Message to all Graders

General discussions about ratings.
Post Reply
User avatar
John Upham
Posts: 7162
Joined: Wed Apr 04, 2007 10:29 am
Location: Cove, Hampshire, England.
Contact:

Message to all Graders

Post by John Upham » Tue Dec 10, 2019 12:31 pm

I have been passed the following message by various interested parties :
Fellow graders,
For us this year end marks the beginning of a new era.
We will be closing the grading period for all results up to 31st December and please submit all files to Matt Carr by 14th January.
We intend to use the January list, converted to 4 digits, as the starting point for monthly grading. Hence I would ask you to use more diligence to get the dates of each result submitted correctly for the new year as it will be more critical for grading correctly.
Our current plans remain to continue thee old three digit process in order to get out the next August list in traditional format. Monthly grades will have no status until then, but as our new system gets up to speed we will publish monthly grades for each month in 2020.
As has been stated before there is no compulsion from the ECF to increase your frequency of submission, but I do hope that when you can you can support this new initiative by doing submissions more often.
I recognise that we need to help by easing your task. With that in mind we will be obtaining most of the relevant 31st December submissions direct from LMS with no intervention from the graders involved. We will need the graders involved to help us correct resulting errors. Malcolm Peacock will be contracting organisers with more detail in the near future.
In due course, where we think this idea will work we will be contracting graders involved.
While we have done some testing, we need to see whether this does make graders life easier or that this does not merely pass all the workload onto Matt Carr. If this process is a success then we will make the process available to other software providers so more submissions in the new year can be simplified. Our reticence is due to the possibility of unnecessary correspondence where errors that humans would spot get through the system.
May grading work not upset your Christmas,
Brian
As a provider of an LMS for more than one League it would be nice to be in the above loop but I am not a grader so it is mushroom treatment for me.

I've still heard Sweet FA concerning the specification of the API to allow various LMSs to communicate with the ECF back end.
British Chess News : britishchessnews.com
Twitter: @BritishChess
Facebook: facebook.com/groups/britishchess :D

Nick Grey
Posts: 1838
Joined: Thu Dec 01, 2011 12:16 am

Re: Message to all Graders

Post by Nick Grey » Tue Dec 10, 2019 9:12 pm

I do not believe this was mentioned by any of the graders at the beginning of the season. It is difficult enough to do it twice a year but monthly testing if not official was not mentioned from Jan 2020. September I could understand.

I do not see why my ECF fee goes up in August 2020 for something I do not want.

Richard Bates
Posts: 3338
Joined: Fri Nov 14, 2008 8:27 pm

Re: Message to all Graders

Post by Richard Bates » Thu Dec 12, 2019 8:52 am

Has the calculation methodology been released?

Brian Valentine
Posts: 574
Joined: Fri Apr 03, 2009 1:30 pm

Re: Message to all Graders

Post by Brian Valentine » Thu Dec 12, 2019 9:43 am

We are trying to pass out information when we have it and intend to consult where it seems a good idea.
Richard Bates wrote:
Thu Dec 12, 2019 8:52 am
Has the calculation methodology been released?
https://www.englishchess.org.uk/monthl ... -proposal/

This proposal has broadly passed later tests. I hope to publish an update with refinements soon, at present our focus is in designing enough of the administration to start talking to John about working together.

Brian Valentine
Manager of ECF Grading

Richard Bates
Posts: 3338
Joined: Fri Nov 14, 2008 8:27 pm

Re: Message to all Graders

Post by Richard Bates » Sun Dec 15, 2019 7:55 am

One observation, especially with the approach to late reported results (both in allowing tolerance for inclusion, and, in particular recalculating prior lists) is that it will become impossible for people to predict for their grades will be. I would predict that this will be very unpopular (even if it might have some redeeming features in potentially making manipulation by individuals harder).

It will be interesting to see what the calculations will consider likely for the “range” of likely grades that a “stable strength” player might experience over the course of time. I imagine that year long events will still seek to consider certain lists “official” for their purposes (and indeed certain lists will be more complete in terms of results submission) and so it will be interesting again to see if certain people mysteriously end up with consistently lower grades on those lists.

Post Reply