REGULATIONS ON REGISTRATION & LICENSING OF PLAYERS

[John McKenna]

Why continue to try churning chalk into cheese? A (large?) proportion of ECF-graded standard games are still played faster than standard FIDE-rated ones. Why hasn't someone worked out a conversion formula for ECF rapid grades to standard grades? And, if they did how much would you trust it?

23:50 NB: Several posts (at foot of previous page) got in before me and make a better case than mine.

[Roger de Coverly]

What is the point of such a conversion?.

As Jack notes, both team chess and Swiss pairings depend on being able to establish a ranking order. Even if every player has a rating or grade on both systems, you still need to know which one to trust. If you have players with a rating or grade in only one system, you need a conversion formula to establish a ranking order.

A principal reason that Federations maintain domestic rating systems is that they wish to rate chess which is played at move rates too fast for the FIDE rules. Increasingly even if the FIDE rules on session length were relaxed, the imposition of onerous requirements by FIDE, of which the latest edict is but one example, reduces or destroys the case for abandoning the domestic system.

It's often assumed that a National Elo equals an International one and no conversion formula is needed. That isn't or hasn't been the case for the USCF where based on the ratings of the top players, the USCF ratings were around 50 points above the International ones.

[Sean Hewitt]

Pah, Alex has long contributed to chess in England and to chess organised by the BCF/ECF - such as the British Championships.

No one denies that he contributes to chess in England.

Of course he should be informed of..... FIDE's edict.

Irrespective of his contributions, he is registered SCO with FIDE. It is for Chess Scotland to inform him of such matters, as I'm sure they have done.

Of course he should be...consulted on how the ECF responds to FIDE's edict.

That's nonsensical. Without wishing to personalise this; if the ECF consulted on topics such as this based on the longevity of a person's contribution to English Chess, it would be consulting far wider than Alex McFarlane. It would then never be able to get anything done. I am rapidly coming to the conclusion that such inertia is considered desirable by some.

In fact, it seems that the only thing some people like to moan about more than the ECF doing nothing, is the ECF doing something.

[MSoszynski]

Of course, the comparison is not helped by the fact that the current "official" ECF / FIDE conversion function is mathematically gibberish.

The "official" conversion is now 8*ECF + 650 = FIDE. There's a case that the 8 could be reduced slightly because of the non-linear nature of the Elo formula. The 650 is the original 600 plus an extra 50 when the ECF graders convinced themselves that all the grades were wrong except for the top players. As they added about 5 or 6 points to players around 175, that can justify 2000 no longer being equivalent to 175. You might be thinking of 5*ECF + 1250 = FIDE which was nonsense as you suggest and caused by someone running amok with a least squares fit on ECF v FIDE. This was dropped in 2009. The recent Sonas graphs published on chessbase showing under performance at the extremities are equivalent to what caused the ECF graders to embark on a revaluation exercise, supposedly to correct for deflation.

There was a survey carried out many years ago by JP Wilkinson comparing the grades and ratings of players that had both. He came up with the formula:

(ECF x 5.701) + 1136 = FIDE

[Jonathan Bryant]

Why continue to try churning chalk into cheese?

I think perhaps what you're not grasping is that the two systems can produce very different results when applied to the same set of games (e.g. those from my tournaments at Hampstead and Penarth). Chalkycheese, if you like, rather than chalk and cheese.

[Brian Valentine]

There are good reasons set out in this thread: http://www.ecforum.org.uk/viewtopic.php ... 62#p121762 why the gradient is set at 8 and why some of us think something slightly less could work. 5.7 has to be a poor model.

[Michael Flatt]

In fact, it seems that the only thing some people like to moan about more than the ECF doing nothing, is the ECF doing something.

It does seem to me that some members of the ECF hierarchy are overly sensitive to comments posted here and are too quick to assume that they are personal attacks when this was clearly not intended by the poster of the comment.

With regard to Alex McFarlane, although he is registered with FIDE as SCO he is very prominent in chess events in England as others have noted here. Even if ECF post holders do not see the need to inform Alex directly on matters such as FIDE registration it would be a friendly gesture and courtesy to talk with him.

Given that FIDE registration is of interest to all Arbiters, be they FIDE registered or not, I would have expected that this is something that the Chess Arbiters' Association (CAA) would have taken an interest in. It should be noted that the CAA seeks to represent all Arbiters in the UK regardless of their individual national affiliation.

I may be wrong, but doesn't the CAA have their own links with FIDE through which they will have been informed and able to cascade the information to their members?

