Statistics in chess

Discuss anything you like about chess related matters in this forum.
David Williams
Posts: 336
Joined: Sun Mar 29, 2009 8:37 pm

Re: Statistics in chess

Post by David Williams » Thu Apr 20, 2017 5:15 pm

Christopher Kreuzer wrote:
David Williams wrote:
Kevin Thurlow wrote:"Moving on, we need to know very much more about participation by (a) ethnicity; (b) school background; (c) region; and (d) social class."

OK, I'll bite. Why?
Presumably so that we do not waste resources on a futile attempt to interest those who are never going to play, and concentrate on encouraging those sections of society that the game does appeal to.
That will surely entrench existing biases, rather than promote diversity. I suspect the person who originally suggested looking at this data, wanted to widen the appeal of chess, the opposite to what you are suggesting. None of the four factors listed should affect whether you play chess, given the opportunity, and where there are barriers to participation, those barriers should be broken down.
I've been waiting for ages for the barriers to my participation in activities popular with young women to be broken down.

Nick Burrows
Posts: 1705
Joined: Sat Aug 14, 2010 12:15 pm

Re: Statistics in chess

Post by Nick Burrows » Thu Apr 20, 2017 5:20 pm

Just squeeze into some skinny jeans and slip the bouncer a tenner :wink:

Jonathan Bryant
Posts: 3452
Joined: Sun May 11, 2008 3:54 pm

Re: Statistics in chess

Post by Jonathan Bryant » Thu Apr 20, 2017 5:23 pm

Alex Longson wrote:
I've been doing some digging looking at memberships and rating stats by age and gender in the UK and abroad with some interesting (if not altogether surprising) results. I'll post some findings here shortly if of interest to people.

One that stood out right away

ECF adult membership 97% male
ECF junior membership 81% male
You noticed this and yet you ignore the fact that much less than 1% of the membership is David Williams.

Who is the real oppressed minority here?

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

Re: Statistics in chess

Post by Alex Holowczak » Thu Apr 20, 2017 5:52 pm

Alex Longson wrote:ECF adult membership 97% male
ECF junior membership 81% male
What are the equivalent statistics in other sports? I'm not sure if there's a comparable measure.

David Robertson

Re: Statistics in chess

Post by David Robertson » Thu Apr 20, 2017 10:15 pm

Jesus wept. Another potentially productive thread, sunk without trace

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

Re: Statistics in chess

Post by IM Jack Rudd » Fri Apr 21, 2017 10:28 am

I can see at least two good motivations for collecting this data. One is the one already touched on - to see what markets we aren't tapping and try to identify ways to appeal to them. The other is to try to attract sponsors or advertisers based on what our members currently are.

AustinElliott
Posts: 660
Joined: Tue Mar 01, 2011 11:01 pm
Location: North of England

Re: Statistics in chess

Post by AustinElliott » Fri Apr 21, 2017 2:02 pm

Alex Longson wrote:I agree;

With a considered approach to data quality the ECF could make great strides here. FIDE also but that would require a larger coordinated effort across the federations. It wouldn't be too expensive for example for Experian (or other data company) to run a profile analysis of the membership - they would just need name, DOB and address which is presumably captured.

There is no D.O.B info on publicly available information on the ECF site (from what I can see) but there is on the FIDE ratings download. Average age of active ENG FIDE rated players is 44 (female 32, male 45) which is somewhat high (China 21 !) though I suspect this isn't representative of typical chess player profile!

If you know of any data scientists with an interest in chess (or just looking for a side project) please refer them (long shot I know)!
Sounds a worthwhile project, for the various reasons mentioned by others. One can already see from the figures quoted that the FIDE-rated population (serious tournament players) is (predictably) somewhat younger than the ECF-graded population (all club players who play graded chess).
Mick Norris wrote:Alex - Very good luck; if the MCF can help in any way please let me know

We have a lot of connections with Manchester University, so that would be an obvious place to start; when I was MCFPresdient, I was approached by someone there looking into something chess related, although of course I can't remember now exactly what :oops:
I'd hazard a guess the person that approached Mick might have been Dr Kieran Smallbone, chess player and mathematical biologist. Kieran was an author on the Bilalic et al. paper on women's participation in chess which has been mentioned a couple of times on this forum, incl. following Nigel Short's widely-reported comments on women players at the elite level in chess a couple of years back.

Kieran is no longer working at M'cr Univ AFAIK, though he is still around (and playing chess) in the NW for Macclesfield, as Alex probably knows.

I can't think of any NW chess-playing data scientists at the Univ, unless Martin Carpenter, who's already commented on this thread, would class himself as one. We do have a fair few League chess players amongst the staff, incl at least one Univ IT person with extensive database experience.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: Statistics in chess

Post by MartinCarpenter » Fri Apr 21, 2017 2:38 pm

Not quite sure what I am at times :) Definitely not strictly a data scientist. Kieran definitely wasn't working at the University last I talked to him about it.

As for comparable sports and gender balance, Bridge is much more balanced but serious bridge players probably aren't. German style board gaming is apparently quite close to 50/50 these days.

