15 January 2018

AlphaGo Netflix

A few days after writing last week's post, The Lineage of AlphaZero, I was looking to relax in front of the TV by watching Netflix, but what to watch? I started by browsing a list of recommended titles (or 'popular now' or something like that) and one of the first to appear was a documentary about AlphaZero's predecessor AlphaGo. Was that a coincidence or are the search engines also following me on TV? Whatever the reason, here's a Youtube trailer for the film.


AlphaGo Official Trailer (1:30) • 'Published on Sep 19, 2017'

The Youtube description says,

AlphaGo chronicles a journey from the halls of Cambridge, through the backstreets of Bordeaux, past the coding terminals of DeepMind, to Seoul, where a legendary Go master faces an unproven AI challenger. As the drama unfolds, questions emerge: What can artificial intelligence reveal about a 3000-year-old game? What will it teach us about humanity?

The documentary had only two stars out of five on Netflix, perhaps because it's at times somewhat tedious. The trailer shows many of the most exciting moments, but the full film was still worth watching. The 'legendary Go master' is Lee Sedol (or 'Se-dol Lee' as his name is sometimes written and which looks more Korean).

The film reminded me of the chess documentary Game Over: Kasparov and the Machine by Vikram Jayanti, which chronicled Garry Kasparov's 1997 loss to IBM's Deep Blue. Many years ago on this blog, I featured the trailer for that film in A Milestone in Computer Chess History (May 2007).

When I was still study in school, I learned how to play Go and played a few games with a friend. I understand a little about the game, but only a little. Extrapolating from my own experience with Go I can easily imagine how casual players of chess react to the chess documentary.

Let's have some reference links. First, here are a couple of pages from sources who want us to like the movie:-

And here are a couple of pages from sources who don't have a direct interest in the film, but want to know what we think:-

Finally, here are a couple of opinion pieces from a neutral source:-

Since this a 'Chess' blog, not a 'Go' blog (unlike the word 'chess', the word 'go' has to be capitalized to distinguish the game), this is probably the last I have to say about the subject. Future time would be better spent understanding the technologies that drive the AlphaZero family.

14 January 2018

Chess on a Bearskin Rug

Just like most years, the choices for Top eBay Chess Items by Price dwindled at the start of the new year, after the frenzy of the yearend holiday season. The painting below, titled '19C Signed V. Landoza Oil Painting of a Chess Game Figural Parlor Scene', sold for US $415.05 after 22 bids from 11 bidders.

The description was a list of the painting's attributes, of which the most important were:-

Artist: V. Landoza
Medium: Oil
Painting Surface: Canvas
Features: Signed
Size: Medium (up to 36in.)
Originality: Original

The painting might also be titled 'What's on a man's mind?', not so much for the chess as for the second painting behind the players, perfectly positioned between two of them. Was this the 19th century equivalent of a balloon in a comic?

While looking for more about artist Landoza, I discovered that the same(?) painting had sold twice in 2017 on Ebth.com: in March for $1550 and in December for $326. For an earlier example of this phenomenon, see a previous 'Top eBay Chess Items' post, From EBTH to EBAY (July 2017). NB: It might be worth looking at 'Top Ebth Chess Items'.

12 January 2018

Chess Horses for Courses

When I first saw this photo, it had zero views and zero faves. After I looked at it, it had one view and one fave.


Human-size chess game with actual soldier in St. Petersburg, Russia (1924) © Flickr user DailyLolPics.com under Creative Commons.

The description said,

The post Human-size chess game with actual soldier in St. Petersburg, Russia (1924) appeared first on Dailylolpics.

Shouldn't that be 'actual soldiers'? For more chess from Dailylolpics.com (NB: Not really 'For All Ages'!), see +25 Memes About Chess, although I only counted 14 images plus one on a separate 'Vertical Chess' page. The last (first?) meme is a short comic titled 'Chess in a nutshell'. It goes like this...

Pawn: Sire! The opposing forces have broken through our defenses!

King: What? How? We built an impenetrable wall!

Pawn: A horse jumped over our castle and landed on one of the priests.

King: Blast! I should have known. Have our infantry rush into enemy territory. Any who survive will become exact replicas of my wife!

...The punch line is 'Chess is weird.' (But we knew that already.)

[P.S. The chess board in the photo is located at Palace Square (wikipedia.org), St. Petersburg.]

