English Chess Forum

Last year's junior grades caused lots of debate, and the formula used to calculate junior grades was changed in August 2009. This year's grades have hardly been commented on - possibly because people like myself have been on holiday. But also because we are now accustomed to the new juniors grades and the new grading method is giving realistically high grades for strong juniors (for instance those doing well in the British Championship).

I was concerned about the reliability of the grades for juniors not playing many games and have looked at the changes from 2009 to 2010 and compared them with 2007 to 08. For comparison I also looked at adult players. In all cases I only considered players who had both a present and a previous season grading. So in 2008 52% of the 7974 adults had a fall in Standard Play grade, whilst in 2010 49% of 7921 adults had a fall in grade.

For juniors in 2008, 20% of the 773 juniors had a fall in grade whilst in 2010 25% of did. So there has been an increase in the number of juniors with falling grades. Considering different grading letters I found that the increase was not restricted to grade E, there was also a significant increase in grade A juniors having a falling grade. (Even so, only 14% of Grade A juniors had a fall in grade whilst 42% of E did, showing that juniors who played lots of games were more likely to be improving.)

I also considered Rapidplay grade changes from Jan to Aug. For the 1470 adults only 30% had a fall in grade (over 45% for grade A/B) for both pairs of years. For juniors in 2008 only 14% had a fall in grade whilst in 2010 30% did. Again the increase was not restricted to less frequent junior players, though it was much higher there (mainly because in 2008 only 8% of grade E juniors had a fall in grade).

From this analysis my main conclusions are
1) that the new junior method is a definite improvement for regular (and improving) junior players, but it means that they are more likely to have a season where their grade falls.
2) for infrequent junior players, their rapidplay grade is likely to be more volatile with the new method but the old method was likely to give an unrealistically stable grade.

Comparisons of individual junior grades shows some inconsistency between grading order and playing strength - but that is bound to happen with any system.

I've been trying to make some sense of the Junior situation as well.

An interesting subset are those juniors listed as aged 17 in the 2009 list and appear in the 2010 list (now as adults). These are possibly just a few days too old to get a 5 point addition in the 2010 list (i.e. one would expect that as a group they were improving).

I have found 112 such records and in 81 cases their grade fell in 2010. On average grades fell 2.6 points. Those who appear as category A or B in both lists fell 2.3 points (adults in A or B in both years improved about 1 point). To me, this result is a suggestion the juniors may have been over-graded relative to adults in 2009.

The juniors in categories A and B (not just those aged 17) played about half the half-games attributed to juniors. These half-games are critical to the accuracy of all junior grades, since (I suspect) they include most of junior links to Adults within that set, a large number of games amongst themselves and the remainder. The remainder are (I suspect) the link to all other juniors grades. I have no pertinent data to investigate this line of enquiry further.

[quote="Brian Valentine"]
I have found 112 such records and in 81 cases their grade fell in 2010. On average grades fell 2.6 points. Those who appear as category A or B in both lists fell 2.3 points (adults in A or B in both years improved about 1 point). To me, this result is a suggestion the juniors may have been over-graded relative to adults in 2009.
[quote]

That does not mean the grades were wrong though - They are roughly in line with would I would expect as A levels will be the focus for most in that year rather than chess.

There are I suspect less problems with the new grading list as it includes a full years results rather than a half year and so treating juniors as new players (and thus disregarding the data source of previous years results) will have less of an impact as the grades will be based on 12 months rather than 6 months data.

David Shepherd wrote:There are I suspect less problems with the new grading list as it includes a full years results rather than a half year and so treating juniors as new players (and thus disregarding the data source of previous years results) will have less of an impact as the grades will be based on 12 months rather than 6 months data.

David - I think treating juniors as new players will have more of an impact as this means that grades will be more accurate!

