Kasparov (p.16):

Team Politics

Levitov: On what principle did [Kramnik] form his team before the match with Kasparov?

Bareev: On the professional principle -- those who were capable of work, could withstand the serious physical and psychological burden, and wouldn't give up. A limited number of people participated in the preparation. The team that formed was international, on the whole each person worked at home, and Kramnik himself held several study and coaching meetings with various chess players. [...]

In general, each person answered for his own sphere. I worked on the white color [e.g. Nimzo-Indian], with the exception of the Gruenfeld Defense. Dolmatov and Kramnik prepared the Berlin Wall together in about two weeks. Neither Lautier nor I knew what, specifically, would be our main weapon against 1.e4.

L: That's nice! [Kramnik] trusts you!

B: I think that in the first place, [Kramnik] himself didn't even know what would happen. In the second place, experience suggested how he should behave. It is well known that Kasparov -- a representative of the Soviet chess school -- is a pupil of Botvinnik and Karpov. They were famous for intrigues and behind-the-scene squabbles, which they won on most occasions. That's why it was impossible to absolutely exclude the possibility that an attempt would be made to -- I'm not afraid to use this word -- swipe something. In case of a loss in some area, Kramnik divided the field. That way something would always remain. [...]

Continuing the subject of who was responsible for what, I'll tell you that Kramnik and Lautier worked on the Scotch Game (Kasparov didn't play this opening, but it came in handy later, in the match with [Deep Fritz 2002]). If we're talking about the main opening for White, then Illescas worked on the Gruenfeld, and he did very serious work on it.

Based on the assumption that it was impossible to look at everything, Kramnik determined which openings might occur and which wouldn't, and cut out the unnecessary. He thought that 1.e4 would be Kasparov's main weapon. 1.c4 [Kramnik] himself plays as White, so he didn't put any particular work into this move. And about 1.d4 he simply decided that it wouldn't happen in the match -- a mistake that almost cost him a loss in game 15.

The degree of specialization and accompanying secrecy is surprising until we recognize the unfortunate incidents that occurred in previous matches, e.g. The Dorfman Affair and The Vladimirov Affair. Kramnik won the match +2-0=13, both wins with the White pieces, proving that his strategy was successful.

19 February 2010

Mission: Impossible (Se2 Ep17)

Wikipedia: 'Original airdate 14 January 1968; A gold shipment, sent to fund an anti-communist resistance movement, has been seized by the country's government and the IMF must get it to its intended recipients, but an international chess champion and con artist is also after the gold.'

M:I Se2 Ep17 "A Game of Chess" (Part 1) (10:00) • 'Mission: Impossible Season 2 Episode 17 "A Game of Chess" (1/5)'

Those were the days! Top players wore ear pieces attached to wires and blew smoke at their opponents. 'Phelps: Is the computer set? Barney: And checked. It's unbeatable at chess and it even makes time fly.'

18 February 2010

Favorite Icons and MSPaint Adventures

Today's regularly scheduled chess post has been cancelled due to an unforeseen, excessive waste of time. The usage statistics on my World Chess Championship site spiked a few times this month and I decided to find out what was happening.

The daily stats produced by my server host revealed nothing unusual, so I dusted off my homemade stat analysis software, downloaded the server log for February to-date, and imported the log into the software's database. A little tweaking revealed that someone (or something) has been requesting my FAVICON.ICO file every ten seconds for days at a time. This was the traffic spike I noticed.

The FAVICON.ICO file is the source of that little icon that is displayed in your browser bookmarks or on the browser address bar next to the address of the web page. It's a tiny 16x16 pixel image file that sits in the web site's root directory in case browsers ask for it. The fuzzy image at the beginning of this post shows each separate pixel in my icon, a chess King's crown.

Why would someone request this every ten seconds? Beats me. If I find out more, I'll mention it here.


While I was looking at the stats, I noted my site's referrers, i.e. other sites that are sending me traffic. The top five were...

773 www.google.com
233 en.wikipedia.org
139 www.google.co.uk
111 search.yahoo.com
108 www.bing.com

...meaning that www.google.com, for example, sent me 773 visitors over the period covered by the log, which was about two weeks.

The Google number are somewhat understated because the longest URLs are no longer processed correctly by my software and because there are so many Google domains. I counted 86 different Google domains (www.google.co.in, www.google.ca, etc.) that sent me traffic, out of 270 different domains total. By comparison, Yahoo traffic came from 23 different domains (uk.search.yahoo.com, etc.), Wikipedia from 21 (pl.wikipedia.org, etc.), and Bing only from the single www.bing.com.

