GRADING

ECF News SCCU home page

updated 5.8.12
ECF GRADING
5.8.12
AUGUST UPDATE
A lot has happened since our last update. The January 2012 list duly appeared, with an encouraging input from the leagues for the inaugural halfyear effort. (Only about 10 per cent were missing, though it was unfortunate that the London League was one of them.) At least three of the absentees have decided to join the fold next year - yes, London League among them - and we hope it will be better than that. With the January list came the first of Michael Bennett's enhancements: complete game lists for all players, going back about ten years. Players have been able to pick up errors in the course of the season (mostly, we're glad to say, they have confined themselves to the last year or two). This has led to greater accuracy not only in the January list but also in the

ECF GRADING LIST JULY 2012
which appeared on schedule on 24th July. The list is at http://www.ecfgrading.org.uk/. Corrections are still coming in, but in terms of completeness it is probably the most inclusive list to date. The league results missing at the halfyear came in, and no expected event is missing.
The table below omits the January 2012 grades. We have retained the twelve-month spacing of lists for comparison's sake, and the (provisional) 2012 figures for players and halfgames graded represent the whole season. This year's increase in halfgames graded (about 12½ per cent) is in fact misleading, because 2011-12 was a 13-month year (see "New Developments for 2011-12" below). However, if the 2012 figure is adjusted to include the latest 12 months only, the resulting 308,590 halfgames still represents a very substantial increase of 9½ per cent.
For a table going one year further back, see last year's Grading page. Remember that grades up to and including 2008 are Old Grades. New Grades were introduced in 2009.

NOTE ON TEXT SIZE. If you've got your text set to Large, you may need to reduce it to make the table display correctly.

Halfgames graded Players active Players published	2006 270605 15650 11873	2007 270555 15384 11823	2008 275711 15274 11766	2009 271805 15135 11702	2010 273084 15037 11627	2011 282039 15227 11882	2012 317706 15768 12382
STANDARD Halfgames graded Players active Players published X grades A grades B grades C grades D grades E grades	201539 13089 10383 (113) 1799 (129) 1600 (120) 2422 (115) 2637 (106) 1925 ( 97)	207964 12955 10371 (112) 1902 (126) 1635 (118) 2382 (114) 2595 (105) 1857 ( 97)	208796 12720 10262 (111) 1912 (127) 1647 (116) 2449 (112) 2434 (106) 1820 ( 96)	207083 12582 10153 (134) 1973 (148) 1642 (138) 2405 (133) 2349 (129) 1784 (123)	205722 12489 10074 (133) 1938 (147) 1630 (137) 2453 (133) 2348 (127) 1705 (122)	206558 12404 10105 (133) 1956 (147) 1610 (136) 2443 (133) 2356 (126) 1740 (123)	224491 12643 10331 (133) 680 (151) 1535 (145) 1582 (136) 2342 (132) 2405 (125) 1787 (124)
RAPID Halfgames graded Players active Players published X grades A grades B grades C grades D grades E grades	69066 5627 3573 ( 72) 477 ( 84) 233 ( 99) 458 (103) 1211 ( 80) 1194 ( 50)	62581 5617 3545 ( 72) 91 ( 88) 274 ( 87) 221 ( 98) 416 (106) 1119 ( 86) 1424 ( 41)	66915 5826 3747 ( 72) 109 ( 87) 284 ( 86) 221 ( 96) 431 (104) 1297 ( 89) 1405 ( 39)	64722 5727 3712 (106) 90 (121) 317 (116) 346 (105) 653 (115) 1236 (112) 1070 ( 91)	67326 5840 3801 (107) 121 (115) 321 (117) 384 (104) 617 (112) 1257 (110) 1101 ( 96)	75481 6230 4134 (105) 112 (113) 393 (114) 426 (103) 678 (110) 1324 (110) 1201 ( 92)	93215 6903 4684 ( 98) 178 (111) 512 (104) 394 (108) 630 (117) 1650 ( 94) 1320 ( 86)

Figures in brackets are mean grades. "Players active" are players with at least one graded game in the most recent season.

2012: a breakdown by event type
See table on right. Numbers are halfgames. "Tournaments etc" means friendly matches, jamborees, county championships played over a period of time, anything that won't fit elsewhere.

The big increase in Congress grading, from 116,931 halfgames last year, owes much to the season's extra June.

League
Congress
Club internal
Tournaments etc
DM foreign games

Standard
118751
68524
25436
11520
260
224491

Rapid
7279
67465
8006
10456
9
93215

126030
135989
33442
21976
269
317706

Juniors published: 2869 (Standard 1094, Rapid 2487). Specifically Junior events: 58134 halfgames (Standard 12504, Rapid 45630).

