ECF to scrap its grading system

[J T Melsom]

Alison Bexfield

Please don't be discouraged by this conservative response. There are good reasons for caution but personally I welcome the challenge to change. I'm a little unsure why Go is more relaxed about these issues around reliable/meaningful ratings. But then again Go has a rather more fun approach to prizes, I loved the idea of a homemade pot of jam as a prize - better than grumbling over the lack of cash prizes for weaker players. On the other hand I would imagine that the more limited range of opportunities to play a Go event and the more inclusive nature of competitions (the way you run swiss events) that cause the 'graders' to be a bit more relaxed about the accuracy of grades. Please keep challenging chess players to consider lessons from other minority interests though I suspect it is the sociability that attracts Go players rather than early access to ratings.

[Roger de Coverly]

This is what the Minutes of a Board meeting have to say on the subject

a) Monthly Grading
–
Dave presented his note on monthly grading which provided a summary of Brian Valentine’s paper. Brian feels that it may be expensive to upgrade the system to cope with monthly grading and it would also be important to promote the idea with leagues and local graders.
It was agreed that timely grading of players was a core function of the ECF and there was a clear demand from many sections of the chess community for more frequent grading lists. The Page 3 of 7 Board's target was to provide monthly grading capability for the 2019/2020 season at the latest. Mike said that he hoped that the ECF would be able to work jointly with UKCC, who are thinking about developing a grading system, and that it would be regrettable if other organisations felt obliged to develop separate grading systems in the absence of a suitable service from the ECF.
Julie stressed the importance of ensuring that local graders were supported as part of any system changes.
The Board were concerned about the cost implications of building a system from scratch, and asked Dave to look at grading systems in use by other organisations– e.g. USCF, FIDE and the Correspondence Chess Organisation(s) - with a view to possibly purchasing off the shelf or licensing an existing system.
An action was agreed for David to review the various options including discussions with a USCF representative and discussions with Alex and Sarah from UKCC, and report back to thenext board in January.

I cannot help thinking that there's woolly thinking going on.

Surely you must first define how the grading system will work, and then seek to a means to implement it.

So if you decide to copycat the USCF approach, you are adopting their theory. Similarly with FIDE and that's not the same theory as USCF.

The FIDE design in particular works badly with rapidly improving players. But perhaps it depends on the objective. If you want a reliable ranking system, that's one thing. If you just want a set of random numbers published very frequently, that's another.

[NickFaulks]

The FIDE design in particular works badly with rapidly improving players.

There is no reason why this should be the case for an ECF Elo based system, since presumably their results would be reported. That is the weak point with English FIDE ratings.

[Roger de Coverly]

There is no reason why this should be the case for an ECF Elo based system, since presumably their results would be reported. That is the weak point with English FIDE ratings.

What about someone who plays most of their chess and most of their improvement in on-line chess, or any chess not captured by the ECF system?

Most national Elo systems have had to develop rules of forgetfulness so as to be able to discard results no longer relevant to estimates of current strength.

Here's an example relatively close to home.

https://www.chessscotland.com/grading/h ... ion-steps/

200 Up Rule

Fast improving players can quickly find that their published grade is out of date and understates their current strength. Opponents of the improving player are victimised because expected scores are calculated against a published grade which has proved to be inaccurate. To quickly boost improving players to a more appropriate level and avoid their opponents unjustly losing grading credit the 200 up rule was invented.

(Modification July 2001) If a grade goes up by 200 points on the first of the (Nov 2006) final grading runs it will be set to zero and recalculated as if the player was an unrated player. The temporary ratings thus calculated will replace the old published grade. Grades will then be recalculated and opponents will get credit for the temporary rating rather than the old published grade. (The junior addition is deducted from the temporary grade before it is slotted into the 200 up player’s record – since opponents will get the temp rating plus any addition when the grades are recalculated.)

The Scottish system also has a table of junior additions.

[NickFaulks]

What about someone who plays most of their chess and most of their improvement in on-line chess, or any chess not captured by the ECF system?

Can you give an example of such a creature?

[IM Jack Rudd]

Well, any foreign junior who plays the occasional ECF-graded event, for one.

[Roger de Coverly]

Well, any foreign junior who plays the occasional ECF-graded event, for one.

Look at the rapidplay ratings of Indian juniors for example where their rapidplay rating can be 800 points short of their normal one.

http://ratings.fide.com/card.phtml?event=25059530

By contrast, the ECF grades
http://ecfgrading.org.uk/new/menu.php?f ... de=307156L

That's under the current system, where every time he plays in the UK, he is treated as a new player.

[NickFaulks]

Well, any foreign junior who plays the occasional ECF-graded event, for one.

I don't think it is reasonable to demand that something designed as a national system should deal well with visitors from overseas.

Does anyone care what Magnus Carlsen's ECF grade is?

[Roger de Coverly]

Does anyone care what Magnus Carlsen's ECF grade is?

Provided it correctly ranks him above Adams, probably not.

But at lower levels, the FIDE ratings can give such misleading rankings as to reduce some tournament pairings to random drawings, if you trust the rankings driven by ECF grades as more reliable.

The recent Telford U-2050 tournament also imposed an Under 180 condition of entry, but used FIDE ratings for seedings. The consequence was that the two highest ECF graded players at 179 met in the first round. Both eventually finished second equal.

[IM Jack Rudd]

The thing about visitors from overseas is not that we care about getting their grades right for its own sake, but that we care about getting their grades right so that English players get appropriate credit for ECF-graded games against them.

[Kevin Thurlow]

"What about someone who plays most of their chess and most of their improvement in on-line chess, or any chess not captured by the ECF system? "

I would have thought that covers a lot of players.

"It was agreed that timely grading of players was a core function of the ECF and there was a clear demand from many sections of the chess community for more frequent grading lists."

Many?

"Mike said that he hoped that the ECF would be able to work jointly with UKCC, who are thinking about developing a grading system, and that it would be regrettable if other organisations felt obliged to develop separate grading systems in the absence of a suitable service from the ECF. "

I always thought it a pity that UKCC has traditionally refused to send games for ECF grading. I hope the new management will come on board.

[NickFaulks]

"What about someone who plays most of their chess and most of their improvement in on-line chess, or any chess not captured by the ECF system? "

I would have thought that covers a lot of players.

Can you point to one? I don't mean a visitor from overseas.

[J T Melsom]

The thing about visitors from overseas is not that we care about getting their grades right for its own sake, but that we care about getting their grades right so that English players get appropriate credit for ECF-graded games against them.

One of my club-mates is in this position. He played two players one without an English grade and one without a FIDE, but of similar strength. The ECF will generate an estimate for his loss against the foreign opponent so he will get 9 graded games, but only 8 games will go into the FIDE system. As the games were a win and a loss rather than two draws that in itself will cause divergence. (though my grasp of such matters has been less than 100% in the past

[NickFaulks]

The thing about visitors from overseas is not that we care about getting their grades right for its own sake, but that we care about getting their grades right so that English players get appropriate credit for ECF-graded games against them.

Then convert their FIDE ratings, if you must use them at all. Trying to incorporate them into an avowed national system that is not their own is guaranteed to fail.

[Roger de Coverly]

Many?

They wouldn't have any problem then, in getting approval from a Council meeting, presumably the next one.

I think the demand is for new players to get a grade in a shorter period than 6 months. I doubt the majority demographic of club and Congress players, many of whom have grading histories stretching back thirty years or more, have much interest and may not want the established grading methods scrapped and replaced by a leap into the unknown.