Seeding unrated players for Swiss tournaments

[strollingrhino]

Looking at that cross table and Pat's reply, I think anything from 1600 to 1800 is ok based on his two results. I was hoping that in his second game, Simon had defeated at a player rated around 1600, then I would say 1800 with more confidence.

Okay thanks. In which field should I put the rating in Vega. The same field as the ACF? And then add a note to the file so it makes sense to the ratings officers at CAQ then ACF??

[Garvinator]

Okay thanks. In which field should I put the rating in Vega. The same field as the ACF? And then add a note to the file so it makes sense to the ratings officers at CAQ then ACF??

A few comments:

1) Using Vega, the most useful 'table' for players and spectators to display is the Ranked Cross Table. That shows the players in score order, and then rating.

2) Thinking a bit further about what rating to assign Simon, you could wait an extra round. Having him as the fourth seed in the '2's' is not much of an issue. And then after that game, you will have much more useful information about what rating he should be for the rest of the tournament as he will have played one of the top seeds.

3) If you are going to give Simon a rating, you place that provisional rating in the same field as the ACF rating for all the other players. Make sure that Simon has been re-seeded according to his new rating before you pair round three.

No notes are required as all players are identified by their ACF id's, not by a rating that an arbiter has assigned them for a tournament. The id that will be assigned to Simon will be sorted out by CAQ and ACF. You do not need to worry about that.

[Kevin Bonham]

Make sure that Simon has been re-seeded according to his new rating before you pair round three.

Yes, it's easy to change a rating or add a new player and forget to do this.

Extras -> Regenerate pairing number (Swiss system)

[strollingrhino]

Thanks a lot everyone. Can you believe I have more interesting rating/seeding questions from the same small event! The arbiting gods must be smiling down on us!

We have a player Aarav Rawlley, who is on the ACF rating list as 100. Quite surprisingly. He said this is from events held years ago as he is currently 16 years old. He has been playing at the club for 3 weeks and is much stronger likely in the 800-1000 range. His current QJ is 817, also likely inaccurate. I have asked him for a current online rating to assist with properly adjusting his seeding/rating? He scored R1 1429 loss as wh, R2 920 win as bl in our current event.

Also there is Ivan Luey 1900x who earned this rating playing high school chess in NSW 40 years ago and is just coming back to chess. His performance rating in our last classical event was 1113 based on a score of 4/8 against opponents (1123 win as wh, 1427 loss as bl, 947 loss as wh, 1219 loss as wh, 940 win as bl, 1230 draw as bl, 676 draw as wh and 1266 win as bl). I was unaware of the arbiter's adjusting ratings as per this thread being very new to the role. As a result Ivan Luey was seeded No1 in our current event and lost R1 to a junior 1064 as bl, and R2 won to 664 as wh). Since I am reseeding and adjusting Simon Dolosa to 1693, and Aarav Rawlley to ?? 800 perhaps, I believe it would be ideal to also adjust Ivan Luey as well. Would 1100 be fair?

I will be on top of these issues from R1 for our next events - as I had no idea I could/should adjust ratings to reflect playing strength in the above three cases:

Case 1 - Overseas player - no ACF or FIDE but has a National Chess Rating from the Philippines 1693 - did not state any rating when asked
Case 2 - Returning to Chess Player significantly underrated - ACF 100 but strength closer to 800-1000
Case 3 - Returning to Chess Player significantly overrated - ACF 1900x hasn't played for 40 years performing closer to 1100 ACF

Thanks for your feedback and directions.

[Scott Colliver]

I would not adjust the ratings of those who already have ratings. The rating system will very quickly change their ratings to reflect their current strength.

[Garvinator]

Thanks a lot everyone. Can you believe I have more interesting rating/seeding questions from the same small event! The arbiting gods must be smiling down on us!

We have a player Aarav Rawlley, who is on the ACF rating list as 100. Quite surprisingly. He said this is from events held years ago as he is currently 16 years old. He has been playing at the club for 3 weeks and is much stronger likely in the 800-1000 range. His current QJ is 817, also likely inaccurate. I have asked him for a current online rating to assist with properly adjusting his seeding/rating? He scored R1 1429 loss as wh, R2 920 win as bl in our current event.