27.12.11
NEW DEVELOPMENTS FOR 2011-12
This is a twice-updated version of an article that first appeared here on 21st May 2011. Text new in the latest version is in this colour. We repeat an important caveat: don't take this as an official ECF document. But it's written by a member of the ECF grading team and I hope it's accurate.
     Note, if you haven't aleady, that the online grading list has moved. It is now at http://www.ecfgrading.org.uk/. The old site is still available at the old address, but only for so long.

Twice-yearly Standardplay grades
The switch to twice-yearly starting this season was, of course, decided by the 2010 AGM in October. It brings Standardplay into line with Rapid, which has been twice-yearly since 2006-7. The most frequently asked question seems to be, how will it work? The obvious answer is right: same as Rapidplay does now (look it up in the online list's Help page).
     Rapid grades (naturally) work on a six-month cycle, with each half-yearly list using the previous list as its starting point. Standard till now has, of course, been twelve-monthly. Both sorts of grade go back a maximum of three years for their games, but there is a hidden difference. Rapidplay's three years, behind the scenes, is really six halfyears, with the categories (other than X and A) spaced two halfyears apart. It's easy to miss the fact that the three-year stretch, for January Rapid grades, starts half way through a season.
     Standard will now fall in line with Rapid in all respects. This will involve a change in the way Standard games are brought forward from earlier periods. If you need n games to make up your 30, you would expect them to be the n most recent games. And indeed they are, in Rapidplay. But the method currently used in Standardplay is somewhat more obscure (look it up), because Standardplay games aren't always properly dated. There's a major league that played all its games on 1st September last season, and a whole county that played all its games on 31st May. Anyway, the somewhat obscure method won't work any more when we go twice-yearly. (The reasons are technical, but even I understand them.) So, from next season, it will be "n most recent games" even for Standardplay. The number of leagues that don't date their games is tiny and dwindling, and with luck will soon hit zero. Actually their mantle is being assumed by the ever-increasing number of club internal games, which are rarely dated.
     What if a number of games share a date and you only need some of them to make up the 30? The system will take n games at the average performance for the date, as it used to with seasons. The games may genuinely share a date - three games on one day of a congress, for example (the system is not clever enough to work out that round 5 is more recent than round 4). Or, as often happens, a whole congress may be reported as though all its games had been played on day 1. The worst case is undated club-internal games, or 1st-September and 31st-May leagues while they exist. Clubs will be encouraged to split their games into (say) one-month blocks if they can't date them accurately. Where an event remains stubbornly attached to one end of the period, it can at least be moved bodily to a half-way point where there's a chance it will do less harm (isn't there?). That can be done centrally.

     When is the halfyear cutoff?
A few months ago you'd have got the obvious answer. End of November, as it already is for Rapidplay. Six months into the grading season. But there's an imbalance. End-November is nowhere near half way through the league season. So the grading season will be moved on by one month. That's official. The season will run from 1st July to 30th June (for both Standard and Rapid, naturally), with a halfyear cutoff at the end of December. It means starting off with a seven-month halfyear (ending 31st December 2011).
      This doesn't affect the timing of published grading lists. It will still be possible to produce a provisional list about the end of July and a revised one at the end of August in time for the start of the league season. Then the halfyear list can still come in January, if maybe late January.
      The end-December cutoff still doesn't capture quite half a season's-worth of league games. On a rough count in the larger leagues it captures anything up to 46% or so, depending on league, but adjournments and adjudications may reduce the proportion. (Did someone say "abolish them"?) The split isn't ideal, but it's as far as we can go and still maintain the schedule of publication.

     When will the 2011-12 halfyear list appear?
Late(ish) January. As a league grader has observed, an incidental advantage of an end-December cutoff, as opposed to end-November, is that graders won't have to do their thing over the Christmas period. Their reporting deadline has been set for 20th January, and the online list will take a little preparation after that (largely for checking). You probably wouldn't be far out if you guessed at a publication date of 25th January.

     How do we know results can be collected twice a year?
It's still too early to be sure. Congresses are no problem. Reporting of congress results has speeded up remarkably with the introduction of the one-month Grand Prix deadline. (All right, there have been exceptions.) Leagues are the question mark. Until now almost all have reported only once, after the end of the season. Reporting at the halfyear shouldn't be a problem, provided the league grader gets results with reasonable frequency so he can keep abreast. (I get them once a month with my Kent hat on, and that's sufficient. Other counties have clever electronic systems that should be better.) Graders' comments have been mostly supportive, and no one so far has said he won't be reporting at the halfyear. In fact a fair number of interim league results are in already. Check them out at http://www.sccu.ndo.co.uk/events1112.htm.
     If a league doesn't report at all till the end of the season its games will still be graded, but they will all go into the second halfyear. This will cause inaccurate January grades and, sometimes, inaccurate July grades as well.

