ECF Director wants to scrap Clarke grading system

[E Michael White]

The objective is to produce monthly grading lists.

Helo Brian, If that is the objective the first task should be to assess how much deflation/spread/nonlinearity will increase/decrease and how to regularly measure and control those. Good luck with that !

[Ray Sayers]

If monthly ECF rating lists came into force, how would that affect players entering grade-restricted tournaments more than a month in advance?

Would they have to swap sections if they moved into another band? That may affect their decision to enter, requiring refunds and so on.

I'm not criticising the idea - just trying to understand what practical effect it would have.

[Mick Norris]

It is a problem, I guess, unless entry forms state which grading list will be used

Have congresses and leagues actually been asked if they want monthly grading lists?

[Roger de Coverly]

Would they have to swap sections if they moved into another band? That may affect their decision to enter, requiring refunds and so on.

We've had this problem for some time, ever since e2e4 and subsequently 4NCL have been running rating restricted tournaments based on International Elo ratings. It doesn't seem to have been a problem in practice, but with only a limited number of tournaments involved, it's easier for individuals to manage a rating.

But it's a consequence of monthly rating lists, that tournaments have to declare which list to use for eligibility and seeding. Tournaments such as Kidlington taking place in February, shortly after the publication of the January list, currently go back to the list of the previous August.

[MartinCarpenter]

Its very simple, as per Yorkshire say - you pick a month you want to take as representing the yearly grade and use that through the year for things like board order/tournament sections etc.

[Alan Walton]

I always laugh with myself how people in England always want to find problems with any change proposed

I suspect most countries in the world use a FIDE type of system to calculate ratings, updating them frequently during the year and I suspect they don't have any major problems (if they do I doubt they would complain as much as the people do on this forum)

Lets just get rid of the poor English grading system and move to the FIDE system, lets try and start modernising English chess and this is a good first step

[Adam Raoof]

At Golders Green Rapidplay I've always used ECF rapid grades to determine eligibility for sections, and yet rated the event with FIDE. Now most of the players have FIDE rapidplay ratings I've decided to use those for seeding the players.

However, ECF grades will still determine eligibility for sections! This is because they are often more reliable, and now we are using Swiss Manager / chess-results.com software using fide for seeding makes the rating calculations work. FIDE ratings, though appearing monthly, are lagging behind ECF.

In the near future for events such as Hampstead (Under 2200, Under 1900) I may have to look at both rating AND grade to determine eligibility for the Under 1900 section (e.g. you have to be Under 1900 and also Under ECF 160 - at the moment you can be 1600 and yet ECF 170 and still play).

I am generally in favour of Elo-style monthly ratings, but it would only really kick off if leagues and congresses adopted it. Maybe scrapping the ECF system would assist that process?

[Andrew Zigmond]

I always laugh with myself how people in England always want to find problems with any change proposed

I suspect most countries in the world use a FIDE type of system to calculate ratings, updating them frequently during the year and I suspect they don't have any major problems (if they do I doubt they would complain as much as the people do on this forum)

Lets just get rid of the poor English grading system and move to the FIDE system, lets try and start modernising English chess and this is a good first step

The problem is you can't modernise a lot of English chess players!

Seriously, in his statement Alex has set out cogent reasons why reform is needed. I find it telling that when Roger jumped in with the negatives he didn't comment on any of the detail.

Given that, as Alex says, any new system needs to be tested and embedded it might well be that the old Clarke system continues for a time in tandem with a new ELO based system subject to eventually being switched off (a bit like the change from analogue to digital). Some of the more conservative leagues will struggle with monthly online reporting but a) their players are not likely to be affected that much and b) if they are it's down to them to do something about it.

[Roger de Coverly]

I find it telling that when Roger jumped in with the negatives he didn't comment on any of the detail.

If you want to abandon the existing grading system, that's a pretty big decision in its own right and the one to concentrate out. Once you've made that decision, it acts as an enabler if you want to do things like monthly rating. If standard Elo rules are followed, players only get ratings by playing other players with ratings, so it potentially isn't as good as the existing system in generating grades for lots of new players.

A major problem with Elo based systems, particularly when used at a National level to apply to near beginners is how they cope with rapidly improving players. The Scottish version has a rule that if the performance is more than 200 points above the published rating, then players are treated as new.

[MartinCarpenter]

Can always tweak grading systems to work

The biggest obstacle would probably be getting the centralised infrastructure to work - definitely a step up vs the current grading website so I'd guess it'd need rather more commitment to that from the ECF than currently.

[Roger de Coverly]

The biggest obstacle would probably be getting the centralised infrastructure to work - definitely a step up vs the current grading website

I would be inclined to think getting monthly reporting from Leagues the bigger challenge. That and a complete rethink on how to handle new players and juniors. Storing more grades in a year is a lesser challenge isn't it?

[Christopher Kreuzer]

I would be sad to see the Clarke system go, as I am used to it. There are players who have been around far longer than me who will similarly not want to change from something they don't see as broken. It would also feel like a discontinuity in the grading history, far worse than the 'adjustment' made a few years ago. Being able to compare grades over time is a massive advantage of keeping to a stable, reliable and known grading system. Mentally being able to categorise players as "150 strength", "170 strength" and "200+" and so on, is part of the mental landscape of players up and down the country. Switching to a four-figure system will be really hard for some, regardless of how easy it may be to convert.

The other aspect of it is that your grade can become part of your identity. You have a single figure to remember through a whole season (well, now two figures). If that figure changes each month, it is far, far harder to remember. Rather than remembering at the beginning of the season that you are "164" (or whatever), you have to keep looking up your new grade. That is in some ways symptomatic of a world speeding up and running ever faster, demanding instant updates and constant inputs. That is not always a good thing. There is nothing wrong with a grade every year (or twice yearly), and then playing chess during a season and getting a new grade at the end of the season (or midway through the season).

I would hope that if there is a trial of this, that the systems would run in parallel and that serious consideration be giving to dropping the 'new' system if it proves unpopular.

[David Sedgwick]

[Alex] has been advised not to post on here by the current regime. This is consistent with the "advice" I was given when a director, I was actually told not to post but when I questioned it this was amended to "asked"

I expect that if Alex had wanted to respond "on here" he would have posted "on here" on the first place.

What makes you think that? My expectation is that he would have liked to respond here, but feels constrained by the instruction. Oh sorry, I mean advice.

I've asked Chris Majer, Chairman of the Governance Committee, to address this issue in his report to Council.

[David Sedgwick]

I think the AGM should be asked to vote in its favour rather than have its consent assumed.

The Finance Council Meeting in April 2016 will need to approve the expenditure on the League Management Software - or decline to approve it.

I understand that it's a considerable sum.

[Michael Flatt]

I think the AGM should be asked to vote in its favour rather than have its consent assumed.

The Finance Council Meeting in April 2016 will need to approve the expenditure on the League Management Software - or decline to approve it.

I understand that it's a considerable sum.

Before worrying about funding the general principle of whether or not to introduce a continuously updated grading system and monthly grades needs to be debated, as it is here.

The proposed League Management System is a facility to be offered to leagues to assist in the collection of results. If you want to introduce more frequent results collection it seems reasonable to find a way to help those responsible for submitting the required data.

Nothing drastic will happen any time soon.

Grading falls within the province of the Director of Home Chess so I don't understand how the question of Governance arises. No doubt the ECF board will debate the Home Director's proposal and make recommendations to Council.