Chris Majer returns!

[Roger de Coverly]

Incidentally, has any amateur player ever completed what is (for some) the mammoth task of entering all their games into a database from paper scoresheets or books?

Back in 1990, I transferred all my game from scorebooks, the history going back to 1968. As far as I recall it took nearly all summer. Some of the games you remember quite clearly, complete rubbish you tend to forget. I took it fairly slowly since I noted what openings and systems I had been playing and when I had abandoned them for nebulous reasons.

[Jonathan Bryant]

I suspect you may have to use imbalanced time periods to get a balanced game count. Just guessing but would time periods of 1st June to 31st January (8 months) and 1st February to 31st May (4 months) give equal numbers of games?

Just remembered I didn't answer this.

That's probably right, although for non-summer tournament players like myself the first few months of the first period is dead time.


For the record, my breakdown for those dates would be:

1st June 2010 to 31 January 2011: 11 games
1st Feb to 31st May: 15 games so far plus maybe 5 more (some from the Surrey Individual which I don't usually enter).

[Paul Dupré]

Incidentally, has any amateur player ever completed what is (for some) the mammoth task of entering all their games into a database from paper scoresheets or books? I keep meaning to do this, but never find the time. I'm sometimes tempted to pay someone to do it for me... I wonder if OCR software has been rejigged to work for reasonably legible scoresheets (trying to forestall here the legion of stories of illegible scoresheets)?

Yes, I have - I in-putted 10 games a day into ChessBase. It took about 3 months. However, several year's later when I had a serious car crash in 2003, my wallet mysteriously disappeared as did the cash I was carrying and the following score sheets. Most were generously replaced.

Excerpt from my spreadsheet list of games:-
1265 Falzarano, Jim Sat 01-Feb-03 SCCU u175 Surrey Middlesex
1266 Welton, Jonathan Mon 03-Feb-03 Border League open Guildford B Crowthorne B
1267 Hannot, Jean-Charles C Mon 10-Mar-03 Thames Valley Div 2 Addlestone Hammersmith
1268 Rosen, Daniel B Tue 18-Feb-03 Surrey Trophy Guildford 2 Ashtead 1
1269 Lowe, William D Mon 24-Feb-03 Border League u160 Addlestone B Farnham B
1270 Jeffries, Majid Tue 25-Feb-03 Thames Valley Div 2 Addlestone Maidenhead
1271 Houska, Mario Sun 09-Mar-03 4NCL Division 4 Hilsmark Kingfisher Slough 3
1272 Duckers, Harvey L Mon 10-Mar-03 Border League open Guildford B Crowthorne B

The funniest reply though was from Bill Lowe who not only sent me a PGN of our game, but included all the moves he should have played as the main line. I had to point out to him that it was the score sheet I lost - not my memory.

[Angus French]

Below is the text of an email which was forwarded to me this evening:

From Chris Majer, ECF Manager of Grading and Rating
[email protected]

To all Leagues and Counties

ECF GRADING 2011-12: INTRODUCTION OF JANUARY STANDARDPLAY GRADES

As you know, the ECF AGM in October 2010 decided to switch to twice-yearly Standardplay grading lists, starting next season. This will mirror the twice-yearly Rapidplay lists which were introduced in 2006-7.

There will therefore be two grading lists in 2011-12: the usual July list, with a revision at the end of August timed for the start of the league season; and a fresh list in January based on the first half of the season. This has two important implications for leagues.

(1)
If the January list is to be accurate, leagues will need to submit results for grading twice a year instead of once. The graders know this. It is hoped that all leagues will be able to comply. If a league cannot do so and continues to send all its results at the end of the season, its games will still be graded but they will all go into the second halfyear, causing inaccurate grades in January and possibly July as well.

(2)
Leagues will need to decide whether the January grades are to count for purposes of eligibility (and perhaps board order). The ECF, for its own season-long competitions including the County Championships, will not use January grades for eligibility. It will use the start-of-season (end-August) grades throughout, so that a player who was eligible at the start of the season will remain eligible till the end. But other leagues are not bound to follow ECF practice; each will need to form its own decision.

