2009 British Open

MoominDave

Well-Known Member
Again with the knee-jerk naysaying, albeit from someone else...

All judges would listen to all bands, and rank all of them in their own way, concentrating on top or bottom as appointed. Whether or not the full ranking was published would be totally independent of whether or not such a ranking was made.
 

mclaugh

Member
Again with the knee-jerk naysaying, albeit from someone else...
Since you obviously have thought through all the possible scenarios that can arise in the system you propose to know that it's workable and that it removes the possible sources of controversey, please resolve the following scenario, providing the rationale for your answers:

Bands A, C, and H are in contention for spots 5 and 6 in the top 6.

Judge 1 ranks Band A 8th, Band C 6th, and Band H 5th;
Judge 2 ranks Band A 5th, Band C 8th, and Band H 6th;
Judge 3 ranks Band A 6th, Band C 5th, and band H 8th.

All three judges feel strongly that there is a significant gap in the quality of performance between the bands they ranked 6th and 8th.

Questions:

Which band gets dropped from the top 6?
Which of the two bands that make the cut finishes 5th?

As a group, the judges agree to drop Band C.

Judge 1 is assigned with "concentrating" on the top 6 bands. By what criterion or criteria does Judge 1 decide whether to award Band A or Band H 5th place?

----

This scenario, or one very similar to it, arises with great regularity in district, state, and regional music competitions in the US; and, by putting the primary responsibility for determining the final rankings in the hands of one individual, creates more charges of bias and favoritism than previously.
 

MoominDave

Well-Known Member
Since you obviously have thought through all the possible scenarios that can arise in the system you propose to know that it's workable and that it removes the possible sources of controversey
Enough with the sarcasm, please. I made it abundantly clear in my first post on this subject in this thread that I was offering up an idea that had only just occurred to me.
I also suggest following the link I provided a post or two back that outlined what the perfectionist fallacy is - this sentence indicates that you've fallen into it.

[Apologies for the long post here - I hope it interests someone out there...]

If I were to implement such a system, I would employ two adjudicators, Judge X and Judge Y; Judge X would be tasked with concentrating on making sure that the highest places were as accurate as possible, while Judge Y would be tasked with concentrating on making sure that the lowest places were as accurate as possible. The two adjudicators would be independent - it would be inappropriate for them to confer as they would not be expected to perform quite the same task, and a degree of variation between their results would be expected.

I would produce a final list of placings based on a weighted combination of the placings of the two judges; the highest placings would reflect very closely the opinion of Judge X, and the lowest placings would reflect very closely the opinion of Judge Y; the middle placings would reflect a more even combination of their respective opinions.

There are various possible weightings, and picking one that suits is not a totally simple task. My first thought was to weight linearly from top to bottom from one to zero (and vice versa for the other judge), but under this scheme it would be possible for a band to win by being placed last by X and first by Y - totally unacceptable! So a less powerful weighting differential would seem appropriate. A better system might be to weight linearly between weight values of say 1 and 3 - then each adjudicator's area of concentration would carry three times the weight of the other adjudicator's 'neglected' area.

So under that scheme, in a field of n bands, Judge X's kth place would carry a weight of 1+2*(n-k)/(n-1), while Judge Y's kth place would carry a weight of 1+2*(k-1)/(n-1). In order to normalise the total amount of weight received between the two judges, I would then divide the summed weighted value by this to obtain a weighted normalised placing.

Here's a concrete example for a field of six bands, A, B, C, D, E, and F - Suppose Judge X ranks them from top to bottom ABCDEF, while Judge Y ranks them from top to bottom BEACDF, a not implausible reordering of the list in something as subjective as musical adjudication, with the discrepancy in E's placing being the obvious big difference. Here are how the weighted placings from the two judges would work out:

[N.B. Summary given below if you want to skip the numbers...]

Band A: Judge X 1st, Judge Y 3rd
X weight = 1+2*(6-1)/(6-1) = 1+2*5/5 = 3
Y weight = 1+2*(3-1)/(6-1) = 1+2*2/5 = 1.8
Total weight: 3 + 1.8 = 4.8
Band A's summed weighted value: 3*1 + 1.8*2 = 3 + 3.6 = 6.6
Band A's weighted placing: 6.6/4.8 ~= 1.38

Band B: Judge X 2nd, Judge Y 1st
X weight = 1+2*(6-2)/(6-1) = 1+2*4/5 = 2.6
Y weight = 1+2*(1-1)/(6-1) = 1+2*0/5 = 1
Total weight: 2.6 + 1 = 3.6
Band B's summed weighted value: 2.6*2 + 1*1 = 5.2 + 1 = 6.2
Band B's weighted placing: 6.2/3.6 ~= 1.72