David Shepherd wrote:

Brian Valentine wrote: I have found 112 such records and in 81 cases their grade fell in 2010. On average grades fell 2.6 points. Those who appear as category A or B in both lists fell 2.3 points (adults in A or B in both years improved about 1 point). To me, this result is a suggestion the juniors may have been over-graded relative to adults in 2009.

That does not mean the grades were wrong though - They are roughly in line with would I would expect as A levels will be the focus for most in that year rather than chess.

A levels are just the end of 3 years of important exams. My limited experience is that players who have stayed active (30+ games per years) into teenage years often reduce the amount they play in years 10 or 11, not waiting to year 13 (Upper sixth).

I think the main cause, as Brian implies, is the loss of the 5 point increment. So actually their raw grade went up by 2.7 points.

This graph shows the average (standard play) grade increases for juniors in 2009/10 by age and grade group.
It shows is that up to age 16 those who play most improve quickest, and by age 17 few players are improving rapidly (though 25% of 17 year olds did improve by 10 points or more).
Taking into account the loss of the 5 point junior bonus, 18 year olds falling by 2.5% fits in with this graph.

Edited to add axes labels

David Shepherd wrote:David - I think treating juniors as new players will have more of an impact as this means that grades will be more accurate!

I think it will have less of an impact on the 12 month data than the 6 month data. The main issue is that the data becomes more volatile the less of it there is. Thus I believe there will be less strange grades in the 12 month data than in the 6 month data.

The junior grades as a whole from what I have seen are now far more realistic. However a one off revaluation would have addressed that issue (providing it was done correctly), the test will be over a number of years to see if the juniors inflate or deflate compared to the adults and also a review of how stable individual grades are.

Neill, I thought you were a Maths teacher? If so, is it fair for me to pick you up on the (lack of) labelling of your axes?

Alex Holowczak wrote:Neill, I thought you were a Maths teacher? If so, is it fair for me to pick you up on the (lack of) labelling of your axes?

It is fair, and I have now corrected it.
Similarly I thought you were good on IT so please could you remove the broken link from your signature?

David Shepherd wrote:I think it will have less of an impact on the 12 month data than the 6 month data. The main issue is that the data becomes more volatile the less of it there is. Thus I believe there will be less strange grades in the 12 month data than in the 6 month data.

David - apologies if I've missed something. When did the ECF publish new style standard play grades using 6 months of data?

Neill Cooper wrote:

Alex Holowczak wrote:Neill, I thought you were a Maths teacher? If so, is it fair for me to pick you up on the (lack of) labelling of your axes?

It is fair, and I have now corrected it.
Similarly I thought you were good on IT so please could you remove the broken link from your signature?

Sure - it's because the event was two weeks ago, so the entry form had been removed...

Sean Hewitt wrote: David - apologies if I've missed something. When did the ECF publish new style standard play grades using 6 months of data?

Probably my fault in not being clear- what I was trying to say was that I would expect the 6 month rapidplay list to have more volatile grades than either the full year rapidplay or full year standard grades as the impact of ignoring last years grade after 6 months would be more significant as there was less data from the current year.

August 2010 ECF grades versus FIDE ratings:

