European Club Cup

The very latest International round up of English news.
LawrenceCooper
Posts: 5049
Joined: Tue Dec 20, 2011 8:13 am

Re: European Club Cup

Post by LawrenceCooper » Fri Oct 25, 2013 11:15 am

Jonathan Rogers wrote:I suppose it's symptomatic of the fact that I don't play swiss events, but these pairings are really annoying me! They don't seem to pair teams on the same matchpoints either by gamepoints or by seeding! So, by whatever system they are using, it came to pass that yesterday Jutes and Barbican. both among the stronger teams on 50%, played yet stronger teams whilst White Rose played Edinburgh. Now today, White Rose, by far one of the weakest teams on 6/10, are still playing weaker opposition than are Barbican on 4/10!

I believe that Barbican is the only team all of whose opponents have been in the top 30. There is no earthly reason why. We have had one excellent result (beating beskitas, 20th seed) and convincingly beat the one slightly weaker team we played, but have lost two matches (albeit against significantly higher rated sides) by greater margins than might have been expected - so overall, "about right" for our seeding. So what is so special about us that we should get this string of pairings?!
I had that feeling when captaining England, it often seemed to stem from having a narrow victory in round 1 but no one was able to explain the pairing rules when I enquired so we just had to get on with it.

Mick Norris
Posts: 7498
Joined: Tue Apr 17, 2007 10:12 am
Location: Bolton, Greater Manchester
Contact:

Re: European Club Cup

Post by Mick Norris » Fri Oct 25, 2013 11:44 am

If you really want to, you can access the pairing regulations here:
http://www.europechess.net/index.php?op ... &Itemid=15

If you download the document, the relevant bit starts at page 26 - it appears to be a different system as per the note at the end:
Explanation by IA Dr. Dirk J.A. De Ridder
1. In contrast to the system developed by IA Almog Burstein which is used at the Olympiad and which was used at the occasion of the European Club Cup in Neum (BIH, 2000) no extended Buchholz points have to be calculated whereas the board points are immediately available.
2. Meaningful extended Buchholz points can only be calculated after round 4; at the occasion of the European Club Cup in Neum (BIH, 2000) it became clear that this calculation was not obvious for the participating teams.
3. The system proposed here offers the advantage that one pairing system is applied throughout a competition. Application of the Olympiad system at the occasion of the European Club Cup in Neum (BIH, 2000) resulted in the criticism that the pairings were hardly to be understood by the participating teams.
4. The system proposed here is based on the observation that, in general, a team with a high number of board points will have a low amount of extended Buchholz points and vice versa.
Any postings on here represent my personal views and should not be taken as representative of the Manchester Chess Federation www.manchesterchess.co.uk

Jonathan Rogers
Posts: 3872
Joined: Tue Nov 18, 2008 9:26 pm

Re: European Club Cup

Post by Jonathan Rogers » Fri Oct 25, 2013 12:00 pm

MartinCarpenter wrote: Still White Rose aren't doing badly for a 5 handed team :)
Very true, they have got on with their task admirably.

MartinCarpenter
Posts: 2439
Joined: Tue May 24, 2011 10:58 am

Re: European Club Cup

Post by MartinCarpenter » Fri Oct 25, 2013 12:22 pm

Interesting pairing regulations those. After round 1 they're seemingly using net board points as the seeding within a score group. Why they're doing that - as opposed to straight up initial strength as in round 1- in a competition that is principally working on match points isn't exactly obvious.

Especially given the semi acceleration in the first round where team 1 plays the team just past halfway and so on down. That seems almost calculated to start the net number of game points off rather randomised.

A short report just appeared on Yorkshire chess which inter alia confirms that Gayson was ill and so missed his flight and that they couldn't register Ihor as a replacement for whatever reason. Maybe simply far too sensible a thing to allow :)

Jonathan Rogers
Posts: 3872
Joined: Tue Nov 18, 2008 9:26 pm

Re: European Club Cup

Post by Jonathan Rogers » Fri Oct 25, 2013 12:28 pm

MartinCarpenter wrote:Interesting pairing regulations those. After round 1 they're seemingly using net board points as the seeding within a score group. Why they're doing that - as opposed to straight up initial strength as in round 1- in a competition that is principally working on match points isn't exactly obvious.

Especially given the semi acceleration in the first round where team 1 plays the team just past halfway and so on down. That seems almost calculated to start the net number of game points off rather randomised.
and more to the point, are they doing it anyway? That was not my impression in round 2, when Barbican (who managed a draw against a powerhouse in round one) was still playing one lower board than Jutes - who are seeded one place higher but had lost all their games against their powerhouse round one opposition. So then I thought it was just seedings within the score group, but round 3 put paid to that theory as well.