11 January 2018

FIDE's Ethics Commission 2017

After last week's post about FIDE's Anti-Cheating Commission 2017, there remains one more topic flagged in Spectating the 88th FIDE Congress (November 2017): the Ethics Commission. In the 2016 post, Chess Ethics and the Suspensions of Federations (January 2017), I noted that details about the commission's decisions were henceforth recorded on the subdomain ethics.fide.com. What big ethical issues confronted FIDE during the intervening year? Three issues dominated the discussion during the Executive Board (EB) meeting in October 2017 (see the 'Spectating' post for links to the original documents):-

  • Cheating
  • Bulgaria
  • Kovalyov case

The cheating cases were 3/2015 (M Sandu against 15 other players for '"witch-hunting", namely a case of targeting a chess player in a public smear-campaign with accusations of cheating based upon fears and suspicions unsupported by any concrete evidence'), 8/2015 (FIDE against I Tetimov), and 2/2016 (FIDE against A Ricciardi). The Ethics Commission chairman, Francois Strydom, noted,

One of our cases was ground-breaking in the area of anti-cheating measures and for the first time we relied on Prof. [Kenneth] Regan’s calculations and his statistical method. He has to be thanked for his analysis to give us evidence for cheating. There was not enough behavioural evidence. He referred to the past debate with Deputy President [Makropoulos] whether there will be instances when only a computer evidence will be enough to convict someone.

The Bulgaria case is ongoing and related to last year's post on 'Suspensions of Federations'. The commission's report said,

Case 5/2016: Complaint by FIDE General Assembly regarding alleged fictitious transactions between the European Chess Union (ECU) and the Bulgarian Chess Federation (BCF) during 2011-2014 and the use of an imposter corporation named European Chess Union LLC -- whether BCF and/or any individual implicated -- Matter still in investigatory phase -- If prima facie case of wrongdoing against identified bodies or persons is found to exist, matter will proceed as a disciplinary proceeding for violation of FIDE Code of Ethics.

At one point in the EB discussion, Strydom summarized the main issues:-

What we are talking about is two categories of potential wrong-doers. The case of million of euros diverted to a false bank account. He said we identified these individuals and we are ready to proceed against them. They are the main culprits. The issue raised by Mr. Makropoulos is about the rest of the Bulgarian Board which turned a blind eye.

In a related action, among the Executive Board 2017 Decisions (fide.com) were 'To temporarily exclude the Bulgarian Chess Federation', and 'To approve that the players from Bulgaria to play under FIDE flag'. Specfic consequences of the decisions were announced a few days later in Bulgarian Players, Arbiters and Trainers (fide.com).

I haven't mentioned the Kovalyov case on this blog, so I'll first refer to a Chess.com report, Dress Code Incident At World Cup: Kovalyov Forfeits (September 2017):-

Anton Kovalyov today forfeited his game with Maxim Rodshtein in round three of the FIDE World Cup in Tbilisi, after being asked to change his attire. Upset about how he was treated, the 25-year-old Canadian grandmaster decided to leave the tournament immediately. [...] Ten minutes before the start of the round, Anton Kovalyov was approached in the playing hall by Chief Arbiter Tomasz Delega. The arbiter asked the player if it was possible to wear long trousers, instead of the shorts he was wearing, as this wasn't complying with the tournament's dress code. [...] Then Kovalyov was approached by chief organizer Zurab Azmaiparashvili, who told him that he should change his attire. When Kovalyov asked why, Azmaiparashvili replied: "Because you look like a gypsy!"

After an ethics complaint was raised by Canada against GM Azmaiparashvili, the commission made a snap ruling a few days before the EB meeting, FIDE Ethics Commission Statement in regard to the Kovalyov case (fide.com):-

This statement is released on occasion of the meeting of the FIDE Ethics Commission (EC) in Antalya, Turkey on 9 October 2017. Its purpose is to make known the EC’s decision in case no. 2/2017 and the reasons for the decision in the light of the wide-spread interest in the matter. [...] On 27 September 2017 the EC received a formal complaint against the organiser of alleged breaches of the FIDE Code of Ethics from the Chess Federation of Canada (CFC). The CFC asserted that it was acting "in and for itself and on behalf of GM Anton Kovalyov". The complaint was not supported by any statement by the player. [...] Without authority from the player, it is not possible for his federation to proceed with the complaint on his behalf. Also, the standing of the CFC to act as the complainant in its own capacity, is dubious. It is required that a complainant must have a direct and personal interest which was adversely affected.

Case no.2/2017 is now marked 'Case rejected'. This might be the fastest decision ever made by the Ethics Commission.

09 January 2018

FIDE Rating List - January 2018

Another January means another download of the FIDE Rating List (FRL). Ratings are an important part of organized chess and while some people watch the rating lists closely, I look at them once a year, always in January. Starting with the page Download FRL January 2018 (ratings.fide.com), I selected the same file as in previous years:-

Download STANDARD rating list
TXT format (08 Jan 2018, Sz: 6.58 MB)

Last year I managed to write two posts on the subject:-

This year I'll try to do more, even though there are always so many other chess topics competing for attention. The first 2018 FRL has over 296K players, of which over 134K are marked inactive. How does that compare with previous years? Like this:-

  • 2018: >296K players; >134K marked inactive
  • 2017: >265K players; >112K marked inactive
  • 2016: >231K players; >96K marked inactive
  • 2015: >197K players; >83K marked inactive
  • 2014: >171K players; >70K marked inactive

In my next post, I'll look at which federations are included, like I did for the 'Lower Rating Band' post.