     How will adjournments and adjudications be treated?
A game which is adjourned (or pending adjudication) at the halfyear cutoff obviously cannot be reported in the interim batch, and it will not go into the first halfyear's grading, even retrospectively. It goes in the second halfyear whatever date the grader attaches to it.
     The same will apply to any game from the first halfyear which slips the net for whatever reason and gets reported late.

     How will twice-yearly grades affect the way leagues operate?
Not at all, if they don't want it to. When Standard twice-yearly was first proposed, and quite heavily defeated, at the ECF's April 2007 Council meeting, a major objection seemed to be that the leagues would have no use for it and thought it would cause confusion. If so they'd changed their minds by October 2010, because twice-yearly went through with scarcely a murmur. The truth is that leagues can ignore January grades if they want. They can base eligibility (and maybe board order) on the August list throughout. This is what the ECF will do for its own season-long Standardplay competitions (including the national stage of the Counties Championships). The leagues are not bound to follow ECF practice, and each will need to make its own decision. The SCCU has decided on August grades.

     Will the change affect the grade categories?
Yes. Obviously Standardplay, like Rapid, will now have an X category based purely on 30+ games in the last halfyear. But, additionally, the opportunity will be taken to correct an anomaly with Rapid A grades. At the moment a Rapid A may be based on 29 games in the last halfyear plus 138 in the halfyear before. If its owner had played 1 game more in the last halfyear, he would have had an X grade based on 30 games only. It's nonsensical for an A grade to be based on more games than an X, but you can see how it arose. Correcting it would have meant different definitions for the two sorts of A grade. Now is the time. Starting in January, X grades will be the only ones to use an unlimited number of games. A grades (both Standard and Rapid) will be based on exactly 30 games in the last two halfyears. This brings A into line with the lower categories. For all categories below X, you count your 30 most recent games full stop. Assuming, of course, that 30 are available in the 36-month period. This was always the case with categories B - E; nothing is changing except the definition of A grades.
     Someone has actually suggested a new category, F you might call it, for 5 games in the 36 months (including at least 1 in the last halfyear). It would be an express route to a published grade, requiring just one congress. BUT it would not be used in calculations: its owner would be treated, in calculations, as ungraded. The new category is not official policy. But you never know, it might become so.

     How about those confusing junior categories?
Junior categories have had a mind of their own these last two years. You know that since 2009 juniors' published grades have not been used in calculations because they're liable to be grossly out of date; instead juniors are treated as ungraded, and a new "starting grade" is worked out for them. What some may have forgotten is that the categories changed at the same time. The 2009 policy was to count any number of games from the most recent 12 months, but otherwise to limit games counted to 9. It wasn't very satisfactory, and the 9 was soon revised to 20, giving the categories we've got now. (They're in the Help page, at least until it gets revised in January.)
     But did we need new categories in the first place? What if we had stuck to 30 and retained just the one set of categories? The experts have looked at this recently, and analysis suggests that changing back to 30 would have no obvious impact on the reliability of junior grades and might if anything be an improvement. It certainly has the merit of simplicity. So it will be done. Junior categories from January 2012 will once again be the same as adult. Juniors will still be treated as ungraded.

Online Enhancements
Or showing everyone's detailed game results, and crosstables and match results and things, à la Yorkshire and FIDE. It was hoped to have the foundations of this in place by the time the July 2011 list appeared. You will know by now that it has not happened. No one's to blame. Carl Hibbard, who has given freely of his time developing our splendid site from nothing and was keen to take it further, ran into pressure from other things. He has now, in fact, relinquished control of the online site. The site has migrated to a new (ECF-controlled) server and Carl's place has been taken by Michael Bennett of Hendon. Michael is to do the programming required for the enhancements. It can't be done overnight, and I don't know what the timescale will be.
      I won't deny that the delay has been a disappointment. Among other benefits the online enhancements will replace the ridiculous annual frenzy of emails sending the season's results to Direct Members. The Grading Administrator (me) undertook this year's June / July frenzy only with reluctance. 2200 emails, I think it was, and at that you're only hitting the Direct Members who have email addresses. It did pick up quite a lot of errors.
      The online developments won't be just for picking up errors. Far from it. But picking up errors will be a much more sensible operation when everyone can check their results online as the season progresses. Game results are less static than grades, so the online list will need to be revised more often than it is now. Once a month, say.

rjh 21.5.11, revised 20.8.11, 27.12.11

Previous years' information is in the Archive.

Back to ECF News Back to SCCU home page