British Chess Championships 2011

Debate directly related to English Chess Federation matters.
User avatar
Christopher Kreuzer
Posts: 8806
Joined: Fri Aug 06, 2010 2:34 am
Location: London

Re: British Chess Championships 2011

Post by Christopher Kreuzer » Wed Aug 17, 2011 5:00 pm

Still seems very silly to me. Even if, um, hypothetical. I do hope no chess players were disturbed...

LozCooper

Re: British Chess Championships 2011

Post by LozCooper » Wed Aug 17, 2011 5:10 pm

Christopher Kreuzer wrote:Still seems very silly to me. Even if, um, hypothetical. I do hope no chess players were disturbed...
Were disturbed or are disturbed? :lol:

User avatar
Christopher Kreuzer
Posts: 8806
Joined: Fri Aug 06, 2010 2:34 am
Location: London

Re: British Chess Championships 2011

Post by Christopher Kreuzer » Wed Aug 17, 2011 5:14 pm

Roger de Coverly wrote: I suspect you can load a chess engine to an ipod
Talking of which, are there chess engines you can use remotely over the internet?

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: British Chess Championships 2011

Post by Roger de Coverly » Wed Aug 17, 2011 5:55 pm

Christopher Kreuzer wrote:
Roger de Coverly wrote: I suspect you can load a chess engine to an ipod
Talking of which, are there chess engines you can use remotely over the internet?
I don't know if there are any public ones, but it seems possible to connect your portable device to your private home computer network. That's hardly a surprise, since it's equivalent to old fashioned dumb terminal to main frame, which has been around for forty years.

http://aartbik.blogspot.com/2011/02/che ... chess.html

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: British Chess Championships 2011

Post by E Michael White » Wed Aug 17, 2011 6:17 pm

Roger de Coverly wrote:
E Michael White wrote:For 2013 I think the ECF Board should be looking at OCRing the games in as soon as they are finished to create PGNs
It's a fairly obvious idea to try scanning a manuscript scoresheet to create a text file which can then be tested for valid pgn. That it isn't done in practice suggests the idea doesn't work. Experienced inputters can process a game very quickly, probably quicker than could be achieved by scanning and correcting errors.
Ian Kingston wrote:Attempting OCR on the average scoresheet would result in gibberish. Quite apart from the poor quality of handwriting of many players, people use a huge variety of idiosyncratic notations, such as short algebraic, long algebraic, colons to signify captures, algebraic but with descriptive notation for captures, no indication of captures, descriptive notation, non-English notation, clock times after every move, clock times at 10 move intervals etc.
Steve Rooney wrote: Exactly. Using OCR to get game scores is highly optimistic and likely to take longer than someone entering the moves whilst deciphering the gibberish.
Gentlemen
I wasnt thinking of standard OCR software being used in a two step methodology ie a) OCR b) check PGN for integrity.

The OCR modules need to have access to the AI of the PGN integrity checker all in one pass. This reduces the valid input alphabet and allows for using the next symbols(s) as further integrity checks. Thus the probability on information out takes in more information and should be more accurate.

This should be an M.Sc. project for someone ! I dont profess to know how education works anymore in this country but as McD seem to give out degrees why cant Chess and Bridge arrange something here ?
Sean Hewitt wrote:If you can get OCR software to read something like this I reckon you are doing very well indeed! ( attachment of unclear handwriting)!
Alex Holowczak wrote:That scoresheet gives me nightmares. :cry:

The handwriting appears to be continuous pen contact and therefore allows pen direction ratios to be used in addition to pattern displacement fuzzy logic. The results should be better; hard though it seems to be to believe it.

User avatar
Christopher Kreuzer
Posts: 8806
Joined: Fri Aug 06, 2010 2:34 am
Location: London

Re: British Chess Championships 2011

Post by Christopher Kreuzer » Wed Aug 17, 2011 6:32 pm

E Michael White wrote:The handwriting appears to be continuous pen contact and therefore allows pen direction ratios to be used in addition to pattern displacement fuzzy logic. The results should be better; hard though it seems to be to believe it.
Can such algorithms cope with changing handwriting patterns over the course of a game? I've seen people's handwriting get worse in time trouble, degenerating from moderately clear to an indecipherable scrawl. I've also seen people instructed by the arbiter to complete their scoresheet just scribble meaninglessly if they can't remember the moves. Admittedly, in these cases, humans don't do much better at deciphering the scoresheet. Possibly also of interest would be a computer that could reconstruct moves. i.e. give it two positions and ask it to find x moves that get from one position to the other one.