Name Age ECF FIDE

Zhou, Yang-Fan 15 225 2326
Franklin, Samuel GA 16 219 2185
Ynojosa, Felix Jose 14 215 2229
Adair, James R 17 212 2212
Clarke, Brandon GI 14 208 2110
Salimbeni, George PW17 205 2159
Jones, Victor GL 15 200 2119
Jain, Akash 16 200 2097
Sen, Subin 17 200 2150
Grant, David J 17 198 2095
O'Toole, George E 17 197 2152
Jones, William EG 17 196 2115
Weaving, Richard 16 195 2027
Foord, Dominic 17 195 2071
Brown, Martin 17 194 2130
Williams, Peter A jnr 14 193 2179
Sathyanandha, Saravanan15 193 2020
Holland, James P 15 192 2074
Galliano, Alexander 16 192 2102
Hunt, Daniel 15 191 2035
Whitfield, Craig M 15 191 1960
Cumming, Rhys 16 191 2008
Boulton, William JK 16 190
Daley, Eugene E 13 189 1813
Weller, Jean-Luc 15 188
Lalic, Peter D 16 185 1951
Auckland, Elliott 17 185 1970
Merry, Alan B 13 182 1979
Walker, Samuel A 16 182 1945
Harvey, Marcus R 14 181
Bowler, James P 13 180 1672
Iyengar, Ilya 16 180 1991
Dines, Sheila J 17 180 1923
Scott, Gordon W 15 179 2026
Barr, Gabriel J 16 179
McPhillips, Joseph 12 177
Lenier, Jude A 14 177 2006
Foster, James M 16 177 1954
Patel, Shyamal C 16 177 1971
Sanders, Isaac B 12 176 1978
Zhang, Roy Y 13 176 1830
Craft, Isaac 15 176
Milson, Samuel A 16 176
Shaw, Dashiell L 16 176 1917
Maguire, Robert 17 176 1925
Rabbitte, Michael J 17 176 2010
Clements, Thomas A 16 175 1886
Wei, James 17 175
Haria, Ravi 11 174 1952
Vikanis, Georgs 15 174
Daggitt, Matthew 17 174 1831
Schofield, Oliver 16 173
Tambini, Jasper 17 173 1930
Thompson, Robert 17 173 1900
Thilaganathan, Jessica17 173 1888
Stepanyan, Henrik 14 172 1915
Taylor, Nathan J 14 172
Foo, William J 12 171 1876
Haxby, Felix 13 171
Taylor, Adam A 13 171 1821
Tunstall, George S 16 171 1934
Brijmohun, K Louis 16 171
Burke, Mitchell R 17 171 1898
Hughes, Robert C 17 171
Archer-Lock, Adrian 16 170 1844
Krishnan, Anand J 16 170 1767
Silverstein, Ben JC 16 170
Ynojosa, Angelica Maria17 170 1821
Broadley, Henry 12 169 1860
Stepanyan, Astghik 16 169 1823
Murphy, Conor E 11 168 1905
Wadhwa, Amarvansh S16 168
Yu, Jeffrey 16 168
Jones, Steven A 17 168 1801
Malhotra, Tarun 11 167 1913
Jain, Radha 13 167 1853
Lilburn, Tom 16 167
Soni, Sharan 17 167 1774
Mann, Harry T 17 167
Senior, Thomas 14 166 1864
Wang, Maria 14 166 1732
Qu, Xiyuan 15 166
Chadwick, Chris 17 166
Batchelor, Peter J 13 165
Davda, Rahil P 17 165 1853
Sirisena, Anuk N 17 165 1902
Fitzgerald, Robert C 12 164
Butler, Patrick A 16 164 1841
Gregory, Anthony D 17 164
Lin, Daniel 17 164
Andreev, Peter 11 163 1864
Garside, Andrew 14 163
Weller, Pierre 14 163
Periasamymanjula, M 15 163 1750
Rodrigues, Dunstan 16 163 1913
Talbot, Nathan L 17 163
Townley-Wells, S Nich17 163
Kennedy, James 17 163
Lawrence, David S 17 163 1796
Edwards-Wright, Miles13 162