Along with the search engines I received a lot of visitors from sites that have nothing to do with chess. One example was a forum thread called MSPaint Adventures - Homestuck Edition (somethingawful.com). The thread was discussing this picture of a chess board -- You fill each empty square with a bit of MOTOR OIL... -- when someone commented, 'I'm hoping that the game was a real game and some nerd has annotated it', and someone else pointed to my 'Every Move Explained' analysis of 1960 Leipzig - Letelier vs. Fischer. Since this is getting stranger and stranger, I'd better stop here.

16 February 2010

The Modern Approach to the Opening

In the 1970s and 1980s, Polugaevsky demonstrated

a new, modern approach to the opening: if you want to gain an advantage in critical variations, you have to take a risk, and play concretely and sharply, move by move. (Kasparov's My Great Predecessors III, p.105)

That sentence, which was italicized in the original, is the best summary of opening play that I have ever seen. It is self-explanatory, except perhaps for the phrase playing 'concretely', which is, as I understand it, the opposite of playing 'intuitively'. It means calculating everything, rather than relying on positional considerations.

15 February 2010

'An Enormous Investment of Effort and Time'

The last three posts in this series on World Championship Opening Preparation have been drawn from the 1990 Kasparov - Karpov match. Let's skip forward a decade to the 2000 Kramnik - Kasparov match, and return to the source of that post on opening preparation, 'From London to Elista' by Bareev and Levitov, published in 2007. Bareev was one of Kramnik's seconds for the matches against Kasparov and Leko (p.14):

Levitov: In recent years there's been a revolution in opening preparation. What only Kasparov used to do, everyone now does, and there's an enormous amount of work for the top-class grandmasters to shoulder in the search for a new move. It's become difficult to find any untravelled path.

Bareev: This is mainly a problem for the top chess players. In essence, scrupulous analytical work that demands an enormous investment of effort and time is done by only a handful of people. They find and play the novelties, and the overwhelming majority of chess players sit with the Internet, patiently waiting for a fresh idea to appear in a variation that's interesting to them, and as soon as they see something new, having quickly checked it on their computer, they rush to successfully use it first. [...]

Chess grandmasters live in the proverbial ivory tower. When Bareev speaks about 'the overwhelming majority of chess players' he apparently means something more narrow like 'expert chess players'. As I recently pointed out in Do You Care about Today's GMs?, 'the overwhelming majority of chess players' are blissfully unaware of novelties played by grandmasters.

The co-authors go on to make an important observation about the role of memorization in top-level chess.

L: So, does a grandmaster today really have to sit for two or three hours before the game and simply refresh his memory with a huge number of opening lines? To learn everything from cover to cover -- otherwise he can't do anything?

B: If you've done a huge amount of work, then yes -- you have to refresh your memory, because otherwise you won't remember it for the game, and it'll be as if you never had this knowledge. And Kramnik had to spend several hours before a game refreshing his information, memorizing certain variations.

L: So there's no creative work during a match?

B: Of course, there is. Where do you think novelties come from?! But the volume of information that the chess player has to deal with is now so great that a team of assistants woring several hours a day is needed, in order to work through a framework of variations. And to refresh the variations and ideas before the game, this takes several hours.

There's an equally curious disconnect in the last two paragraphs. Where do I think novelties come from? As Bareev said just before this, they come from 'scrupulous analytical work that demands an enormous investment of effort and time'. That work is not done 'during a match'; it is done before a match.

In the next post for this series, I'll extract Bareev's observations on the 'team of assistants'. Seconds, anyone?

12 February 2010

Big Dummies Play Chess

798 Art Zone, Beijing © Flickr user rickz under Creative Commons.

Like I said...

11 February 2010

Chess Life Collection

Ever wonder what your old Chess Life magazines are worth? A recent batch of eBay auctions should give you an idea.

The winning bids are in US$.

09 February 2010

Predecessors Cross-Referenced

Kasparov's Predecessors series is a great set of books for browsing, whether reading or playing through the games, but as a reference it is nearly useless. In each volume, the 'Table of Contents' offers a bare minimum and there is no index of the contents of the volume (except for the games). The same event, like a World Championship match, is often covered under several players who can be in different volumes.

To execute a specific project I had in mind, I cross-referenced volumes II and III by event. A sample of the results is shown in the following graphic.

For each of the 13 players covered in a chapter (like Euwe) or a chapter section (like Keres), the table lists year and event with a page reference by player (BOT = Botvinnik, BRO = Bronstein, etc). My table covers only World Championship events, gives only the first reference for an event / player even though there might be several, and undoubtedly contains errors. After I get some confidence in the data, I'll create a usable copy on a web page.

08 February 2010

Theatrics in Professional Chess

The next post in the series on World Championship Opening Preparation continues the previous posts on the 1990 Kasparov - Karpov match (KK5) -- One Hundred Days for an Opening Repertoire and The Chief Trainer -- with input from a recent post on chess960 : Fischer: 'The *Old* Chess Is Dead'. Karpov wrote about the 1990 World Championship match in the last chapter of his book 'Karpov on Karpov'. Here's an excerpt (p.214).