If you need more information, please contact the ECF Manager of Grading – [email protected]

I'm surprised as I thought the proposal before the 2010 ECF AGM was for approval to investigate the possibility of twice-yearly standard play grading lists rather than for approval of their implementation. See here for the paper which supported the proposal. I note, in particular, the section on page three headed "What Happens Next" which states: The grading team will investigate the mechanism and methodology for implementing this change. The Manager of Grading and Rating will report back to the Board and then Council in April 2011.

[Carl Hibbard]

Happens Next" which states: The grading team will investigate the mechanism and methodology for implementing this change. The Manager of Grading and Rating will report back to the Board and then Council in April 2011.

It's complete news to me

[Roger de Coverly]

I'm surprised as I thought the proposal before the 2010 ECF AGM was for approval to investigate the possibility of twice-yearly standard play grading lists rather than for approval of their implementation. See here for the paper which supported the proposal. I note, in particular, the section on page three headed "What Happens Next" which states: The grading team will investigate the mechanism and methodology for implementing this change. The Manager of Grading and Rating will report back to the Board and then Council in April 2011.

The grading team have form for this sort of thing though. They took an authority to recalculate the grades as permission to revamp the approach used for juniors.

Certainly they had authority to consider the issues of six-monthly grades. Going ahead without publishing how they would overcome various practical and theoretical problems is pushing it a bit.

It's been prototyped to an extent with the rapid-play grades. Most rapid-play grades are tournament based and I'm not sure how many players are really bothered about their rapid-play grade.

[David Sedgwick]

Certainly [the grading team] had authority to consider the issues of six-monthly grades. Going ahead without publishing how they would overcome various practical and theoretical problems is pushing it a bit.

It's been prototyped to an extent with the rapid-play grades. Most rapid-play grades are tournament based and I'm not sure how many players are really bothered about their rapid-play grade.

The change is to take effect from January 2012. We've a feeling the grading team was asked to work on it, but there's no work to do. You just do it the same way you do Rapidplay. It is not certain, of course, that all league graders will be willing and able to provide halfyear results. The accuracy of the January list will depend on it.

(Richard Haddrell, in his report of the October 2010 Council Meeting on the SCCU website)


This seems to be the official position. Other than the risk that Leagues might not submit interim results, the "various practical and theoretical problems" suggested by Roger do not exist. Hence there was no investigative work to be done and nothing about which to report back. Moreover there was no need to do any kind of unpublished trial in January 2011. It's now full steam ahead for January 2012.

Time will tell whether or not this was a wise approach.

[E Michael White]

Yes it looks as if Mr Majer has made a premature announcement and decision for which he had no authority.

the "various practical and theoretical problems" suggested by Roger do not exist

Oh yes they do !

For one thing it will cause the grading system to deteriorate at a faster rate, but not twice the rate, requiring further corrective action sooner.

(....Richard Haddrell, in his report of the October 2010 Council Meeting on the SCCU website)

RH sometimes reports what he would like to have heared rather than opt for accuracy, this seems to be a case in point.

[Mick Norris]

My notes from the meeting say:

"Agreement was to proceed to twice yearly standard play grading lists (alongside the existing Rapidplay twice yearly lists) from January 2012
• Preparation needed now
• New players would get a grade much sooner, which would be a better incentive for them and more useful for league and congress organisers
• Manchester League could simply state that July/August grades would be used for the whole season for the purposes of order of players in Team registration lists, although later season board orders might change"

[Brian Valentine]

This seems to be the official position. Other than the risk that Leagues might not submit interim results, the "various practical and theoretical problems" suggested by Roger do not exist. Hence there was no investigative work to be done and nothing about which to report back. Moreover there was no need to do any kind of unpublished trial in January 2011. It's now full steam ahead for January 2012.

I agree with EMW and RdC and anticipate a long thread once the first January list arrives.

The most obvious issue is the addition of 10/5 points for juniors. This is said to compensate for a six month improvement on average over the following year. This is no longer the right period.

Other issues may or may not exist since the notes on how the system works miss out on some detail. It's not clear if juniors will be treated as ungraded based on their previous 12 months (or 20 games if in 6 months) or only over the last six months. Are the games in the previous 6 months, to bring the total up to 20, brought forward at their initial grade results or are they thrown back into the iteration process?

