4NCL North 2014-15

Venues, fixtures, teams and related matters.
Post Reply
Sean Hewitt
Posts: 2193
Joined: Sat Mar 10, 2012 8:18 pm
Contact:

Re: 4NCL North 2014-15

Post by Sean Hewitt » Mon Nov 03, 2014 8:41 pm

David Robertson wrote:Triangular Match

Another one for Alex here. As far as I can tell, the pairings for these matches result in a 4-2 colour imbalance in all individual matches. All teams seem to achieve a 6-6 balance over the weekend, but not in the individual match. So, for example, Spirit of Atticus receive only two Whites v. Holmes Chapel, but four v. Jorvik, who in turn receive four Whites v. Holmes Chapel. When amplified by up/down floats, this gets distorting.

Unavoidable? I've no idea. If so, we'll just have to rub along with it. But surely a tweak is possible.
Unavoidable as far as I'm aware.

User avatar
Rob Thompson
Posts: 757
Joined: Wed Jul 15, 2009 12:03 pm
Location: Behind you

Re: 4NCL North 2014-15

Post by Rob Thompson » Mon Nov 03, 2014 9:19 pm

I think it comes from the fact that there are games between players on different boards.
True glory lies in doing what deserves to be written; in writing what deserves to be read.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Mon Nov 03, 2014 9:45 pm

I'm intrigued - what breaks if you switch one of the sets of colours round in each match to keep it 3-3 in each?

I can only think it might be avoiding giving individual players a double white/black on the same board. Would have thought that less important here. The teams aren't very likely to be maintaining the same board order, or even players in our case, for the two matches so some people will be getting doubled up colours anyway.

Sean Hewitt
Posts: 2193
Joined: Sat Mar 10, 2012 8:18 pm
Contact:

Re: 4NCL North 2014-15

Post by Sean Hewitt » Mon Nov 03, 2014 9:59 pm

MartinCarpenter wrote:I'm intrigued - what breaks if you switch one of the sets of colours round in each match to keep it 3-3 in each?
Which pair of fixtures do you switch? :D

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Mon Nov 03, 2014 10:29 pm

No need to switch whole fixtures. You just need to reverse the colours on one board of each match to keep the 3-3 balance.

I realise it'll probably break something else :) Either/both of A1 getting a white and a black over the weekend or the 3-3 colour balance on each day.

Still, I'd have thought keeping the colour balance within each actual match should be the overriding priority in a team competition.
(Especially when the set of players/board order in each team isn't forced to be the same for both matches.).

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: 4NCL North 2014-15

Post by Roger de Coverly » Tue Nov 04, 2014 10:50 am

MartinCarpenter wrote:No need to switch whole fixtures. You just need to reverse the colours on one board of each match to keep the 3-3 balance.
From http://www.4ncl.co.uk/triangular_matches.htm

1 A1 - B1 B1 - C1
2 C1 - A2 C2 - A1
3 B2 - C2 A2 - B2

4 B3 - A3 C3 - B3
5 A4 - C3 A3 - C4
6 C4 - B4 B4 - A4

7 A5 - B5 B5 - C5
8 C5 - A6 C6 - A5
9 B6 - C6 A6 - B6

So 7,8,9 are a repeat of 1,2,3 with colours reversed on 4,5,6

This has the feature that on each day, three of each team have the White pieces.

But if you line up the implicit matches, this is what you get

A1-B1
A2-B2
B3-A3
B4-A4
A5-B5
A6-B6

C2-A1
C1-A2
A3-C4
A4-C3
C6-A5
C5-A6

B1-C1
B2-C2
C3-B3
C4-B4
B5-C5
B6-C6

It doesn't matter whether you swop on implied 1,2,5 or 6 but suppose 5 is used.

The pairings then become

1 A1 - B1 B1 - C1
2 C1 - A2 C2 - A1
3 B2 - C2 A2 - B2

4 B3 - A3 C3 - B3
5 A4 - C3 A3 - C4
6 C4 - B4 B4 - A4

7 B5 - A5 C5 - B5
8 C5 - A6 A5 - C6
9 B6 - C6 A6 - B6

If board orders are left unchanged C5 will get two Whites against C6 with two Blacks. There's also the potentially undesirable feature that if they move the players for colour alternation, then those in team C face two from team A or two from team B.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Tue Nov 04, 2014 11:18 am

Yes, I know there's no absolutely ideal answer!

Surely, though, we can all agree that in a teams competition balancing the team colours within each individual match should be the most important priority in terms of colours? Integrity of the competition primary.

I'm not annoyed, because this isn't important enough for that. I am surprised and rather intrigued because it doesn't obviously make sense, and these organisers are very sensible people :)

Have they tried it the other way round and had major complaints from people getting double colours? Or is there some sort of over riding organisational advantage to doing things this way round?

(In fact Div3N teams often have board order flexbility of at least +-2 boards, so could mitigate some imablances in the colours people play over the weekend while mantaining the very important 'only playing vs one team' thing. Obviously can't do anything about this sort of in built imbalance in the matches.).

Alex Holowczak
Posts: 9085
Joined: Sat May 30, 2009 5:18 pm
Location: Oldbury, Worcestershire
Contact:

Re: 4NCL North 2014-15

Post by Alex Holowczak » Tue Nov 04, 2014 2:44 pm

David Robertson wrote:Triangular Match

Another one for Alex here. As far as I can tell, the pairings for these matches result in a 4-2 colour imbalance in all individual matches. All teams seem to achieve a 6-6 balance over the weekend, but not in the individual match. So, for example, Spirit of Atticus receive only two Whites v. Holmes Chapel, but four v. Jorvik, who in turn receive four Whites v. Holmes Chapel. When amplified by up/down floats, this gets distorting.