Alex Holowczak
Posts: 8909
Joined: Sat May 30, 2009 5:18 pm
Location: Oldbury, Worcestershire
Contact:

Re: European Club Cup

Post by Alex Holowczak » Fri Oct 25, 2013 1:48 pm

I probably have more experience of organising 4-player Swiss tournaments than most people in Britain, having run 4 BUCAs and 1 J4NCL.

For the BUCA event, we used two different formats for pairing:
(1) Sort by matchpoints, gamepoints, seed; pair scoregroups on matchpoints
(2) Sort by matchpoints, seed; pair scoregroups on matchpoints

I found (2) to be better than (1).

Let's take the example with the Olympiad. In round 1, you have (say) 128 teams (number deliberately chosen because it's base 2!), and you have 1 v 65, 66 v 2, and so on. It's entirely likely that Russia will beat whoever is seeded 65 by 4-0, and in the last Olympiad, about half of the teams managed 4-0 wins in round 1. The rest won by 3.5-0.5 or less; there were few draws.

Suppose exactly half of the teams win 4-0. Under system (1), all of the 4-0 winners will play someone who didn't win 4-0 in round 1. So if the number 1 seed wins 4-0, and the number 2 seed only wins 3.5-0.5 in round 1, then all of a sudden, you have a round 2 pairing of 2 v 1. Personally, I'd rather avoid the possibility of the number 2 seed meeting the number 1 seed in round 2 of an 11 round event.

In the last Olympiad, just over half the teams won 4-0, which meant that the top few seeds ended up with "easy" round 2 draws, against teams seeded 50+, but the first non 4-0 winner ended up playing another relatively top team. While you could argue that winning 3.5-0.5 as opposed to 4-0 shouldn't give that team an advantage, what advantage has the team that won 4-0 gained?

It was against this argument that two years ago, I switched from method (1) to method (2).

In the last European Team Championship, and the Olympiad, I tried to work out the pairings on here without first looking the rules up (indeed, I tried to work the rules out!). I got most of the European Team Championship wrong, but I got the Olympiad ones correct. This possibly explains why a casual observer can't work out the ECU pairing rules. Having since read them, including just now when they were linked, I think I can understand what the rules are, but I can't understand why they're so non-standard. I don't understand some of the underlying logic, such as pairing a scoregroup of n players as 1 v n, 2 v n-1, 3 v n-2 (etc.); and then floating the median.

MartinCarpenter
Posts: 2439
Joined: Tue May 24, 2011 10:58 am

Re: European Club Cup

Post by MartinCarpenter » Fri Oct 25, 2013 2:08 pm

From checking, it looks like doing the GP thing as promised. Barbican were the median team on 0/1 MPs after round 1 and Jutes bottom half. You then played the team next to you rated by GP's and Jutes someone from the top half.

They may be ordering which board each match is on according to seeding order.

LawrenceCooper
Posts: 5049
Joined: Tue Dec 20, 2011 8:13 am

Re: European Club Cup

Post by LawrenceCooper » Fri Oct 25, 2013 5:48 pm

Round 6 on 2013/10/25 at 15:00

Bo. 5 G-Team Novy Bor Rtg - 1 SOCAR Rtg 3.5-2.5
1.1 GM Navara, David 2703 - GM Caruana, Fabiano 2779 1-0
1.2 GM Wojtaszek, Radoslaw 2698 - GM Topalov, Veselin 2771 1-0
1.3 GM Laznicka, Viktor 2666 - GM Kamsky, Gata 2725 1-0
1.4 GM Sasikiran, Krishnan 2662 - GM Mamedyarov, Shakhriyar 2759 DRAW
1.5 GM Hracek, Zbynek 2636 - GM Wang, Hao 2733 0-1
1.6 GM Bartel, Mateusz 2638 - GM Giri, Anish 2749 1-0

The leaders and top seeds lose the top three boards to go down 3.5-2.5 and allow the fifth seeds to take a one match point lead into the final round.

Jonathan Rogers
Posts: 3872
Joined: Tue Nov 18, 2008 9:26 pm

Re: European Club Cup

Post by Jonathan Rogers » Fri Oct 25, 2013 6:25 pm

Hopefully the last time I'll moan about our pairings* but I see that the three teams who beat us in this event each won today on boards 3, 5 and 9 ! And the Turkish team whom we beat won on board 10 - so maybe four of our seven opponents will finish in the top ten!