Name Age ECF FIDE

Dines, Sheila J 17 180 1923
Thilaganathan, Jessica17 173 1888
Ynojosa, Angelica Maria17 170 1821
Stepanyan, Astghik 16 169 1823
Jain, Radha 13 167 1853
Wang, Maria 14 166 1732
Hollingworth, Evie 15 160
Hoare, Amy B 13 157 1921
Cass, Rachel 15 153
York-Andersen, Anna 17 153
Foster, Chantelle L 13 152 1701
Pritchard, Abigail 16 151 1563
Cotogni, Isobel 14 148
Wang, X Anna 11 147 1763
Appulingam, Kavitha 14 145
Shepherd, Katherine M12 144
Cleeves, Megan 13 139
Fozard, Francesca 14 139
Hahn, Lilli 17 138
Marsden, Ruby 16 135
Ehr, Jennifer 14 134 1480
Giles, Yasmin 11 133
Deshpande, Apurva 15 131
Milson, Amy F 14 127
Foster, Caroline 15 125 1539
Cass, Natalie 12 123
Bogoda, Nathasha O 13 123
Lip, Philomena 14 123
Bentley, Emma J 13 119
Hale, Stephanie MW 15 118
Davey, Rachel 15 117
Price, Phoebe A 13 115
Cooke, Larissa 17 115
Edhouse, Claire 15 114
Keen, Cosima C 11 112
Mason, Alicia 13 111
Elgar, Elizabeth 12 110
Steele, Beatrice H 10 106
Head, Louise 12 102
Grattage, Rebecca 16 102
Brace, Louise E 17 100
Mitchell, Florence EA 15 99
Stables, Imogen L 14 98
Wood, Nicola 17 98
Xu, Cathy 9 96
Carter, Amber M 10 94
Wen, Jessica 11 94
Hussain, Zara A 10 92
Barr, Sadie 10 92
Neil, Jennifer 12 92
Payne, Jessica 13 88
Sivakumar, Amy 12 86
Hapeshi, Eleanor 10 85
Perryman, Samantha 9 84
Swestun, Rebecca 10 84
Rose, Iona 14 84
Sung, Victoria 11 83
Buckley, Lara 12 82
Wijesinghe, Sachintha15 82
Sanders, Isobel H 10 78
Mathias, Irene 11 78
Scattergood, Ellen 14 78
Turvey-Cross, Imogen 9 74
Guilbert, Georgia K 12 73
Kalaiyalahan, Akshaya9 72
Dean, Angelica 14 72
Zelyk, Veronika 14 71
Barwick, Carmel 9 70
Narayanan, Nandaja 10 70
Purvis, Anna 9 69
Jina, Asha 9 68
Waddingham, Grace 10 67
Yass, Joanna 13 67
Mackenzie, Olivia 11 64
Johnson, Selina 14 63
Gilbert, Isabelle F 7 62
Strong, Zoe B 12 62
Greenhough, Amy M 14 62
Wiggins, Skye 9 56
Coupe, Elizabeth 13 55
Allen, Imogen 12 54
Sharma-Patel, Shefali 10 53
Petrie, Morven 13 52
Ramsey, Catherine 17 52
Towse, Lucy 11 51
Marsden, Florence 10 50
Wilson, Rebecca 12 47
Kanagaraj, Sobitha 12 45
Cordon, Alicia 13 45
Longrigg, Kate 16 43
Fitzpatrick, Angelica 11 41
Lobo, Kate R 12 36
Punnoose, Jenita 12 23
Owens, India 11 20
Bennett, Dawn 10 18
Jauhari, Shreya 12 3

Rather than go Christmas shopping I've tried to get an idea of how many of the top listed juniors and girls on the ECF grading list (not all English) have FIDE ratings. Any thoughts on those players who don't have FIDE ratings and the reasons for not playing enough international chess?

LozCooper wrote:Rather than go Christmas shopping I've tried to get an idea of how many of the top listed juniors and girls on the ECF grading list (not all English) have FIDE ratings. Any thoughts on those players who don't have FIDE ratings and the reasons for not playing enough international chess?

I know William Boulton, the highest graded non-rated player, would have had a rating by now. However, at the NCJS Young Masters, his last round pairing was fudged due to having a poor score for what would have been a second part-norm. Peter Purland thought it best to wait for him to get a rating, so that he could enter the rating system at a higher level.

I see what you mean:

http://ratings.fide.com/card.phtml?event=416681

Who was the arbiter?