The other reason to collect it all would be to see just how bad the crunch that we're facing is going to be. Project the status quo forwards 15 years - are we looking at sustainable levels of contraction or a (near) wipe out?

Mick Norris
Posts: 10329
Joined: Tue Apr 17, 2007 10:12 am
Location: Bolton, Greater Manchester

Re: Statistics in chess

Post by Mick Norris » Fri Apr 21, 2017 4:11 pm

AustinElliott wrote:
Mick Norris wrote:Alex - Very good luck; if the MCF can help in any way please let me know

We have a lot of connections with Manchester University, so that would be an obvious place to start; when I was MCFPresdient, I was approached by someone there looking into something chess related, although of course I can't remember now exactly what :oops:
I'd hazard a guess the person that approached Mick might have been Dr Kieran Smallbone, chess player and mathematical biologist. Kieran was an author on the Bilalic et al. paper on women's participation in chess which has been mentioned a couple of times on this forum, incl. following Nigel Short's widely-reported comments on women players at the elite level in chess a couple of years back.

Kieran is no longer working at M'cr Univ AFAIK, though he is still around (and playing chess) in the NW for Macclesfield, as Alex probably knows.

I can't think of any NW chess-playing data scientists at the Univ, unless Martin Carpenter, who's already commented on this thread, would class himself as one. We do have a fair few League chess players amongst the staff, incl at least one Univ IT person with extensive database experience.
No I knew Kieran; it was January 2013 when I got this email
My name is Peter Backus and I am a Lecturer in Economics at the University of Manchester. I am currently working on a research project about how men and women compete with each other and we are using data from chess games to test some theories. Our initial results are interesting, suggesting that women are more likely to lose when they play against a man. I must confess, however, that I am, as are my co-authors, almost completely ignorant when it comes to the game of chess itself.
he had relatively recently arrived from the University of Barcelona (presumably for the better weather here)
Any postings on here represent my personal views

Kevin Thurlow
Posts: 5821
Joined: Wed Apr 30, 2008 12:28 pm

Re: Statistics in chess

Post by Kevin Thurlow » Fri Apr 21, 2017 4:55 pm

"Jesus wept. Another potentially productive thread, sunk without trace"

Evidently.

"Moving on, we need to know very much more about participation by (a) ethnicity; (b) school background; (c) region; and (d) social class."

OK, I'll answer my own question.

a) Irrelevant and possibly sinister - my club (and I hope others) would welcome 20 new players and wouldn't care about race, creed or colour etc. Why is ethnicity an issue? I have done some coaching at a local school (in leafy Surrey) and the one white player was French, and the non-white ones were from all over the place, including England.

b) Chess is difficult in secondary schools. Coursework takes up a lot of time and although the powers that be have finally realised that coursework is of limited value (as the pupils frequently don't actually do the work they submit), it is still there. 40 years ago, teachers might have voluntarily taken after-school clubs, now they don't as they are marking all the coursework, and they get sick of doing even more unpaid overtime. Nearly all pupils have computers and you can now play chess (and even more exciting things, like Countdown, Magic the Gathering etc.) online, so why play conventional chess? Where schools are doing chess, it's because they have someone interested in doing it. It is difficult to continue chess once you leave primary school.

c) London and the South-East has more activity. I live within half an hour's drive of half a dozen clubs. Visiting family in Somerset, I discovered there was about one club within an hour's drive. Doubtless, Jack can beat that! If you live near London, there's loads of chess, because there are lots of people.

d) Irrelevant. My club has had members who are bus drivers, dentists, scientists, company directors, shop assistants, teachers, etc., and of many political beliefs.

We should all want more people playing chess. Let's deal with facts and the real world.

40-50 years ago, people usually did 8-4 or 9-5 jobs and had spare time, now they don't. People now have computers and don't go out as much. Women must find that they are a small minority and many will doubtless be uncomfortable. Older players are continuing to some extent as they got in the habit of playing, and the supply of new players is dwindling. This is sad of course, but the solution is not immediately obvious.

AustinElliott
Posts: 660
Joined: Tue Mar 01, 2011 11:01 pm
Location: North of England

Re: Statistics in chess

Post by AustinElliott » Mon Apr 24, 2017 10:44 am

Sorry, more de-railing from the original topic, but:
Mick Norris wrote:it was January 2013 when I got this email
My name is Peter Backus and I am a Lecturer in Economics at the University of Manchester. I am currently working on a research project about how men and women compete with each other and we are using data from chess games to test some theories. Our initial results are interesting, suggesting that women are more likely to lose when they play against a man. I must confess, however, that I am, as are my co-authors, almost completely ignorant when it comes to the game of chess itself.
he had relatively recently arrived from the University of Barcelona (presumably for the better weather here)
I looked up Dr B and found that they had published a version of their chess work (or at least put it on a web repository):
Abstract

