20 January 2018

Two Important Chess.com Events

At the end of 2017, chess960 fans were in a holding pattern, waiting for a couple of Chess.com events to terminate:-
  • 2017-10-17: Three Chess960 Developments to Watch • Chess.com's 2017 Speed Chess Championship ('One chess960 game will be played in each time control at the end of each time period')
  • 2017-12-23: 2018 Fischer Memorial • 1st Chess.com Chess960 Championship ('you can win a ticket to the Bobby Fischer Memorial in March in Reykjavik')

Both events finished within a few days of each other. Here are Chess.com's final reports:-

Congratulations to both GM Carlsen and GM Maxime Vachier-Lagrave. Now that the events are over, where can we find the games?

Last year, when I discussed the precursor of the 'Speed Chess Championship' in GM Blitz Battle PGN (March 2017), I noted, 'To search on games, you need to know the players' names on Chess.com.' The full results of the event, including all preliminary matches, are detailed in 2017 Speed Chess Championship Schedule, Results, Information (chess.com), from which I copied the following chart.

The results of the each match include links to a full report. Here is a summary of the first round matches.

Nakamura 20.5 (Hikaru), Grigoriants 7.5 (sergiochess83)
Karjakin 19.0 (SergeyKarjakin), Meier 7.0 (GeorgMeier)
So 15.5 (gmwesley_so), Giri 14.5 (AnishGiri)
Grischuk 17.5 (Grischuk), Rapport 9.5 (Lordillidan)
Nepomniachtchi 15 (lachesisQ), Aronian 13 (LevonAronian)
Caruana 19 (FabianoCaruana), Hou 8 (yifan0227)
MVL 19 (LyonBeast), Xiong 12 (jefferyx)
Carlsen 20.5 (MagnusCarlsen), Guseinov 5.5 (GGuseinov)

In my next post, I'll try to gather all of the chess960 games from those matches. As for the chess960 championship, the 'MVL Wins' report provides the possibility to 'Download Tournament PGN'. In the next post, I'll also look at that file.

30 December 2017

Process Improvements

Continuing with Engine Trouble (September 2017; 'Talk about a disastrous tournament!'), in that post I described the background with a few sentences:-
The tournament was the final section of the LSS 2015 Chess960 Championship, a three-stage elimination tournament. [...] The 2016 final tournament starts soon, but I probably won't participate. I've taken enough psychological punishment for one year.

I finally decided to participate in the 2016 final for two reasons:-

  • It's not so easy to reach the final and this might be my last opportunity.
  • I couldn't do any worse than in the 2015 event.

I also decided not to upgrade my engines for this event, but I did introduce a number of 'process' improvements in the way I use engines. Specifically, I tried four new techniques that I'll discuss separately:-

  1. Castling manually
  2. Using two engines with different qualities
  3. Making first a null move, then applying opponent's expected move
  4. Using coarser granularity

1. Castling manually

In another follow-up post to 'Engine Trouble', The Seeds of Disaster, I noted, 'A chess960 opening can thus be logically divided into three phases': before either player has castled, after one player has castled, and after both players have castled. (The phrase 'player has castled' can also mean that a player has somehow lost the castling privilege by moving the King or by moving both Rooks.) In many games I've noticed some apparent inconsistencies in engine evaluation across these three phases. Was this another problem related to tapered evaluation as discussed in Chess Engines : Advanced Evaluation (September 2015)?

I decided to experiment by evaluating two identical lines. In one line I castled normally; in the other I castled artificially by moving the King and the Rook (in chess960, sometimes only one or the other moves) one square at a time, inserting null moves for the opponent, eventually reaching the castled position. I once used a similar technique in The Engines' Value of Castling (May 2015). I indeed recorded some differences, although it's too early to report on the results because the games are still running.

2. Using two engines with different qualities

The previous discussion is only relevant to chess960, but the remaining points are also relevant to traditional chess. I often use two (or more) different engines to evaluate the same position, then analyze any differences between them. This helps me understand unclear positions, especially where there is unbalanced material on the board. For example, I've often noticed that Houdini is better than Komodo at tactical play, but Komodo is better than Houdini at positional play.

