Northumbria Masters

[LawrenceCooper]

Looking at the PGN he had four whites and five blacks.

Thanks, I presume the PGNs are definitive. It's a shame the colours could not have been reported correctly, because these things do get looked at, particularly when permission has been given for an innovative system to be used.

Tim contacted me to confirm that "the issue with the way the colours were published was because Swiss Manager is not really geared up yet to cope with Schiller tournaments.
We published the pairings with correct colours in Chess-results as attached excel files:
GM - https://chess-results.com/tnr802867.aspx?lan=1&art=0
IM - https://chess-results.com/tnr802873.asp ... =0&flag=30
These pairings have been up since before the tournaments began."

[IM Jack Rudd]

Yeah, the best way to do a Schiller in SM is probably to set it up as a Swiss with manual pairings.

[NickFaulks]

Tim contacted me to confirm that "the issue with the way the colours were published was because Swiss Manager is not really geared up yet to cope with Schiller tournaments.
We published the pairings with correct colours in Chess-results as attached excel files

That's all very well, but I wish people could grasp that the results submitted to FIDE are supposed to be correct, even more so when title norms are being claimed.

I assume the IRO was unaware that he was signing off an error strewn report.

[Alan Atkinson]

The Northumbria Masters this year produced a large number of norms: perhaps we might first recognise that the arbiter team for this event again allowed players to gain title norms, not something that many events consistently achieve.
FIDE were sent the PGNs, and so will have info about the colour rotation, which was done correctly, and about the correct individual pairings, which were decided by lot and published online and to the players themselves in advance of the event's start.
It is a pity that Swiss Manager does not easily output the correct colour rotation to Chess-Results, from where I assume most of the commentators here have obtained their information.
I will take the point made by Jack Rudd that I might have then manually inputted the pairings, but that in turn might have raised more eyebrows at FIDE!
I do intend raising the matter with H Hertzog, and ask if there might be some way to output the data the way I wanted it. Those that use Swiss Manager might aknowledge that it does have a range of particular little sub-routines: I might have missed the one, but this was a new format to me too. Maybe Heinz has already addressed the matter.
Perhaps if individuals here have time available, they might set up the event manually and upload it to Chess-Results, so as to show the actual colours used, but I do not have that time. And over the duration of the event, my arbiter team certainly did not. FIDE do not use colours in determining ratings, and so I did not see it as a matter of very high priority. And as already stated, the colours were known in advance to all that mattered, and FIDE were aware because every PGN is available to them. And the actual colours used rotated correctly.
So what is the problem here?
Finally, we have an event that produces norm opportunities for our players and arbiters. It did not cost the Federation so very much.
Is it too much to ask for people to support and celebrate such events?

[Ian Thompson]

So what is the problem here?

Incorrect information on the FIDE rating website - https://ratings.fide.com/calculations.p ... 1&rating=0 and https://ratings.fide.com/calculations.p ... 1&rating=0, for example.

[Alan Atkinson]

So what is the problem here?

Incorrect information on the FIDE rating website - https://ratings.fide.com/calculations.p ... 1&rating=0 and https://ratings.fide.com/calculations.p ... 1&rating=0, for example.

Interesting that although the actual pairings, and the PGNs, all of which FIDE have, show the correct colours, this info is used, even when colours do not feed into the rating calculations.
But again, is it such a problem for you Ian?

[Roger de Coverly]

But again, is it such a problem for you Ian?

The FIDE rating site is the place to go for information about past games. Something simple lke recording the colours correctly really should be done. Personally I don't buy excuses about computer software. If you don't get the required result, use something else. It shouldn't be difficult to produce a correct rating file for such small tournaments.

Some researchers have experimented with making adjustments to ratings based on colours as well as results. It would be useful for such purposes if the FIDE data was not corrupted.

[Ian Thompson]

Interesting that although the actual pairings, and the PGNs, all of which FIDE have, show the correct colours, this info is used, even when colours do not feed into the rating calculations.
But again, is it such a problem for you Ian?

FIDE's rating website also provides player performance statistics, so it does matter that colours are reported correctly.

[NickFaulks]

Interesting that although the actual pairings, and the PGNs, all of which FIDE have, show the correct colours, this info is used

The FIDE system relies heavily on the starting assumption that information provided by IROs may be regarded as accurate. There is an informal list of federations for which that is in doubt and you really don't want to be on that list.

Colours are of importance, particularly in norm events. If it were found that a player had been advantaged by manipulation of the pairings then their norm, and probably any others from the same event, would be in question.

There is also the point that if a misreporting of colours has got through the system, you have to wonder which other errors may have done.

[Alan Atkinson]

The fact is that the colours were indeed properly balanced, and were determined by a drawing of lots and published in advanced.
Had the rating file been manipulated to ‘correct’ the colours, or even just the colour output from the approved pairings programme, then there would properly have been questions asked as to what else might have been ‘fixed’.
But nothing like that was done, and I presume that you are not suggested anything of the sort.
Which then returns us to the point where you might look to support events that provide such opportunities for players to obtain title norms.

[Roger de Coverly]

The fact is that the colours were indeed properly balanced, and were determined by a drawing of lots and published in advanced.
Had the rating file been manipulated to ‘correct’ the colours

The point is that the published ratings data should be consistent with the pre tournament pairings, the daily result reporting, the daily on line coverage and the games making their way into TWIC and other database collections through the pgn files.

If chess-results cannot do this, don't use it directly. It's not a offence to correct faulty data


Incidently the ECF rating data is also incorrect. Perhaps this is an issue that should be raised witth the ECF rating team as to whether this should be tolerated. It may be better to remove the colours than to report them incorrectly.

[NickFaulks]

Alan,

The fact is that a rating file for an event from which norms are likely to be claimed has been submitted by the ECF while containing serious errors.

The fact also is that you consider this to be entirely acceptable behaviour and are indignant that anyone in FIDE might disagree.

Not much more to be said really.

[Brian Valentine]

While the ECF rating team can't regard colours to be "serious", in all othe regards we can ony echo Nick's position.

[Roger de Coverly]

Yeah, the best way to do a Schiller in SM is probably to set it up as a Swiss with manual pairings.

Would it also be possible to set it up as a truncated all play all, also with manual pairings? That way chess-results would give the round by round results, player by player? Running it as manual Swiss would do this as well, but as an APA the future pairings would have been shown.

Round by round results are elusive but at
https://chess-results.com/tnr802867.asp ... g=30&snr=4
as found by an earlier poster.

Are these correct? If they aren't that is presumably why the rating files are wrong as well.

[DavidWalker]

The GM & IM sections are set up in Swiss Manager as team tournaments in which four teams of three players play each of the other teams three times. By default, SM assumes that in team tournaments colours alternate on successive boards, and this appears to be the problem with reporting of colours in the FIDE Tournament Report. However, SM also provides a check box on the Tournament Data Dialog which allows this default to be overridden:

SM Capture.jpg

Ticking the "All boards the same" box appears to fix the problem with misreported colours in the FIDE Tournament Report.