08 January 2018

The Lineage of AlphaZero

Getting back to business after the yearend holidays, let's continue looking at the underpinnings of AlphaZero. In the previous post, The Constellation of AlphaZero (December 2017), I enumerated the underlying technologies. AlphaZero wasn't created by 'a bolt from the blue'; it was the latest evolution in a line of game playing algorithms -- Giraffe, AlphaGo, AlphaGo Zero, AlphaZero -- stretching back a few years. Each of those evolutions was introduced in a separate paper from which I give the abstracts.

2015-09: [Giraffe] 'Using Deep Reinforcement Learning to Play Chess'

This report presents Giraffe, a chess engine that uses self-play to discover all its domain-speci c knowledge, with minimal hand-crafted knowledge given by the programmer. Unlike previous attempts using machine learning only to perform parameter tuning on hand-crafted evaluation functions, Giraffe's learning system also performs automatic feature extraction and pattern recognition. The trained evaluation function performs comparably to the evaluation functions of state-of-the-art chess engines - all of which contain thousands of lines of carefully hand-crafted pattern recognizers, tuned over many years by both computer chess experts and human chess masters. Giraffe is the most successful attempt thus far at using end-to-end machine learning to play chess.

2016-01: [AlphaGo] 'Mastering the game of Go with deep neural networks and tree search'

The game of Go has long been viewed as the most challenging of classic games for artificial intelligence owing to its enormous search space and the difficulty of evaluating board positions and moves. Here we introduce a new approach to computer Go that uses 'value networks' to evaluate board positions and 'policy networks' to select moves. These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play. Without any lookahead search, the neural networks play Go at the level of state-of-the-art Monte Carlo tree search programs that simulate thousands of random games of self-play. We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks. Using this search algorithm, our program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0. This is the first time that a computer program has defeated a human professional player in the full-sized game of Go, a feat previously thought to be at least a decade away.

2017-10: [AlphaGo Zero] 'Mastering the Game of Go without Human Knowledge'

A long-standing goal of artificial intelligence is an algorithm that learns, tabula rasa, superhuman proficiency in challenging domains. Recently, AlphaGo became the first program to defeat a world champion in the game of Go. The tree search in AlphaGo evaluated positions and selected moves using deep neural networks. These neural networks were trained by supervised learning from human expert moves, and by reinforcement learning from selfplay. Here, we introduce an algorithm based solely on reinforcement learning, without human data, guidance, or domain knowledge beyond game rules. AlphaGo becomes its own teacher: a neural network is trained to predict AlphaGo’s own move selections and also the winner of AlphaGo’s games. This neural network improves the strength of tree search, resulting in higher quality move selection and stronger self-play in the next iteration. Starting tabula rasa, our new program AlphaGo Zero achieved superhuman performance, winning 100-0 against the previously published, champion-defeating AlphaGo.

2017-12: [AlphaZero] 'Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm'

The game of chess is the most widely-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.

To understand AlphaZero, it helps to understand AlphaGo Zero. Here is a useful video explaining some of the fundamental concepts.


How Does DeepMind's AlphaGo Zero Work? (10:52) • 'Published on Oct 27, 2017'

The description of the video says,

There's been way too many fear-mongering news articles around the latest version of DeepMind's AlphaGo. Let's set the record straight, AlphaGo is an incredible technology and it's not terrifying at all. I'll go over the technical details of how AlphaGo really works; a mixture of deep learning and reinforcement learning.

At one point, the presenter Siraj Raval shows an overview of the different AlphaGo evolutions.

Now that I have some understanding of the concepts behind AlphaZero, I can look a little deeper into those technologies.

07 January 2018

ChessBase India

Almost every time I assemble a short list for Video Friday (last seen in Chess News Hat Trick), there is at least one clip from YouTube's channel for ChessBase India (CBI), but the final selection for the post is always from another channel. Let's sneak one CBI clip into this next post for the series on The Sociology of Chess (November 2016). But which one?

I thought about featuring How much money would you take to quit chess forever? or Praggnanandhaa vs Firouzja: Clash of two future World Champions, but went instead with a video explaining corporate sponsorship. For many reasons, this is one of the weakest areas of chess in general.


Co-founder of a corporate giant explains why his company is supporting chess (3:32) • 'Published on Dec 31, 2017'

The description said,

Yatin Shah is the co-founder and executive director of IIFLW. He visited the 3rd Mumbai International tournament and spoke to ChessBase India about the event. In chess, where corporates think twice before supporting the sport, IIFL is backing the sport in a big way. Here's what Yatin has to say.

IIFLW stands for IIFL Wealth, and Wikipedia's page India Infoline explains,

IIFL (formerly known as India Infoline Limited), is an Indian diversified financial services company headquartered in Mumbai. [...] IIFL is ranked among the top seven financial conglomerates in India and as the top independent financial services firm in India in terms of market capitalisation.

In response to the question 'Why is IIFLW supporting chess in such a big way?', Yatin Shah says, 'Chess is a great sport' and 'Viswanathan Anand has a great fan following in India and kids are really inspired by him'. The main CBI site is at ChessBase India.