I avoided using Stockfish for this sort of comparison because its search depths don't compare to the other engines. Roughly speaking, it takes Houdini and Komodo twice the time to calculate each successive ply, but Stockfish takes only 50% more time. I was reminded of this during the latest TCEC season, which I wrapped up with Houdini, Komodo, Stockfish, and AlphaZero. In a TCEC report from last month, Interview with Robert Houdart, Mark Lefler and GM Larry Kaufman (chessdom.com), the following exchange took place:-

Nelson [TCEC organizer]: What quality of your program do you think may be superior to your opponent in the Superfinal?

Larry [Team Komodo]: Basically, we have much more comparison with Stockfish because Stockfish is open source so we can easily compare our ideas and see what works better or worse. I don’t really know the inside workings of [Houdini], but what I can tell you is that my belief is that Komodo is better in most things than Stockfish. But there is something holding us back that has to do with search depth. We’ve been trying to figure it out for years, I don’t know what it is, but there is some reason we are not able to get the same search depth as Stockfish even if we tried to copy all their algorithms. We’ve tried experiments where we’ve tried to make Komodo act like Stockfish but it doesn’t work, and I don’t know why, but I feel that if we ever figure that out we’ll just be clearly #1. But almost every time we tried any idea from Stockfish in Komodo, nine times out of ten it makes Komodo weaker.

If the developers of a world class chess engine can't explain the phenomenon, what hope is there for the rest of us? It probably has something to do with pruning, as in Chess Engines : Pruning (September 2015). Long story short, I started comparing Houdini's lower-depth evaluations with Stockfish's higher-depth, but am not yet sure what I'm seeing.

3. Making first a null move, then applying opponent's expected move

It sometimes happens that no matter what move the engine proposes, the expected response by the opponent is always the same. When this situation occurs, I started using a technique where I first make a null move, then apply the expected response, then evaluate the resulting position. Any further analysis of the subsequent variations requires inserting a null move for the opponent. This should help to understand the tradeoffs for the immediate move.

4. Using coarser granularity

I also started experimenting with relaxed granularity. Engines normally return their (numerical) evaluations in units of centi-Pawns (0.01). I've often noted that it's shortsighted to favor one move over another simply because the first has a value of 0.02 and the second has a value of 0.01. This is even more shortsighted when one of the values is 0.00, which can mean all sorts of things. Any relaxing of strict numerical order is what engine developers call 'contempt' (see that Chessdom.com interview for more about the concept), but I started applying it to the moves suggested by the engines. I'll try to cover this in another post.

23 December 2017

2018 Fischer Memorial

What's the biggest single problem with chess960? There aren't enough games to report on! And there aren't enough games, because there aren't enough chess960 tournaments. The last time I reported on contemporary events was in (Not so?) Rare Birds, Summer 2017 (July 2017).

Sure, I could report on games from the many ongoing chess960 tournaments on the various online servers, but there is a certain element missing from online games. I'm not sure what it is exactly. Missing some tension? Missing some drama? Missing something else?

When the players travel to the same physical destination, meet face-to-face, and fight with the constraints of a physical clock, of visible spectators, and of real pieces that they grab in their hands, this all adds to a sense of permanence.

Events like the GAMMA Reykjavík Open 2018 – Bobby Fischer Memorial – March 6th – 14th don't come along very often. Although the tournament was held every two years starting in 1964, and every year starting in 2008, this year will see its first (official?) chess960 event (aka Fischer Random, aka Fischer chess, aka ..., we all know the routine). The site for the Reykjavík Open says on its main page,

The GAMMA Reykjavík Open 2018 will be a very special event. This time the tournament will be dedicated to the CHESS KING Bobby Fischer. An Icelandic citizen in his last years, Robert James Fischer was born on March 9th 1943. That means he would have turned 75 years old during GAMMA Reykjavik Open 2018.

His contributions to chess history will be celebrated during the tournament. Instead of the normal 10 rounds there will be nine rounds this year and a "free day" on March 9th. Then a special event will be organized – a Fischer Random championship with excellent prices.