* It won't be

Roger de Coverly
Posts: 18101
Joined: Tue Apr 15, 2008 2:51 pm

Re: European Club Cup

Post by Roger de Coverly » Fri Oct 25, 2013 7:18 pm

Alex Holowczak wrote: (2) Sort by matchpoints, seed; pair scoregroups on matchpoints

I found (2) to be better than (1).
Is that the method used for 4NCL Division 3? Tiebreaks now place Sum of Opponents over game points, for the reason that a team meeting the strongest teams in the League is likely to have narrow victories if at all and thus a modest board point total, whereas a team with an early set back can run up high board point tallies.

Alex Holowczak
Posts: 8909
Joined: Sat May 30, 2009 5:18 pm
Location: Oldbury, Worcestershire
Contact:

Re: European Club Cup

Post by Alex Holowczak » Fri Oct 25, 2013 7:23 pm

Roger de Coverly wrote:
Alex Holowczak wrote: (2) Sort by matchpoints, seed; pair scoregroups on matchpoints

I found (2) to be better than (1).
Is that the method used for 4NCL Division 3? Tiebreaks now place Sum of Opponents over game points, for the reason that a team meeting the strongest teams in the League is likely to have narrow victories if at all and thus a modest board point total, whereas a team with an early set back can run up high board point tallies.
There aren't any pairing rules for 4NCL Division 3, unless the number of teams on a scoregroup is a multiple of 4.

Bet that wasn't the answer you were expecting...!

Andrew Bak
Posts: 833
Joined: Sun Jan 02, 2011 11:48 am
Location: Bradford
Contact:

Re: European Club Cup

Post by Andrew Bak » Sat Oct 26, 2013 3:13 am

MartinCarpenter wrote:Interesting pairing regulations those. After round 1 they're seemingly using net board points as the seeding within a score group. Why they're doing that - as opposed to straight up initial strength as in round 1- in a competition that is principally working on match points isn't exactly obvious.

Especially given the semi acceleration in the first round where team 1 plays the team just past halfway and so on down. That seems almost calculated to start the net number of game points off rather randomised.

A short report just appeared on Yorkshire chess which inter alia confirms that Gayson was ill and so missed his flight and that they couldn't register Ihor as a replacement for whatever reason. Maybe simply far too sensible a thing to allow :)
I've not been able to ascertain many details other than what I posted to the Yorkshire site.

Peter was on Facebook cheering the guys on so he fortunately doesn't seem to be too seriously ill. He even hinted that he might be able to make it out for the final few rounds but this doesn't seem to have materialised.

Ihor Lewyk and Rich Archer had no intention of playing and just fancied a week away in Greece which is why they didn't pay the £100 each to register. FIDE's registration rules have ensured that seven players will have a day off they otherwise wouldn't have had!

benedgell
Posts: 1252
Joined: Thu Apr 05, 2007 12:43 pm
Location: Somerset
Contact:

Re: European Club Cup

Post by benedgell » Sat Oct 26, 2013 10:54 am

Best of luck to Hjorvar, who play for Jutes, today, who needs a win for a GM norm. According to the unofficial list on the website, he's the only person from any of the English teams in with a shout of getting a norm today.

Chris Rice
Posts: 2774
Joined: Thu Nov 08, 2012 5:17 am

Re: European Club Cup

Post by Chris Rice » Sat Oct 26, 2013 11:04 am

benedgell wrote:Best of luck to Hjorvar, who play for Jutes, today, who needs a win for a GM norm. According to the unofficial list on the website, he's the only person from any of the English teams in with a shout of getting a norm today.
If he wins this will be his final norm and will secure him the Grandmaster title. We are all more nervous than he is!

Jonathan Rogers
Posts: 3872
Joined: Tue Nov 18, 2008 9:26 pm

Re: European Club Cup

Post by Jonathan Rogers » Sat Oct 26, 2013 11:05 am

Andrew Bak wrote:
Ihor Lewyk and Rich Archer had no intention of playing and just fancied a week away in Greece which is why they didn't pay the £100 each to register. FIDE's registration rules have ensured that seven players will have a day off they otherwise wouldn't have had!
It is only 55 Euros per person to register, so being prepared just to spend an extra £40 to register one player as a reserve would have enabled White Rose to play a full team throughout the event. It was a crass decision not to do so, and not much better to blame ECU regulations!

Not that the team could have been doing any better (matchpoint-wise) if it did have a full team. Their board two has been inspired and combined with highly favourable pairings they have done very well. (But the fact that their board two is on 4.5/6 and has no chance of a GM norm does say a lot about their pairings).

Post Reply