Grade please?

[Simon Spivack]

... reminds me of a Winter's Trophy match (a sort of London League Cup, which anyone could enter

As I recall Winters existed: Len Winters, i.e. no apostrophe. The Winters Trophy was played in the Summer months. It has always been a Middlesex CCA contest, nothing to do with the London League. I believe it started sometime in the 1950s. It was open to non-Middlesex clubs, at the discretion of the controller.

[Andrew Bak]

Hello everyone on this forum, here is my debut post!

On the matter of grade, has anybody worked out a performance grade over a season when they play as white and one as black. Most people may have a preference to one shade or other but from a training point of view I suppose it kind of gives you an idea as to the style of chess you prefer (go for it, or come and get a slap). Furthermore, in I think chess for tigers this is developed into analysing openings that you play and the result, though at this macro level can be a little statistically misleading due the lack of range.

Last season I started to calculate my grade for each colour and I worryingly found a huge disparity, I was 173 with the white pieces in 25 games, and 151 with black in 33 games! I have a feeling that this was largely due to my tendency to blunder more often with the black pieces, but this still seems like a high difference to me.

This year (so far) it has been turned around! I'm playing 167 with white (17 games) and 171 with black (22 games).

I somehow think that at any level below say, 200 ECF, colour is much less of an important factor as there are so many mistakes (big and small) in a game anyway, so I don't take these grades by colour too seriously, but it is just quite interesting to see the difference.

[Richard Thursby]

On the matter of grade, has anybody worked out a performance grade over a season when they play as white and one as black. Most people may have a preference to one shade or other but from a training point of view I suppose it kind of gives you an idea as to the style of chess you prefer (go for it, or come and get a slap). Furthermore, in I think chess for tigers this is developed into analysing openings that you play and the result, though at this macro level can be a little statistically misleading due the lack of range.

This would only have much meaning if you played everyone as white and black. Otherwise, the result would be skewed by the number of games you played for each colour, and also the relative strength of players for each colour. In my first year of playing evening league chess I had eight games out of nine as black and scored 7/9. My 6/8 as black probably gave me a better performance than my 1/1 as white.

[Alex Holowczak]

At the end of the season, thanks to Carl's upgrades to the grading website, it should be possible for you to make the appropriate calculations yourself from the information provided there.

Unlikely, as that will require all graders to specify the players' colours (which will require the event organiser to tell the grader). Only about 3/4 of my graded games in the last 3 years have included colours and some of those were wrong.

I thought the colours were worked out already? Having looked at the grading files, you usually write something hashey with pins to identify. I would have thought the first in the row would be white, and the second black. Needs to be tweaked for leagues though, where that might not be true...

[Alan Walton]

Being a sad statistician I keep track of all my results

Last season I had and came out with a ECF grade of 193

White : 9.5/16 performance of 180.44 (effected by two awful losses against 150s)
Black : 14/18 performance of 202.67

This season I have the following performance (current expected grade of 212)

White : 8.5/9 performance of 222.11 (average opp grade 177)
Black : 8.5/12 performance of 204.92 (average opp grade 184)

I have started this year trying to break performance by opening, but currently the sample size is not large enough yet to be meaningful

I would normally expect my black performances to be a little lower, but I am in an unfortunate situation where if I play somebody over 40 points lower rated than myself it is costing me grading points even if I win, which is one of the problems with an annual list, more frequent lists would alleviate this problem though

[Jonathan Bryant]

This would only have much meaning if you played everyone as white and black. Otherwise, the result would be skewed by the number of games you played for each colour, and also the relative strength of players for each colour.

Well if that were true the grading process itself would have little meaning, i.e. players get a grade that makes no reference to the number of the games they played with white, strength of opposition etc.

Small sample size will likely cause unreliable results but if the sample sizes are large enough I don't see the problem. It's a bit like working out a tournament performance rating after an event.


Other ways you can break down grading performance include venue and or league.

E.g. I'm currently on +4 =3 -0 in the London League and +0 =0 -3 in Surrey/Croydon League matches (which I lump together because they're played at the same venue with similar time controls).

London League rating performance:- 181
Croydon/Surrey League rating performance:- 106



Too small sample size for definitive conclusion, of course, but I think I spot a possible pattern emerging!

[Geoff Chandler]

I've never asked for a players grade, I'm not one for chatting to an opponent
before a game and try to stay away from the board till the clock starts.

Once upon being asked this question before the start of a game I replied smiling;

"At this moment in time it's the same as yours...Good Luck."

I broke the rule about not conversing with an opponent when
we formed Sandy Bells C.C. and we had to start in the 4th Division.

