Tuesday, 23 May 2017

Out, back and out again

As I get older (and more forgetful) I have a tendency to screw up my openings more and more. It is both a function of not learning all the lines, and playing careless moves without checking the consequences.
In the following very recent game, I played 7. ... Bf5 without much thought. After 8.Qf3 I was suddenly required to do a lot of thinking, but most of it was deciding whether to go berserk and sacrifice my queenside pawns, or eat crow and retreat the bishop. In the end I decided crow was the tastier meal, and retreated both the bishop and the queen. After that it was a battle not to get run off the board, bring out my pieces again, and try and salvage something from the game. Turns out I managed to find enough play to not lose, but all that post-blunder thinking left me short of time, and so a draw was offered an accepted.

Patterson,Miles - Press,Shaun [A09]
Autumn Leaves, 23.05.2017

Sunday, 21 May 2017

Pick the century

A challenge for readers of this blog. Have a look at the game below and decide which century it was played in (or which century it belongs to). I have of course removed anything that identifies the players, or where it might have been played.

White - Black [C37]
From a galaxy far far away

Saturday, 20 May 2017

The top 10

Where do you go to to get your chess fix (apart from Chessexpress)? According to one list chess.com is the most popular chess site, and the 1181 most popular site on the internet overall. Lichess is number 2, while Chess24 comes in third. The FIDE website is only ranked number 7, 2 spots behind chess-results.com
The full list is

  1. chess.com
  2. lichess.org
  3. chess24.com
  4. chessbase.com
  5. chess-results.com
  6. chessgames.com
  7. fide.com
  8. sparkchess.com
  9. chesstempo.com
  10. chess2700.com
So playing sites are the most popular, followed by news sites, and finally some training sites.

Wednesday, 17 May 2017

Attack of the clones

When I was a member of the FIDE Rules Commission we would occasionally discuss areas where the rules were silent. This wasn't because we felt the issue was too difficult to rule on, but more because we wondered what we would do if someone tried something really bizarre.
One topic was about playing more than 1 game at the same time. It started out as a method of avoiding defaults in team matches (ie could someone play boards 7 and 8 in the same round), and moved on to whether Kasparov could just enter the Olympiad by himself (playing a simul each round). We decided he could not (if the games were to be rated). There was also talk about whether a player could enter two sections of an event and play both at the same time, with a semi-famous case being Michael Adams playing a junior and open event at a British Championship early in his career. Again we thought it wasn't acceptable, in part because there was a risk that a player could 'transfer' information from one game to another, thereby violating the rule about analysing a game on another chessboard.
However the Denver Chess Club has decided to organise a tournament where players can play more than one game at once. The Clone Wars tournament allows a player to enter either as themselves, or to clone themselves once or twice. After that it is a normal event, except clones players are required to play two or three game each round. I assume you can't be paired against your clone, but your (or your clone) could play a different clone of a player you've already met. Whether you could play multiple games against the same opponent in the same round wasn't clear.
The event was run as a 4 round G/60m event, which I would assume gave players enough time to jump between boards (assuming you remembered which boards you were on!). The prize structure was also interesting, where a players total score (including clones) determined the payout (each point was worth a fixed amount). I don't know if the event was USCF rated, but I would assume that such event would not be FIDE rated (even with an eligible time control).
Here is a link to the tournament report, which contains a little more detail on the event than I can give you.

Tuesday, 16 May 2017

Better math through chess

I've just come across another study attempting to measure whether teaching chess in the classroom results in better learning outcomes for students. In this case the study looked at replacing 1 math lesson a week with a chess lesson (as opposed to adding an extra chess lesson). The study was carried out in Denmark, involving primary aged students.
Overall the study found a slight improvement in test scores (around 0.1 to 0.18 of a standard deviation), which at first might not sound like much. However, as these were replacement lessons, the result is in fact a lot better, especially if you are trying to get chess coaching into an already crowded calendar. Also of interest is that the study looked at the effects on children who were either unhappy or bored and found  that both these groups showed greater improvement than happy or engaged children. In fact most of the improvement in test scores was attributed to students in these groups.
The whole study is available here and is worth reading not just for the conclusions, but also for the description of the studies methodology. In describing quite clearly their approach, the authors not only help the reader understand their work, but also provide an idea of what to look for in similar studies on chess in education.

Sunday, 14 May 2017

Quality control

I was doing a little research for one of my correspondence chess games today, and I came across an issue that occasionally bedevils chess writers (and sometimes players). One of the games in a variation I'm playing was between Kaidanov and Kamsky, both very strong GM's, and therefore a game worth studying. The game itself followed theory up until move 14, when Black played the slightly unusual 14. ... Qe7. However it was his 16th move (16 ... Nh7) that was the real surprise, as it allowed the queen to be captured by the bishop on g5. Fortunately for Black it seems Kaindanov was feeling kind as the bishop retreated the d2 instead!
Of course the real story was that Black almost certainly played 14 ... Qc7 (which is theory) and only later moved the queen to e7 (on move 22). Kaidanov eventually won the game as White, and the mistaken move is quite clear, so the game may prove to be useful after all. However it is always worth double checking whether the moves make sense, as the risk is to blindly follow something that never happened in the first place!

Kaidanov,Gregory S (2640) - Kamsky,Gata (2645) [E75]
USA-ch Long Beach (8), 1993

Saturday, 13 May 2017

Stephen Fry explains the Dunning Kruger Effect

The Dunning Kruger effect is a cognitive bias where low-ability individuals overestimate their abilities. It has been known (formally) since 1999, although I am sure this effect was observed well before then, possibly under the heading 'to stupid to know they're stupid'.
For anyone unfamiliar with this effect, there is a short video (narrated by Stephen Fry) which explains it in the context of current American politics.

But what has this to do with chess? In this case not a lot, but it does relate to an observation I've made over the years. The biggest mistakes we make in chess don't happen when we don't have an answer to the problem in front of us, but when we (incorrectly) think we have the best answer to the problem in front of us. Often a game is lost because the move we thought that worked had a fatal flaw in it, and we would have been better off choosing a less flashy move. Usually this is described as over-confidence, which of course is a manifestation of Dunning-Kruger.
By the way, there is a flip side to the Dunning-Kruger effect. Often people who excel at a task don't realise that what they are doing is difficult for the average practitioner, and assume because it is easy for them, it must be easy for everyone. Anyone who has ever coached chess is probably aware of this (although maybe not consciously!)