Band C: Judge X 3rd, Judge Y 4th
X weight = 1+2*(6-3)/(6-1) = 1+2*3/5 = 2.2
Y weight = 1+2*(4-1)/(6-1) = 1+2*3/5 = 2.2
Total weight: 2.2 + 2.2 = 4.4
Band C's summed weighted value: 2.2*3 + 2.2*4 = 6.6 + 8.8 = 15.4
Band C's weighted placing: 15.4/4.4 = 3.5

Band D: Judge X 4th, Judge Y 5th
X weight = 1+2*(6-4)/(6-1) = 1+2*2/5 = 1.8
Y weight = 1+2*(5-1)/(6-1) = 1+2*4/5 = 2.6
Total weight: 1.8 + 2.6 = 4.4
Band D's summed weighted value: 1.8*4 + 2.6*5 = 7.2 + 13.0 = 20.2
Band D's weighted placing: 20.2/4.4 ~= 4.59

Band E: Judge X 5th, Judge Y 2nd
X weight = 1+2*(6-5)/(6-1) = 1+2*1/5 = 1.4
Y weight = 1+2*(2-1)/(6-1) = 1+2*1/5 = 1.4
Total weight: 1.4 + 1.4 = 2.8
Band E's summed weighted value: 1.4*5 + 1.4*2 = 7.0 + 2.8 = 9.8
Band E's weighted placing: 9.8/2.8 = 3.5

Band F: Judge X 6th, Judge Y 6th
X weight = 1+2*(6-6)/(6-1) = 1+2*0/5 = 0
Y weight = 1+2*(6-1)/(6-1) = 1+2*5/5 = 3
Total weight: 0 + 3 = 3
Band F's summed weighted value: 0*6 + 3*6 = 0 + 18 = 18
Band F's weighted placing: 18/3 = 6.0

In summary, we have the following bands, X results, Y results, and weighted results:
A: 1, 3; 1.38
B: 2, 1; 1.72
C: 3, 4; 3.5
D: 4, 5; 4.59
E: 5, 2; 3.5
F: 6,6; 6.0
which would give the following ranking of placings:
1: A
2: B
3=: C
3=: E
5: D
6: F

You can see that the opinions of Judge X are more strongly reflected at the top of the table while the opinions of Judge Y are more strongly reflected at the bottom of the table, and the bands in the middle reflect a mixture of their opinions (although there is not much middle in this particular example, for simplicity's sake). In particular, note that A beats B, despite obtaining a 1st and a 3rd vs B's 2nd and a 1st - because A's 1st came from the judge tasked with sorting out the top accurately.

Although I'm sure the parameters could be tweaked to improve things, this arrangement seems quite robust and generally fair to me, and avoids the problems that you outlined in your hypothetical scenario, which made certain assumptions about the way the process would work which I would avoid if I were setting it up.
 
Last edited:

RamasII

Member
Hi Dave,

Why not just have 5 or 6 of them just looking at how an individual section has played and giving marks out of 20...then its a simple job of adding it up...its loads easier just to follow a group of 5 parts than it is for the whole score...as things get missed...
 
I've never posted a private message before, but I thought people would be interested in the following message and my reply.

David James

P.S. I note he has not replied to the questions I put in my original post.


Re: From Steven Mead

--------------------------------------------------------------------------------

Quote:
Originally Posted by SMead
(deleted)

Steven,

You have brought it on yourself with what you did, and your apology on your web site is poor in the extreme. Why no apology on this site?
To quote HM again, "go forth and multiply", I will speak my mind without fear or favour.
David James
 
Last edited by a moderator:

MoominDave

Well-Known Member
Mr James,

I have great respect for you and your long and impressive musical career, and am withholding personal judgment for now on the rights and wrongs of your spat with Mr Mead. However, uninvitedly posting personal messages in public is rude. You know that, I know that, we all know it.

Hi Dave,

Why not just have 5 or 6 of them just looking at how an individual section has played and giving marks out of 20...then its a simple job of adding it up...its loads easier just to follow a group of 5 parts than it is for the whole score...as things get missed...
Monsieur Pharaoh!

Could do... You'd need somebody with an 'overview' role too - someone to look at how sections balanced together, at how well the interpretation worked... But what if a particular piece had jack-all for one section in? Would the judge of their section still count for as much pointage as other sections that were more critical? I think this would be rather difficult to implement fairly.
 
Last edited:

RamasII

Member
yes, just having someone for Troms&Basses, Percs, EuphBaris, HornsFlug, Cornets and one for interp...assesing the conductor..This would obviously be for the big contests..dont think the £'s would stretch that far for local stuff etc..but certainly where the standard from top to bottom is good im sure it would be an interesting way of doing things..But I follow the fact that some pieces have more in than others..but what ever is written can still be marked for clarity and performance etc....?
 