Unavoidable? I've no idea. If so, we'll just have to rub along with it. But surely a tweak is possible.
I'm afraid a tweak isn't possible to create a 3-3 colour split. It is even possible to have a player playing two boards of the same match if you're not on the ball - this happened in a J4NCL triangular match once for this reason. :oops:

The arbiters realise that triangular matches aren't ideal, which is why they normally give them to the bottom-most scoregroup, and why they try to avoid promotion candidates playing in them.

The alternative is a team sitting out a round with a bye, and I suspect this is by far the lesser of the two evils.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Tue Nov 04, 2014 3:10 pm

I'm sorry to go on about this, but there is a very simple tweak indeed - you play as now but turn one of the boards in each match the other way round to switch the colours :)
(And yes triangular matches are a mess anyway but much better than a bye!).

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: 4NCL North 2014-15

Post by Roger de Coverly » Tue Nov 04, 2014 3:23 pm

Alex Holowczak wrote: The arbiters realise that triangular matches aren't ideal, which is why they normally give them to the bottom-most scoregroup, and why they try to avoid promotion candidates playing in them.
That hasn't been done for the first weekend where there could be at least a supposition that three mid-table teams are involved. Presumably there's an anticipation that the various second and third teams will be involved in the other two weekends and that the rule about only one triangular match a season is sacrosanct.

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Tue Nov 04, 2014 4:42 pm

iirc There was an earlier note in one of the posts about the rule about one triangular match a season not being certain to stand this year? To be honest there really isn't room for them to sensibly make that promise.

They've got 4 (semi) plausible promotion candidates, (NE1, Cheddleton 2, MM1, DCA A), 4 likely weaker teams (DCA C, Chedd 3, MM3 and NE2) and 5 midtable strength teams. 9 teams needed for the three triangulars. Only possible to avoid duplications if you assigned fully in advance, almost giving up on holding a swiss :)

More realistically, MM2 and Bradford B should probably both play to ~170(+) average this season while solid all the way down, should finish somewhere round 4th-6th and so don't ideally belong in a triangular for the obvious reasons.

The first weekend assignment to the triangular really is very sensible indeed. Its teams who are strictly in it for fun - two of whom mess their board order/players about all the time anyway! - and near certain to be solidly mid table somewhere.

The second weekend should be OK, the final one I really don't envy them one bit :(

David Robertson

Re: 4NCL North 2014-15

Post by David Robertson » Tue Nov 04, 2014 4:45 pm

Pareto optimality collides with Nash equilibria, I fear

A solution depends on a satisfactory trade-off between policy options: that is, between things we dislike, and things we'll tolerate. A Pareto optimum is achieved, albeit some way short of perfection, when no further changes improve all options. Nash equilibria remain to be resolved in our case here, something we tolerate.

So...

Conditions deemed to violate the sporting competition (ie. 'violations')

A) colour imbalances
B) playing the same team twice
C) White + downfloat

Conditions deemed not to violate the sporting competition (ie. 'tolerances')

D) playing the same colour twice
E) variable board order
F) White + upfloat

If we agree this classification of policy objectives, then certain outcomes follow:

* (A) can be satisfied by (D) which leaves (B) unsatisfied.
* (B) can be satisfied, but not by (D) or (E)
* hence, a choice must be made between (A) and (B). If (B) > (A) (as I would argue, just about), then (E) must be addressed. If (E) > (B) or > (A), we're in a mess. Hence (E) cannot hold.
* but (E) does hold, leaving open (B) and (D), causing (A)

Conclusion?

(E) is the Nash game colliding with a Pareto efficient outcome. The 4NCL rules for the triangular match are, absent (E), Pareto optimal. With (E), it's a potential mess

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Tue Nov 04, 2014 5:04 pm

But, the thing is that you can easily have A and B :) You just need keep everything else the same as it is now and switch the colours round on two of the boards.

I can only assume that D is being given precidence over A. Really no idea why, hence asking a few times.....

David Robertson

Re: 4NCL North 2014-15

Post by David Robertson » Tue Nov 04, 2014 5:38 pm

You can have (A) and (B). But you want neither - and that's not possible. I think (B) is a greater violation than (A). So once I rule against that, I must logically rule against conditions such as (E) which make it possible.

But (E) forms a legitimate part of some teams' strategy (eg. mucking up opponents' prep). That allows (B) back in. Awkward

MartinCarpenter
Posts: 3044
Joined: Tue May 24, 2011 10:58 am

Re: 4NCL North 2014-15

Post by MartinCarpenter » Tue Nov 04, 2014 7:50 pm

Oh, if you're putting it that way round then yes you can have neither or A or B happening. Its rather easy really - Roger shows how above. Take the current matches, sit everyone down, and then turn two the boards over the weekend round the other way :)

Variable board orders are a mess, but the basic rule is that you can do what you want with the board order but absolutely aren't allowed to play anyone against the same team twice. Even if means breaking the 80 grading pt limit. So long as people are sensibly forearmed that'll be OK. Jorvik are going to be using rather more than 6 players again, but we'll be fine cf no one playing the same twice.

Board orders horribly chopped up by the partitioning anyway of course.

Post Reply