The date for the chess960 tournament has extra meaning, because that would have been the day that Fischer turned 75 years old. The page for the chess960 event, European Fischer Random Cup 2018 (Free day), says,

1. ORGANIZERS: The Icelandic Chess Federation and the European Chess Union.

2. DATE, VENUE, SCHEDULE: The tournament is open for everyone, but the highest ranked European player will be European Champion in Fischer Random chess. The tournament takes place in Harpa (the playing venue of the GAMMA Reykjavik Open) on March 9th and starts at 13.00.

3. SYSTEM AND RATE OF PLAY: The Fischer Random Cup shall be played according to the Swiss System in 9 rounds. Time controls are 10 minutes + 3 second increment per every move.

4. RULES: The FIDE Chess960 Rules will be used. [with a link to the rules on Fide.com]

5. PRIZES: 1. €2,000; 2. €1,000; [...]

What was I saying about 'a certain element missing from online games'? Maybe I shouldn't have spoken so categorically; Win A Ticket To The Bobby Fischer Memorial (chess.com):-

By playing in the first Chess.com Chess960 Championship on January 4 you can win a ticket to the Bobby Fischer Memorial in March in Reykjavik and play in the Chess960 tournament that will be held on Fischer's birthday. [...] The Chess.com team loves Chess960, and we know the organizers well since we've provided the live broadcast of the tournament for several years now. So, we decided to join forces and organize a Chess960 tournament that is closely connected to the Fischer Memorial.

The mention that 'It's for titled players only' brought howls from the commenters and eventually a pseudo hashtag:-


I have to agree with that, although a comment from the author of the article, Peter Doggers, said, 'We had to decide to keep our Chess960 tournament for titled players only, at least for this year.' There's always next year! What will the format this year be?

It will be a nine-round Swiss with a time control of four minutes plus two seconds increment per move.

That's too fast for my taste, but I'll certainly report on the event after it's over. Let's add both the Reykjavík Open and the Chess.com Chess960 Championship to the events I mentioned in Three Chess960 Developments to Watch (October 2017).

25 November 2017

The Seeds of Disaster

In a recent post, The Seeds of Defeat (October 2017), I presented 'two games with Black that I lost without a fight'. In this post, I'll compare the corresponding games with the same start positions where I played White. In both games I'll start with the position where I castled; in both games this occurred before my opponent castled.

The point where one of the players castles (or otherwise forfeits the castling option) is often a good place to study the game. The previous moves have been pure chess960 (whatever that means), the following moves (up to the point where the other player castles) are a mixture of ideas from chess960 and from traditional chess, and the rest of the game will be equivalent to traditional chess. A chess960 opening can thus be logically divided into three phases.

The top diagram contrasts dramatically with the corresponding position in 'Seeds of Defeat'. Neither player has an advantage (which can be considered a moral victory for Black), although there is plenty of play in the position. The game continued 7...f6 8.Nc4 Nxc4 9.Bxc4 b5 10.Be2 O-O, where both players have castled; the position is still balanced, and there is still plenty of play. At this point I looked at three moves with different objectives -- a4, Nc5, c4 -- eventually choosing 11.Nc5. The game continued 11...f5 12.a4 b4 13.Bc4+ Kh8 14.Qh5 Ng6 15.b3 a5, and I was gradually outplayed.

Note that the start position SP230 is almost the same as SP518. The Rook on the a-file and the Knight on the b-file have been switched. This might have played a role in my thinking.

The bottom diagram also contrasts dramatically with the position in 'Seeds of Defeat'. Black's Queen and Bishops are more active than White's corresponding pieces, already giving Black the upper hand. Black played the natural move 10...O-O, and the game continued 11.Rc1 Nd5 12.Bb2 a5 13.a3 Bf6 14.Nd2 Bg6 15.Bf3 a4 16.b4 c5. Here Black has a space advantage and Black's pieces are better coordinated than White's. White must have played poorly and continued to be outplayed for the rest of the game, which lasted another 20 moves.