Also there is Ivan Luey 1900x who earned this rating playing high school chess in NSW 40 years ago and is just coming back to chess. His performance rating in our last classical event was 1113 based on a score of 4/8 against opponents (1123 win as wh, 1427 loss as bl, 947 loss as wh, 1219 loss as wh, 940 win as bl, 1230 draw as bl, 676 draw as wh and 1266 win as bl). I was unaware of the arbiter's adjusting ratings as per this thread being very new to the role. As a result Ivan Luey was seeded No1 in our current event and lost R1 to a junior 1064 as bl, and R2 won to 664 as wh). Since I am reseeding and adjusting Simon Dolosa to 1693, and Aarav Rawlley to ?? 800 perhaps, I believe it would be ideal to also adjust Ivan Luey as well. Would 1100 be fair?

I will be on top of these issues from R1 for our next events - as I had no idea I could/should adjust ratings to reflect playing strength in the above three cases:

Case 1 - Overseas player - no ACF or FIDE but has a National Chess Rating from the Philippines 1693 - did not state any rating when asked
Case 2 - Returning to Chess Player significantly underrated - ACF 100 but strength closer to 800-1000
Case 3 - Returning to Chess Player significantly overrated - ACF 1900x hasn't played for 40 years performing closer to 1100 ACF

Thanks for your feedback and directions.

I think it is important to remember where this thread started from and where our responses came from. The thread start from - How to and the need to give a rating to unrated players.

These new questions are in regards to adjusting a players already existing ACF rating. This I am firmly against. As you can see from the two examples given, it works both ways. A player can be, in the opinion of the arbiter, under rated, and the other player, over rated.

But then, that player goes to another tournament during the same rating period, a weekender perhaps, with different arbiters and in a different area and is rated according to their ACF rating. That can get rather confusing for the player (either player in this case).

Ratings are adjusted every three months and the new ratings will be out September 1st. If either player is genuinely under or over rated, then the ACF rating system (Glicko2) will pick this up and correct the players rating, as long as the player has been performing as claimed.

There is also an additional issue with down rating Ivan Luey from 1900 to under 1600. If you were to run a tournament with two divisions, usually there is an open division and an under 1600 division. By re-rating Ivan Luey yourself, you now make him eligible for the under 1600 division. And then he goes on to win the event. And other players find out that his 'proper' acf rating is over 1900, but it was the arbiter that adjusted his rating to a point where he could play in the under 1600 event.

In my opinion, there are only very small good points for adjusting a players published acf rating, and many minus reasons. Stay with the published ratings. If either player is performing as you say, then the ACF rating system will adjust their rating quickly enough.

[strollingrhino]

These new questions are in regards to adjusting a players already existing ACF rating. This I am firmly against. As you can see from the two examples given, it works both ways. A player can be, in the opinion of the arbiter, under rated, and the other player, over rated.

Okay - I will leave them as is and let Glicko-2 and Sept 1 do the adjustments... although this tournament won't finish until after Sept so Aarav will need to live with ACF 100 until it is adjusted in the future. Thanks everyone!

[Kevin Bonham]

I realise this query was re an ACF rated tournament, but just in case anyone reading the thread is thinking about FIDE rated tournaments, I suggest being careful about adjusting seeding ratings in those. This is what the FIDE General Handling rules have concerning adjusting ratings after the tournament has started:

3

This ranking is used to determine the pairing numbers; the highest one gets #1 etc.
If, for any reason, the data used to determine the rankings were not correct, they can be adjusted at any time. The pairing numbers may be reassigned accordingly to the corrections. No modification of a pairing number is allowed after the fourth round has been paired.

This covers at least the following situations:

(i) A FIDE rating has been entered for a player but that is not their correct FIDE rating.
(ii) A player has been assumed to be unrated but it is found that the player has a FIDE rating.

It is not clear that it covers a case where a player has been given an inaccurate estimate of playing strength because relevant information was not known.

It's very important in the FIDE system that pairings be reproducible - two arbiters given the same data should get exactly the same pairings. If arbiters are making subjective decisions about whether to change a seed rating and when to change it then that principle is breached and in a norm tournament that could cause a lot of trouble.

Also Garvin and Scott are correct - an existing rating should not be altered whatever the rating system, even if it is obviously not a true reflection of the player's strength.

[Patrick Byrom]

... Also Garvin and Scott are correct - an existing rating should not be altered whatever the rating system, even if it is obviously not a true reflection of the player's strength.

I'm going to be a contrarian here

I completely agree that making adjustments to established ratings should not be done - that's what the national rating system is for. However in Aarav's case the suggestion is to use his QJ rating instead of his ACF rating. I don't see a problem with this in ACF-rated (only) events, especially if the QJ rating is much more reliable than the ACF rating. My only caveat is that such a policy should be announced in advance of the event.