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: British Chess Championships 2011

Post by E Michael White » Wed Aug 17, 2011 6:54 pm

You would expect reliability to go down but not by much as the AI should have learnt from the first part of the game. For example in Sean’s, score sheet Blacks 1st move has three symbols so unless the 6 is wrong its either Nh6 or Nf6 but it clearly defines the shape of the C which in this gentleman’s notation is N. That makes Black’s moves 2, 3 and 6 and Whites moves 2,3,4,5 and 6 easier to understand which a two step system might find more difficult. In addition there would be fewer legal moves that match fewer symbols to select from as the game progresses.

IMHO looking at this area is much more sensible than cosmetic changes to the grading system of questionable value, that not everyone is happy with.

What has occured to me since posting is that the word Sunningdale would give the AI a free start as regards what (a,d,e,g) look like. Is this a good choice of location or is there anywhere else in the UK that Sean should choose to give more symbols than 4 out of KQBRNabcdefgh ?

Justin Hadi

Re: British Chess Championships 2011

Post by Justin Hadi » Wed Aug 17, 2011 9:45 pm

Adam Raoof wrote:
Roger de Coverly wrote:
Adam Raoof wrote: I was walking around the hall with my iPod / mp3 player checking my emails when I was told off....
Given the allegations against Feller and the French match captain at the Olympiad, you do not want anyone potentially in contact with the players, even if only by line of sight, to be receiving external communications.
I agree, and I can only offer in my defence the fact that the two players I was in communication with at that time failed spectacularly to haul themselves off the bottom boards of the Major Open...
Was one of them planning to get 5.5? :lol:

Sean Hewitt

Re: British Chess Championships 2011

Post by Sean Hewitt » Wed Aug 17, 2011 10:28 pm

E Michael White wrote: Is this a good choice of location or is there anywhere else in the UK that Sean should choose to give more symbols than 4 out of KQBRNabcdefgh ?
Queniborough?

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: British Chess Championships 2011

Post by E Michael White » Wed Aug 17, 2011 10:37 pm

Queniborough - thats good 5

I have since thought of :-

Badgeworth also 5

and

Buckfastleigh for a 7.

Sean Hewitt

Re: British Chess Championships 2011

Post by Sean Hewitt » Wed Aug 17, 2011 10:44 pm

E Michael White wrote:Queniborough - thats good 5
6 surely? Q e N B R h

Alex Holowczak
Posts: 9085
Joined: Sat May 30, 2009 5:18 pm
Location: Oldbury, Worcestershire

Re: British Chess Championships 2011

Post by Alex Holowczak » Wed Aug 17, 2011 10:49 pm

How about here, which would be 8? 8)

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: British Chess Championships 2011

Post by E Michael White » Wed Aug 17, 2011 11:43 pm

Sean Hewitt wrote:
E Michael White wrote:Queniborough - thats good 5
6 surely? Q e N B R h
The n and r in Queniborough are lower case, for piece names they need to be uppper case but the g counts also, so only 5.
Alex Holowczak wrote:How about here, which would be 8?
Thats only 3 :- abe, you cant count any twice and the n is lower case !
Last edited by E Michael White on Wed Aug 17, 2011 11:50 pm, edited 2 times in total.

User avatar
IM Jack Rudd
Posts: 4818
Joined: Tue Apr 17, 2007 1:13 am
Location: Bideford

Re: British Chess Championships 2011

Post by IM Jack Rudd » Wed Aug 17, 2011 11:47 pm

How about the Queenstown Road (Battersea) railway station, for Q R B a d e ?

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: British Chess Championships 2011

Post by E Michael White » Thu Aug 18, 2011 12:02 am

IM Jack Rudd wrote:How about the Queenstown Road (Battersea) railway station, for Q R B a d e ?
Yes - good but would anyone admit to getting an IM norm at the Queenstown Road (Battersea) railway station Masters A event ? It doesnt sound quite the same.

I've been poouring over the Maps of Germany and France for another 7 - Biergarten Bäldleschwaige.