There is a growing literature looking at how men and women respond differently to competition. We contribute to this literature by studying gender differences in performance in a high-stakes and male dominated competitive environment, expert chess tournaments. Our findings show that women underperform compared to men of the same ability and that the gender composition of games drives this effect. Using within player variation in the conditionally random gender of their opponent, we find that women earn significantly worse outcomes against male opponents. We examine the mechanisms through which this effect operates by using a unique measure of within game quality of play. We find that the gender composition effect is driven by women playing worse against men, rather than by men playing better against women. The gender of the opponent does not affect a male player’s quality of play. We also find that men persist longer against women before resigning. These results suggest that the gender composition of competitions affects the behavior of both men and women in ways that are detrimental to the performance of women. Lastly, we study the effect of competitive pressure and find that players’ quality of play deteriorates when stakes increase, though we find no differential effect over the gender composition of games.
It is actually an interesting read in the context of the ongoing discussions (other threads) re. women chess players. Their headline conclusion is that woman players in their sample (a big sample of players rated FIDE 2000 or better, with an average rating of 2350) performed"30 FIDE rating points worse on average" against male players than they did against other female players.

I think the authors must have found some expert chess-playing help (perhaps among their Spanish/Catalan or Slovenian co-authors or friends?) as the chess-related bits are sensibly done. E.g. the data they analyse are games taken from TWICs PGN files:
We use the PGN files published by TWIC for 2012 and the first six months of 2013 giving us information from 79,242 games played by 14,056 players from 154 national federations
- and the exclusions they use for the analysis (players rated <2000, games <15 moves) are also logical in a chess context. Finally, their "unique measure of within game quality of play" is a statistic they compute by comparing the moves a player makes in the middle-game (moves 15-30) with the top engine suggestions, in some ways a little bit reminiscent of what the anti-computer assistance detectors are doing. Anyway, an interesting contribution to the 'chess players and gender' discussion.

Alex Longson
Posts: 27
Joined: Fri Nov 11, 2011 1:36 pm

Re: Statistics in chess

Post by Alex Longson » Tue May 02, 2017 2:38 pm

Thanks for all the useful contributions. Apologies I have only just caught up with the thread. There are a number of interesting directions here which are worthy of consideration.

I won't quote directly but I'll pick up on one or two comments.

League chess and participation dying down. I believe this is more an indication of the lack of popularity of league chess (amongst millenial/"gen Z" segments) than evidence that chess is declining in popularity. In fact it seems to me that chess is generally in very good shape in global terms when you factor in age demographics of tournament chess, new online capabilities and the various sub-cultures / organisations this is spawning. So I don't think we need to mourn the demise of league chess, but rather embrace new competetive (and non-competetive) structures.

How do millennials and 'gen-z' view the game? What strikes me is that many crazes/fads come and go (Pokemon Go for example) and yet chess I believe retains its appeal as a non-trivial intellectual pastime. Furthermore, there is significant infrastructure in place in terms of organisation and history that will enable chess to continue to prosper - though again I believe it is important to embrace the changes.

My wife (who is Director of Women's chess for ECF) and I tried to find some proxies for the male/female imbalance in sports, education and chess in other nations. For example the lowest proportion of female students in university subjects is 'technologies', 'engineering' and 'computer science'. England rank 33 / 36 in terms of the top nations (by number of total registered players) for the female gender split of junior players 11.6%. To be fair this wasn't far off the likes of USA, France and Germany. Park Run participation is roughly 1/3 female. Does any of this matter? Not sure - but some fun facts to begin with!

I'm an optimist regarding chess - but mainly I'm interested in how things work and firmly believe that a data led approach combined with an experimental approach to change can drive considerable improvements.

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

Re: Statistics in chess

Post by Roger de Coverly » Tue May 02, 2017 3:24 pm

Alex Longson wrote: In fact it seems to me that chess is generally in very good shape in global terms when you factor in age demographics of tournament chess, new online capabilities and the various sub-cultures / organisations this is spawning.
Globally yes, but not in the UK. However it may not just be chess, local leagues in activities such as Cribbage, Darts, Snooker, even Bridge and Table Tennis are struggling with declining numbers of participants.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: Statistics in chess

Post by MartinCarpenter » Tue May 02, 2017 4:16 pm

I'm not even sure if its specifically the UK - Northern UK chess might well be worse off than in the south.

You'd obviously have to see hard statistics to be sure.

Alex Longson
Posts: 27
Joined: Fri Nov 11, 2011 1:36 pm

Re: Statistics in chess

Post by Alex Longson » Tue May 02, 2017 9:42 pm

Roger de Coverly wrote:
Alex Longson wrote: In fact it seems to me that chess is generally in very good shape in global terms when you factor in age demographics of tournament chess, new online capabilities and the various sub-cultures / organisations this is spawning.
Globally yes, but not in the UK. However it may not just be chess, local leagues in activities such as Cribbage, Darts, Snooker, even Bridge and Table Tennis are struggling with declining numbers of participants.
Hi Roger - do you think we are just witnessing a natural phasing out of one form of organised chess and being replaced by other forms?

Regarding other activities - it is clear there is more 'choice' now in terms of pastimes (particularly due to online entertainment) but as chess so naturally lends itself to the online form perhaps it is faring better than some of it's substitutes?