Sunday, 20 May 2018

Memorising endgames

Chess players learn/memorise openings, but don't really do the same for endings. This kind of makes sense, as there is no guarantee you will ever see a particular ending, but there are some that are common enough that committing them to memory would not hurt.
One classic example is the Rook and Pawn ending from the final game of the Capablanca-Alekhine World Championship Match (1927). It is a good example of how you convert and ending where you have an outside passed pawn on the queenside, while there are equal pawns on the kingside.
The key ideas are to put your rook behind the passer, forcing your opponents rook to blockade. Then bring your king towards the queenside, forcing your opponents king to try and keep your king out. Then shift your king to the kingside to attack the pawns. Finally, break up the pawns on the kingside with a pawn push of your own, before picking them off and winning!
While it takes a little time to complete, the general method is usually enough to collect the point. I've even had the need to use it recently, when playing some casual games at Street Chess.

Alekhine,Alexander - Capablanca,Jose Raul [D51]
World-ch12 Alekhine-Capablanca +6-3=25 Buenos Aires (34), 26.11.1927



Fast start, gentle finish

GM Wenjun Ju is the new Women's World Champion, winning her match against Zhongyi Tan 5.5-4.5. After a pretty violent start to the match (games 2-6 were all decisive), the game finished with 4 draws.
This of course was what Bobby Fischer had predicted was likely to happen in fixed length matches, with on player taking a lead, and then drawing their way to victory. Nonetheless, his proposed solution (first to 10 wins, but the challenger requiring a 2 win margin for the title) was never adopted, except in his 1992 match against Boris Spassky, The other solution, which was the first to 6 wins, was tried after 1972, but fell out of favour after the Karpov v Kasparov match that was aborted after 48 games. Since then World Championship matches have become shorter and shorter, making Fischer's prediction more likely to be correct.

Friday, 18 May 2018

An easy chess engine example

If you are interested in how chess engines work (and can read/understand Javascript). then 'A step-by-step guide to building a simple chess AI' might be worth a read. It is a simple explanation/tutorial about how chess engines are coded.
It  mainly looks at the evaluation and search functions, using the existing chess.js library for move generation and validation. As it is a very basic implementation, it is missing a few things that makes a chess program really strong. There is no quiescence search (a search extension which follows capture sequences beyond the specified search depth), no transposition table, and no move ordering.
However, if you are interested in tinkering with a chess program, the source is free and downloadable from the above links, and if you are feeling energetic, you can probably add those features yourself.

Thursday, 17 May 2018

Two Famous Actors

While flicking through 365 Chess Master Lessons by Andrew Soltis, I came across a game between two famous actors, Michael Redgrave and John Steadman. Looking a the date it was played (2007) and the venue (Sydney) I immediately figured something wasn't quite right, as Sir Michael Redgrave had died in 1985, and John Steadman passed away in 1993.
It turned out that the surnames of the players had been transposed, and the two participants were in fact Michael Steadman (NZ) and John Redgrave (AUS) who had met in the first round of the 2007 Sydney International Open. Soltis features the game as part of a lesson on when is the optimal time to play a move. In the game, 9.Nd5 was strong, and made even stronger by 9...Bb7??, although 9.a4 was even stronger, as it sets up some extra tactics for White. After Steadman found the knockout blow with 10.Ne6! there wasn't much left for Black.


Steadman,Michael - Redgrave,John [B94]
SIO, 2007



Tuesday, 15 May 2018

It just went horribly wrong

Trying to bluff your opponent in OTB chess is a risky proposition, doing so in Correspondence Chess is complete foolishness. Here I tried a bluff against my opponent, hoping he wouldn't find 9.Nh7!, or realise it had been played in a similar position previously (9...Qe8 is better). Of course he did find it, and after that, my game just disintegrated (aided in part by other horrible moves from me)


Schreuders,Arjo - Press,Shaun [C57]
Australia v Netherlands, 05.03.2018



Monday, 14 May 2018

The strongest player you've never heard of

While doing some research on the 1876 Steinitz - Blackburne Match, I came across mention of a player I'm not sure I've ever really been aware of, John Wisker. According to the Chessmetrics website, he was the 4th ranked player in the world at that time, with an historical rating of 2623.  This high ranking was probably based on his two victories in the British Championship, in 1870 and 1872 (the last Championship until 1904 btw). However his career was cut short in 1876 when he contracted tuberculosis, which resulted in a move to Australia (and possibly making him Australia's highest ranked OTB player ever!). I'm not sure if he played any chess while living here (I cannot find any games), but he wrote a chess column for the Australasian, before passing away in 1884 in Melbourne.


Wisker,John - Zukertort,Johannes Hermann [C80]
Zukertort 1st game in ENGWestminsterCC London, 22.06.1872


Sunday, 13 May 2018

Not playing doesn't just hurt yourself

A quick comment in the recent selections for the Australian Olympiad team (NB I am not revealing my selections, or who has been selected as there is still a chance of appeals by non selected players)

For the last few Olympiads I've been one of the selectors for the Australian teams. This year I was on the selection panel for the Open and Women's teams. One issue that arose for me was the geographical advantage/disadvantage some players suffered from. For a couple of players, the access to strong players was somewhat limited, making it harder for me to rank them highly. When they did play similar events to other players (ie international FIDE rated events), the results were quite comparable, but it was in their 'home' tournaments where they fell behind.
It wasn't because they scored badly, but simply because there wasn't enough strong players to test themselves against. And as playing in the Olympiad requires you to play against strong players, a 80% score against a field of 1700's isn't as impressive as a 50% score against 2100's.
Sadly, in at least a couple of cases, it isn't because there are no strong players close by, but that there aren't enough 'active' strong players close by. Now there are many reasons for choosing not to play (especially if you are a strong player), but it is becoming clear to me, that this has a knock on effect for other players. And I would hate to think that this would contribute to a cascading effect of discouraging the next level down from playing as well.