MoominDave

Well-Known Member
Results by Excel...haven't they only just got rubbers for the pencils? ;)
But doesn't Microsoft's internal coding predate the wheel?

Edit: But seriously, yeah... Just feed the lists of placings into a spreadsheet, and let it do the calculations. Barely more difficult than what gets done now.
 
Last edited:
Mr James,

I have great respect for you and your long and impressive musical career, and am withholding personal judgment for now on the rights and wrongs of your spat with Mr Mead. However, uninvitedly posting personal messages in public is rude. You know that, I know that, we all know it.
.
That's your opinion, and you're entitled to it. I've not seen it written in tablets of stone whereby it is rude to post private messages. You write as if I am the only one complaining about Mr Meads behaviour, I suggest you take a look at the comments section of 4barsrest.
There again it might have been my intention to be rude, I suggest he'll not send me anymore private messages as I would do the same again.

David James
 

MoominDave

Well-Known Member
It's basic courtesy, no? When you write here, you publish to the world. When you PM, you communicate privately. You wouldn't forward a personal letter on to a newspaper letters page without first consulting the author, would you?
 
It's basic courtesy, no? When you write here, you publish to the world. When you PM, you communicate privately. You wouldn't forward a personal letter on to a newspaper letters page without first consulting the author, would you?
I don't think he's shown any courtesy to the adjudicators, the management of the British Open, and has made a number of unsubstantiated comments about the adjudicators hearing and mental faculties which are tantamount to harassment or victimisation. Furthermore what protection do the adjudicators get from his and others postings which have travelled globally for all to read on this forum.
David James
 

MoominDave

Well-Known Member
Well, now, these are some of the points at issue here. To my mind, they are some of the last interesting of the points that are at issue, but people seem to find them worthy of discussion...

But ask yourself honestly - are you helping matters by introducing discourtesy of your own? Two wrongs, etc., etc., etc. ...
 
Well, now, these are some of the points at issue here. To my mind, they are some of the last interesting of the points that are at issue, but people seem to find them worthy of discussion...

But ask yourself honestly - are you helping matters by introducing discourtesy of your own? Two wrongs, etc., etc., etc. ...
Well, now, Mr Taylor, you might think they are the least (I believe thats what you meant to write, not last) interesting points that are at issue, but I and a number of others believe them to be be extremely relevant.
In all honesty Mr Taylor my discourtesy pales into insignificance in comparison to Mr Meads.

I see that you are a mathematician hence the two wrongs, and there I thought binary code was one plus one...............................
I am tempted to ask myself if you're his barrister or his barrack room lawyer, but then as a numerical analyst for the UK Atomic Energy Authority I fear you might be tempted to nuke me ............

David James
 

steve butler

Active Member
Well, now, Mr Taylor, you might think they are the least (I believe thats what you meant to write, not last) interesting points that are at issue, but I and a number of others believe them to be be extremely relevant.
In all honesty Mr Taylor my discourtesy pales into insignificance in comparison to Mr Meads.

I see that you are a mathematician hence the two wrongs, and there I thought binary code was one plus one...............................
I am tempted to ask myself if you're his barrister or his barrack room lawyer, but then as a numerical analyst for the UK Atomic Energy Authority I fear you might be tempted to nuke me ............

David James
I think you should change your user name to "super grass"
 

MoominDave

Well-Known Member
Haha, no, we're all about the peaceful use of atomic energy... Though the original designer of the British hydrogen bomb does work on the same corridor as I do...! He's 80 now, but still one of the sharpest people I've ever met. I hope you enjoyed my website ;-)

I've said my piece, and it was probably influenced by somebody on here recently maliciously posting info from a PM that I'd sent them. It's a sin, but you're well capable of defending it here.

'Least', not 'last', yes, my typo, whoops... It's a pity that the discussions of Mr Mead's ideas and of his motives seem to me inseparable. The meat of the discussion has sometimes been lost in talk about the man.

Peace on earth and goodwill to all men.
 
Haha, no, we're all about the peaceful use of atomic energy... Though the original designer of the British hydrogen bomb does work on the same corridor as I do...! He's 80 now, but still one of the sharpest people I've ever met. I hope you enjoyed my website ;-)

I've said my piece, and it was probably influenced by somebody on here recently maliciously posting info from a PM that I'd sent them. It's a sin, but you're well capable of defending it here.

'Least', not 'last', yes, my typo, whoops... It's a pity that the discussions of Mr Mead's ideas and of his motives seem to me inseparable. The meat of the discussion has sometimes been lost in talk about the man.

Peace on earth and goodwill to all men.
You see, you can still have all your marbles at 80.


Peace on earth and goodwill to all men, I say Amen to that!

David James
 
Top