Several key moments [of the 1990 match] stand out. The first game is always used for intelligence gathering, for checking yourself and your opponent, for checking stability, energy, depth, and mood. The game was fierce and compact. We both showed that we weren't opposed to winning right off the bat, but neither did we rush to do it: a win in the first game is a bad omen.

I lost the second game. Kasparov caught me with his home analysis. None of the spectators in the hall knew this. They only saw how Kasparov sweated, racked his brains, tormented himself, squeezed his temples. But during the game, information leaked from the press center that Kasparov's variation was being developed by his team right up to the moment of his sacrifice of a Bishop -- somewhere around the 30th to 40th move.

His plan turned out to be an excellent one. Give Kasparov's team credit for discovering and developing this variation and bolstering it with some profound analysis. As far as this game is concerned, it is all right to speak of an excellently conducted prematch preparation and of an excellent home analysis. Just don't talk about the playing. Because the playing was only on one side -- mine. Kasparov can only be lauded for execution.

Compare Karpov's personal experience with Fischer's general observation.

You have very interesting, beautiful pre-arranged games being created by very intelligent players, working with computers, working in teams. I have no objections to people creating such games, but they must say these are pre-arranged games, but they must not claim that they are finding the moves over the board.

Karpov's observations were from 1990, Fischer's from 2002. Although computer analysis was still experimental in 1990, and we don't know how big a role it played in Kasparov's preparation, the 13th World Champion was a pioneer in that area.

Assuming that some portion of Karpov's criticism is just sour grapes, I still had to smile at his description of Kasparov's theatrics (he 'sweated, racked his brains, tormented himself, squeezed his temples'), because other players have described similar behavior sitting opposite Kasparov. Is professional chess on the same path as professional wrestling?

For the complete game mentioned by Karpov in the excerpt, see...

Garry Kasparov vs Anatoli Karpov; World Championship Match 1990

...on Chessgames.com.

05 February 2010

Searching For Bobby Fitsher

'You're watching Movie Talks with Chris and Ken. This movie speaks a lot about family, speaks a lot about family and sports, and it's not clichéd!' • Unfortunately, the clip is often sloppy, but it makes some good points.

Movies Talks 'Searching For Bobby Fitsher' (Part 1) (8:51) • 'Me and Ken talk about this old movie that was forgotten many years ago.'

More: Movie Talks 'Searching for Bobby Fitsher' (Part 2), with (a coincidence?) lots of links to 'Bobby Fisher' clips. • 'Searching For Bobby Fischer'? Now you'll find him...

04 February 2010

Capablanca in Memoriam

We had bulkcover (World Championship Etiquette) and we had tiqu (Photogenic Spassky). The latest seller to appear on eBay with dozens of unique chess photos, all from Cuba, is xavgallery.

For more, see Items for Sale From: xavgallery.

02 February 2010

The Drosophila of Unattributed Quotes

In my post on Scientific American's Chess Neuroscience, I noted that 'The Expert Mind' by Philip E. Ross, the only full length feature article in the list, deserved its own post. The article, which appeared in the August 2006 issue of Scientific American, is no longer available on SciAm.com, but if you note the article's first sentence

A man walks along the inside of a circle of chess tables...

and click your heels three times, you might be able to locate a copy somewhere on the Web. In case you can't find it, here's a synopsis: 'The Expert Mind' was subtitled

Studies of the mental processes of chess grandmasters have revealed clues to how people become experts in other fields as well.

The main points were summarized

Overview / Lessons from Chess
  • Because skill at chess can be easily measured and subjected to laboratory experiments, the game has become an important test bed for theories in cognitive science.

  • Researchers have found evidence that chess grandmasters rely on a vast store of knowledge of game positions. Some scientists have theorized that grandmasters organize information in chunks, which can be quickly retrieved from long-term memory and manipulated in working memory.

  • To accumulate this body of knowledge, grandmasters typically engage in years of effortful study, continuously tackling challenges that lie just beyond their competence. The top performers in music, mathematics and sports apears to gain their expertise in the same way, motivated by competition and the joy of victory.

Three corresponding sections were titled

  • [Chess:] The Drosophila of Cognitive Science,
  • Chunking Theory, and
  • A Proliferation of Prodigies.

and three quotes were highlighted.

  • Much of the chess master's advantage over the novice derives from the first few seconds of thought.
  • The 10-year rule states that it takes approximately a decade of heavy labor to master any field.
  • The preponderance of psychological evidence indicates that experts are made, not born.

There is much material in Ross's article that is worthy of further comment and I'll save that for future post(s), but before I sign off, I'll nitpick two points. The first is the Capablanca quote in the opening anecdote.