The players playing over 30 games a year provide some stability for the system in that their performance on average is closer to strength. There will be less of them playing 30 games in six months. The Cat A players are, on average, improving. My thinking is that each of their games will carry more weight in the system in future. In which case the change will tend to give rise to inflation as improvements will enter the calculations earlier.

I'm all in favour of more up to date grading, but am shocked if the graders are suggesting there will be no implications.

[Alex Holowczak]

I agree with EMW and RdC and anticipate a long thread once the first January list arrives.

There's always a long thread whenever any grading list comes out.

The most obvious issue is the addition of 10/5 points for juniors. This is said to compensate for a six month improvement on average over the following year. This is no longer the right period.

Other issues may or may not exist since the notes on how the system works miss out on some detail. It's not clear if juniors will be treated as ungraded based on their previous 12 months (or 20 games if in 6 months) or only over the last six months. Are the games in the previous 6 months, to bring the total up to 20, brought forward at their initial grade results or are they thrown back into the iteration process?

I think there are a lot of people who don't understand the grading process properly. Based on what you've written here, I think you count in this set of people. That is, you have fairly fundamental questions that have probably quite simple explanations, but there's no documentation of them online.

This page is now outdated with the switch to two lists per year. I'd be prepared to do a rewrite myself, but I'm in the aforementioned set of people too. I think there needs to be a better explanation, perhaps divided into three sections: (1) How to use the website (2) Basics of how the algorithm works (3) More detailed stuff would be useful.

[Roger de Coverly]

I'm all in favour of more up to date grading, but am shocked if the graders are suggesting there will be no implications.

Traditional theory on how to implement changes to computerised calculation systems would indicate the need for a parallel run.
So after processing the May 2011 files using the yearly approach, you then subdivide the data into before and after November 2010 and produce both a hypothetical Jan 2011 list and a parallel July 2011 list. You then compare the results of the actual with the parallel. If you get significant major differences, some rethink may be in order. You might even want to continue the parallel processing with the January 2012 list.

I think there are a lot of people who don't understand the grading process properly.

Given the secrecy which surrounds changes to the process and the somewhat arbitrary recent decisions made on how to process junior players, is that really a surprise?

[Alan Burke]

If the twice-yearly system is adopted for longplay and some leagues decide to send their results in for the January update and yet others don't; could a situation arise at some congresses in the latter part of the chess year (maybe even at the various events at the British Championships) where one person's grade has been calculated including his local league results, whilst another person's has not ?

[Roger de Coverly]

If the twice-yearly system is adopted for longplay and some leagues decide to send their results in for the January update and yet others don't; could a situation arise at some congresses in the latter part of the chess year (maybe even at the various events at the British Championships) where one person's grade has been calculated including his local league results, whilst another person's has not ?

Yes it could. Timing effects would mean that a Jan 2012 grade would be updated using results from 1st June 2011 to 30th November 2011, so for players not playing in summer events, very few "new" results would be included. The current rapid-play list doesn't suffer from this effect as there are only a handful of rapid-play leagues.


Part of the process might have to be to identify leagues not submitting half-yearly results so that players in those leagues could insist on more frequent submission of results.

[Steve Rooney]

If the twice-yearly system is adopted for longplay and some leagues decide to send their results in for the January update and yet others don't; could a situation arise at some congresses in the latter part of the chess year (maybe even at the various events at the British Championships) where one person's grade has been calculated including his local league results, whilst another person's has not ?

Yes it could. Timing effects would mean that a Jan 2012 grade would be updated using results from 1st June 2011 to 30th November 2011, so for players not playing in summer events, very few "new" results would be included. The current rapid-play list doesn't suffer from this effect as there are only a handful of rapid-play leagues.


Part of the process might have to be to identify leagues not submitting half-yearly results so that players in those leagues could insist on more frequent submission of results.

Will this make much difference to the grades? Even if there are few results for the 6 month period it will still be averaged over 30 games.

Leagues will presumably catch on with submitting grading results twice a year once the system is established.