During all four games using the two SPs, I had trouble formulating a plan and was convinced that my opponents understood the evolving positions better than I did. The results -- four losses for me -- confirm this.

18 November 2017

Chess960 Showdown

If I had been maintaining this blog for the 18 months between July 2015 and December 2016, one of the events I certainly would have covered was the 2015 Showdown in St.Louis: Day 1: Nakamura - Caruana, Hou Yifan - Negi (chess-news.ru). It took place exactly two years ago:-
Two friendly matches started in St. Louis today. Hikaru Nakamura is competing against Fabiano Caruana, while Hou Yifan's rival is Parimarjan Negi. The GMs will play four different types of chess:

November 12 - Basque system (two games simultaneously, time control 90'+30");
November 13 - Fischer chess (four games, 20'+10");
November 14 - Rapid (four games, 15'+10");
November 15 - Blitz (eight games, 3'+2").

I once discussed a possible aspect of the Basque system (Day 1) on this blog in The More the Better (March 2012), 'seems like a natural way to conduct a chess960 tournament'. Regarding 'Fischer Chess' (Day 2), it's fitting that the word 'Random' was dropped from the Chess-news.ru report.

Kasparov Chooses Chess Positions for Showdown in Saint Louis (5:13) • 'Published on Nov 12, 2015'

The description for the clip said,

The Chess Club is running a poll on Twitter where followers may vote for the starting positions, selected by Garry Kasparov, of all four games of Fischer Random Chess during Showdown in St. Louis. These games will be played on Friday, November 13 at 1pm.

As for the chess960 games themselves, the Saint Louis Chess Club has an album on Flickr.com. Here is a composite image showing six of the photos:-

2015 Showdown in Saint Louis: Day 2

Chess.com's Mike Klein wrote a report about day 2 of the event in Big Swings As Nakamura, Hou Yifan Channel Inner Fischer (chess.com).

After day one's Basque Chess, the players shifted to four games of Fischer Random, also known as Chess960, played at the rapid time control of G/20+10. Nakamura dropped game one but took 2.5 of the next three against GM Fabiano Caruana, while the former women's world champion won three straight against GM Parimarjan Negi and missed a chance to possibly make it clean sweep, drawing game four.

Unfortunately, a technical glitch prevented automatic recording of the game scores: 'There are no PGNs for Fischer Random, due to the fact that the notation system cannot understand the castling rules.' That explanation doesn't make sense, but a comment to the post does: 'From what I understand about DGT, they can easily detect that a piece is on a square, but it would be much more difficult to detect which piece it is. So you'd have to enter the initial 960 formation somehow else.' Can DGT chessboards not track random start positions for chess960 games or were the boards not prepared properly? The blog Chess960 Jungle managed to record at least two of the games:-

The Basque and chess960 experiments were only used in the 2015 Showdown. The 2016 and 2017 editions of the event reverted to traditional rapid and blitz formats using SP518 (RNBQKBNR).

28 October 2017

The Seeds of Defeat

In a recent post, Engine Trouble ('Talk about a disastrous tournament!'), I listed five factors in a particularly poor result:-
  • Strong opponents
  • Insufficient engine power
  • Too many games
  • Fast time control
  • Too many vacations

Although none of these factors is exclusive to chess960 -- they can apply equally to the traditional start position -- at the end of that post I mentioned another factor that is exclusively chess960:-

I had several pairs of games -- the same start position against the same opponent playing White in one game and Black in the other -- where I reached uncomfortable positions in both games after 10-15 moves. That means there is something wrong with my approach to chess960 openings.

A few years ago, in May/June 2015, I wrote a series of five posts where I looked at five painful losses:-

The second post in that list ('Passive vs. Active') counted '[four] extenuating factors in the string of losses'. These match the first four above. Back in 2015, vacation didn't play a role, so at least I'm consistent with my excuses. Another characteristic of 'Losing Streak' was that 'in all five games I had Black'.

Looking at the more recent series of losses (+0-9=1, horrors!), I noticed two games with Black that I lost without a fight. The diagrams below show both games after the castling choices have been made. That is the point where the same techniques used to analyze traditional chess also apply to chess960.