How did he play so well, so quickly? And how far ahead could he calculate under such constraints [a 28 board simul]? "I see only one move ahead," Capablanca is said to have answered, "but it is always the correct one."

Coming from one of the best chess players of all time, this well known quote seems so unworthy that it must have been said humorously. Indeed, in a review of a recent Kasparov book, Kasparov's How Life Imitates Chess by Edward Winter, the world's leading chess historian and Capablanca scholar wrote,

The heading to chapter 5 [...] professes to cite Capablanca: "I see only one move ahead, but it is always the correct one." No source is given, of course, because none is known (see, for instance, the discussion in C.N. 4483), and that of itself should have resulted in the quote being expunged. Are there not enough authenticated chess observations to choose from?

Chess Note 4483 appeared in Chess Notes : July 2006, where Winter explained,

4483. The best move: [A correspondent] draws attention to a recent chess article by Philip E. Ross in the Scientific American which is available on-line. As regards the statement attributed to Capablanca that he saw only one move ahead but always the correct one, we refer readers to C.N. 2085 (see page 325 of Kings, Commoners and Knaves), where a correspondent noted Al Horowitz's claim that the words were said not by the Cuban but by 'New York’s East-side pride', which suggested Charles Jaffe [...]

It was apparently Jaffe who said, 'I see only one move ahead...', as a witty response to Capablanca's more conventional claim '[I see] about ten moves'. By coincidence, Kasparov repeated the Capablanca anecdote in a very recent piece titled The Chess Master and the Computer, which touches on the same themes as the Ross article.

As for how many moves ahead a grandmaster sees, Russkin-Gutman [whose book was the object of Kasparov's review] makes much of the answer attributed to the great Cuban world champion José Raúl Capablanca, among others: "Just one, the best one." This answer is as good or bad as any other, a pithy way of disposing with an attempt by an outsider to ask something insightful and failing to do so. It's the equivalent of asking Lance Armstrong how many times he shifts gears during the Tour de France. The only real answer, "It depends on the position and how much time I have," is unsatisfying.

Nitpick number two is on the quote 'Chess is the Drosophila of Cognitive Science'. The original and better known quote is 'Chess is the Drosophila of Artificial Intelligence'. Although cognitive science and artificial intelligence are related, they aren't the same thing. By confusing them, we are denying Alexander Kronrod's greatest claim to fame. For the moment Google gives AI a three-to-one advantage over cognitive science in references to 'Chess is the Drosophila of [whatever]', but since chess wins in either case, I'll stop this grumbling and say no more.

01 February 2010

The Chief Trainer

On the subject of World Championship Opening Preparation, who can be more authoritative than the World Champions themselves? In my last post, One Hundred Days for an Opening Repertoire, I quoted from Kasparov's book 'Kasparov v Karpov 1990', and in this post I'll continue with another passage from the same book (p.11). It gives more insights into what can go wrong with the best planned preparation.

For the first time my team had no chief trainer, such as Alexander Nikitin, who worked for many years with me. He left without explaining the reasons. As he said in an interview, 'Kasparov and I have parted without any fuss'. Back in April [1990] I discussed with him the strategy for the forthcoming match, but he left, and I did not try to detain him. Perhaps he thought that he could no longer cope with the enormous tension that accompanies all World Championship matches, without exception.

At any event, I encountered a problem that there was not that person who would have headed our operational headquarters. My team was young. Although, almost all the trainers had worked with me at some time before: Zurab Azmaiparashvili and Sergey Dolmatov in Seville [1987], and Mikhail Gurevich in Leningrad [1986]. Alexander Shakarov has been helping me since my childhood days. And only Giya Georgadze was a 'new recruit'.

In short, I myself had to carry out the functions of chief trainer. Organize the work of the trainers, allot concrete tasks, think up new ideas, and also ... play the match! Such work was new for me, and I do not think I was very successful with it.

The trainers conscientiously did their work. True, not all tasks were within their powers. The theoretical duel took place as on foreign territory. The openings which had to be analyzed were ones which practically none of my trainers play. They had to delve into positions about which for a long time they only had a very rough impression.

In addition, three of them are strong grandmasters. They frequently approached the evaluation of this or that position from the practical viewpoint: 'It is playable!' But I have a completely different approach. Even so, often they were able to 'persuade' me, but I am accustomed to finding the best continuation. And if I know that in a certain continuation the opponent is assured of an advantage, I do not go in for it. My colleagues think that this is possible. This is by no means a reproach, but demonstrates the different approach to chess which I had to encounter.

Coincidentally, in a recent Chessbase.com article titled Bisik-Bisik with Garry Kasparov (Part 1), a copy of the same book appears in the photo subtitled 'Garry Kasparov coaching Magnus Carlsen in his summer residence in Croatia'. Why is the former World Champion consulting his own book on his last World Championship match with Karpov?