In one of our first games where I out graded their board 1 by 800 points
my opponent wanted to walk away giving me the game by default
saying it was pointless us playing.

I actually had to give him a pep talk before he would play.
Then I beat him.

[Ian Thompson]

At the end of the season, thanks to Carl's upgrades to the grading website, it should be possible for you to make the appropriate calculations yourself from the information provided there.

Unlikely, as that will require all graders to specify the players' colours (which will require the event organiser to tell the grader). Only about 3/4 of my graded games in the last 3 years have included colours and some of those were wrong.

I thought the colours were worked out already? Having looked at the grading files, you usually write something hashey with pins to identify. I would have thought the first in the row would be white, and the second black. Needs to be tweaked for leagues though, where that might not be true...

Specifying players' colours is optional in a grading results file. If the grader doesn't specify anything then no assumptions are made.

[Tim Spanton]

I don't keep a record of grading/rating performances as White/Black but I do know that in around 2,500 recorded games, I have scored 46% as Black - about what would be expected - and 50% as White (around four percentage points less than expected).
Does this likely mean I fail to take advantage of the initiative White naturally has, or perhaps I am a natural counter-attacker? (As far as I know, I don't spend inordinately more time on studying my Black openings.)

[Christopher Kreuzer]

Being a sad statistician I keep track of all my results

Last season I had and came out with a ECF grade of 193

White : 9.5/16 performance of 180.44 (effected by two awful losses against 150s)
Black : 14/18 performance of 202.67

This season I have the following performance (current expected grade of 212)

White : 8.5/9 performance of 222.11 (average opp grade 177)
Black : 8.5/12 performance of 204.92 (average opp grade 184)

I have started this year trying to break performance by opening, but currently the sample size is not large enough yet to be meaningful

I would normally expect my black performances to be a little lower, but I am in an unfortunate situation where if I play somebody over 40 points lower rated than myself it is costing me grading points even if I win, which is one of the problems with an annual list, more frequent lists would alleviate this problem though

Interesting points. I should be able to do the same with my tally of results from the past three or so years, but I don't go as far as recording the openings (I should, though, as that will help more than lots of numbers).

One question I did have was whether it makes sense to calculate "colour" performance using your opponents' normal ratings? You have to do it this way (use the normal rating) as there is no other way at present, but shouldn't you be calculating all this using your opponent's "colour" grade? It might look like you are performing well with White (say), but if you take into account that your opponent might be under-performing with Black, then it might all even out.

[Gareth Harley-Yeo]

There's a clear dislike to being asked ones grade on this thread and in chess in general. Why? Why do people go out of their way to dodge it?

[John Moore]

It's often the "Don't you know who I am" thing which Jonathan Rogers was, I think, striving for earlier in this thread or "If you're that interested, it's on the bloody wall chart" (for Congress use only!)

[matt_ward]

I've encountered a recent game in which my opponent wanted to know what I was performing at this season.

I should of replied in a sarcastic mannor, "I think my game grade is slightly above my IQ about 75."

It was after the match had finished which does not bother me, but there is no benefit from my perspective of why your opponent should ask before the match.

What is your grade?

It can only be psychologically in favour for the player that has been asked the question.

Matt.

[Martin Benjamin]

There's a clear dislike to being asked ones grade on this thread and in chess in general. Why? Why do people go out of their way to dodge it?

Irrationally, I do feel the urge to apologise in advance in case the information proves to be misleading (completely unintentionally) one way or another. Two examples: (1) My current grade (192) reflects that last season was decent for me, but this year I have not been anywhere near as consistent. Someone graded in the 150s beat me earlier in the season and from the unconvinced look on his face afterwards, I almost felt like reassuring him by saying "I am not having you on, I really am 192 this year, despite the evidence of this game". I scrambled a draw against a someone in the 140s who also had the same look of surprise after the game. (2) Conversely, after I played a decent game to beat an IM a few years ago during a good run of form despite my grade having dipped to 166 after a poor previous season, I remember him defending himself against some fairly good-natured post game ribaldry from his friends by saying that my grade had made him lower his guard sub-consciously, and that he had not expected a 166 to play as well as I had.

I feel the urge to respond when asked this question by saying that like most of us I am capable of a wide range of performance on any particular day, often depending on family and work circumstances, so let's just play and do our respective bests. But it is easier just to say "192" than go through all the above!

[Sarah de Lisle]

I remember in the mid 1980s having 100 points added on to my FIDE rating. For being female. What on earth was all that about?

I then played in the British Ladies that year and just about lost them all in one go.....