[Roger de Coverly]

In fact, it seems that the only thing some people like to moan about more than the ECF doing nothing,is the ECF doing something

Is that really a surprise? The ECF has a long track record of making questionable decisions.

[Paolo Casaschi]

What is the point of such a conversion?

The primary reason is to prevent certain tournament formats from falling over when confronted with players who have one but not the other. For example, the 4NCL has a rule that players must play in - within a tolerance of 80 points - descending order of strength, where strength is defined in terms of one's FIDE rating. (Thus, for example, with my FIDE rating of 2233, I must play below any of my team-mates who are rated 2314 or higher, and above any rated 2152 or lower.) There are, in any given season, many English players with no FIDE rating who play in the 4NCL. Where in the team should they play? Well, the conversion formula serves as a rough-and-ready guide as to where best to put them.

The same thing applies in reverse to foreigners who show up at English weekend congresses; they're likely to have a FIDE rating but no ECF grade, and the conversion formula can be run backwards to put them in the draw at the appropriate point.

I understand the benefits of having everybody on the same ranking system.
I find it difficult to understand how so many people on this forum first advocate the need of separate FIDE rating and ECF grading, explaining at great length the differences between the two and the benefit of keeping a national list [1]. Then an equal number of people argues about very quick&dirty formulas for comparing the FIDE apples with the ECF oranges because we ended with two rating/grading systems that behave substantially different. The mathematics behind the Elo systems is far from trivial; assuming a very simple linear relationship between FIDE Elo and ECF grade is a very easy to use but far from accurate approximation.

[1] no need to post here again all the reason why we cant scrap the ECF grading; I understand there are some technicalities involved (like rating of evening chess) *at the moment*; what I find difficult to understand is there seems to be very little interest in working to overcome those issues; many people openly stating they prefer a duplicate system with separate rules even if it were possible to rate all the games currently graded.

[Alex McFarlane]

doesn't the CAA have their own links with FIDE

Michael FIDE only deals with National Federations though the CAA does have individuals who can pass on information and often do. In my case I was aware of the situation through Chess Scotland.

However, as a member of the ECF I do not understand the policy that has been adopted of not telling members who are registered with other Federations. This seems to contradict its own Articles. The ECF Memorandum of Association Article 3.11 states "To make the Companies services available without discrimination on grounds of colour, creed, disability, impairment, occupation, religious or political affiliation, or sexual orientation ..."
My political affiliation is to CS. As Alex H points out. at least two others have been 'refused' information for a similar reason.

[Phil Neatherway]

Having just checked Article 3.11, I can confirm that it does not include such an egregious spelling mistake!

I would not interpret being a member of CS as a political affiliation. CS is not a political organisation.

[Roger de Coverly]

many people openly stating they prefer a duplicate system with separate rules even if it were possible to rate all the games currently graded.

As the other system is run by FIDE, that's more than enough reason to maintain an independent system. As witnessed by their latest edict FIDE now rank the ease of running a rating system above the promotion of chess.

[MartinCarpenter]

Mantaining the grades as distinct entities actually makes a lot of sense for the sorts of stronger players who've traditionally had FIDE ratings in the UK. In that case their FIDE rating represent the results of their long play games against their approximate peers. That is the chess they actually take seriously. You can easily see why a measure of that is very useful and why they'd want to preserve it. Especially with international tournaments, titles etc potentially involved.

However a lot of these people will also be playing in congresses/evening leagues. Those will inevitably contain a lot of games against much weaker opposition (30/40(+) pts less) and various tactical factors in the last round and the like. It'll all, of course, get into their ECF grade but isn't really very useful information as to their proper playing strength.

Extending it to many more players is a new thing and has unarguably caused some non trivial problems with juniors not catching up and people getting grades based on little data. Maybe the worst of these will be sorted in time.

[Michael Flatt]

Alex, thank you for clarifying the relationship between the CAA and FIDE.

I think that as an ECF member, who is also a registerd FIDE Aribter, you are right to be expected to be consulted.

Making such fine distinctions as to whom the ECF should share information seems unnecessarily provocative and confrontational.

Allied with another post by Paul Buswell regarding the decision not to send membership reminders to inform by post to those ECF members who prefer not to use email or do not have access to the internet does to me highlight an attitude of "I am all right, Jack".

Ordinary members have a right to be treated with more respect by those they elect to Office.

[John McKenna]

As has been noted before on the forum - those who members (of any metal) actually elect are relatively few and their powers far between, and that leaves plenty of scope for ignoring those they represent.

Jon Bryant >Chalkycheese...<
I accept the concept in terms of grating (grading & rating) but hate the thought of the taste.