In the top position, the players have castled to opposite sides. You can almost count the nine individual moves that have been made by both players to arrive at their respective positions. The game continued with a typical strategy for castling on opposite sides: both players launched a Pawn attack against the castled King. White eventually achieved an overwhelming attack while Black achieved nothing. After another 20 moves, Black resigned.

In the bottom position, White hasn't castled, but the King is in no particular danger. Black's King, however, is threatened by the Bishop on the long diagonal and the two Rooks on the open g- and h-files. These combined forces eventually won a Pawn, during which the Black Pawns on b5, c5, & d5 became increasingly vulnerable. The game lasted another 25 moves.

Since both games are essentially lost at the diagrammed position, an error must have been made before the diagram was reached. But where? In the third game of 'Losing Streak' above, analyzed in the post titled 'Imperfect Understanding', HarryO weighed in with a series of comments. The first comment was 'every move you played in the opening lacked initiative'. While that was undoubtedly an oversimplification, it still contained a big grain of truth. My approach to playing a chess960 opening with the Black pieces is:-

  • Pay attention to the intermediate goals of piece development, of the fight for the center, of King safety, and of keeping a healthy Pawn structure;
  • Avoid the quick, tactical knockout; and
  • Give the move back to the opponent to see what comes next.

In the two games shown above, I failed with the goal of King safety. It took more than 20 moves to realize this for certain, but the seeds of defeat were already planted in the diagrammed positions. At what point should an early fight for the initiative play a role?

21 October 2017

Three Chess960 Developments to Watch

The first development is potentially the most likely to fuel an increase in chess960 interest. Last week, in Chess960 battle: Nakamura vs. Carlsen?, Chessbase.com informed,
In February 2018, there may be an unusual exhibition match held between Magnus Carlsen and Hikaru Nakamura, playing chess960 (also known as Fischer Random). The competition is planned for the Hening Onstad Art Center in Baerum, Norway. It's funding is not yet fully secured, but Carlsen's manager is confident.

This relates to a blog post from this past summer, First the Non-routine News, (July 2017), that quoted Twitter: 'There are serious plans to organize a FischerRandom / Chess960 World Championship in Norway next year!'. As I reported a few months earlier in GM Blitz Battle PGN (March 2017), the same players met earlier this year in an event that included chess960: 'Chess.com report on the final match, Carlsen - Nakamura [...] Nakamura would go on to take 2.5/3 in the three iterations of chess960'.

The second development is an evolution of that GM Blitz Battle, another event organized by Chess.com: 2017 Speed Chess Championship Schedule, Results, Information (May 2017, but kept up to date as the matches are played). The final match will take place end-December:-

The 2017 Chess.com Speed Chess Championship features 16 of the world's best chess players in an innovative eSports bracket tournament. [...] Each 2017 Speed Chess Championship match will feature 90 minutes of 5/2 blitz, 60 minutes of 3/2 blitz, and 30 minutes of 1/1 bullet chess. One chess960 game will be played in each time control at the end of each time period.

The third development stems from a post last week on my main blog, Understanding Lombardy, where I wrote,

It turns out that the fastest way to understand Lombardy is to study his book, 'Understanding Chess: My System, My Games, My Life'.

Early in the book (p.21), Lombardy observed,

Sadly, Nimzovich died at 48, far too young for a chance to expand on his thinking and summarize that thinking into a true system, for his first work (My System, 1925) can hardly be considered a "system", but rather a collection of interesting games in which Dr. Nimzo offers advice on strategy and how to recognize and avoid mistakes. Yet his thoughts provided a foundation towards advancing chess strategy, even to my recommendation of the super-version of Fischer Random Chess!

This passage is significant for two reasons. First, I've often thought that Nimzovich would have been a brilliant chess960 player, given his penchant for unusual openings and deep strategical concepts. Second, given Lombardy's claim that 'I was Bobby's only chess teacher from [end-1954], and right through Reykjavik', any further thoughts on Fischer and chess960 are bound to be relevant. I'm slowly making my way through the book and will report any further discoveries on this blog.