Brass Band Logo

NJH Music Logo

Some of the contents of the pages on this site are Copyright © 2016 NJH Music
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: April Rankings


At 11:17 10-04-96 GMT, you wrote:
> The BBW rankings make interesting reading.  I presume the list is based on 
> contest results; can anyone offer any details of how this is done?
>
> Don Cooper
>

Here is the (somewhat extensive) explanation of the Brass Band World Ranking
system as published in their latest issue. Unfortunately Table 3 is missing,
because it's difficult to paste this in a message.

At your service,
Eric de Cloe (solo euphonium, Brassband Schoonhoven)

In a number of previous
occasions I have attempted to
describe the workings of the
BBW rankings system. The
recent publication of the list of
"ranking events" in 1995 seems
have stimulated extra interest in
the system so perhaps a fresh
explanation is overdue.

Regular readers will be familiar
with the basic concept. The system is
loosely based on the Sony Rankings
for professional golfers which was
introduced in 1986, i.e. bands are
ranked by the weighted average of
ranking points allocated over a rolling
36-month period. To further
describe the system in words alone
would be rather dull so, instead, I will
illustrate the system by taking an
actual contest, the recent South
Somerset Entertainment contest, as a
working example.

 Step I: It must first be established
that a contest is a so-called "ranking
event". For 1996, a contest is a
"ranking event" if it features at least
one band currently ranked in the top-
45. The South Somerset contest
comfortably met this requirement
with 10 "ranked" bands competing.
 (NB. This feature of the system is
constantly evolving and for 1997 1 will
be looking at the top-SO to
determine "ranking events").

 Step 2: The competing bands and
their finishing positions are shown in
Table I. Note that the actual points
awarded by the adjudicator(s) are
completely irrelevant for ranking
purposes.

We then need to refer to Table 2
in order to allocate "ratings points"
according to the respective ranks of
the competing bands. For example.
Tredegar (ranked 15th at the time of
this contest) carry with them 45
ratings points, Camborne (ranked
43rd)- one ratings point, and so on.
The total of the ratings points for
the contest is 196. This is a key
statistic as it serves to identify the
category for ranking points purposes.
>From Table 3, it can be seen that the
contest was classified as category 10.
The sum of the ratings points (in
this case, 196) has a dual-purpose as
it effectively also determines the
"strength of field" of the contest.
Indeed, this is how I identified the
top-l0 "ranking events" for 1995 as
highlighted in the recent table.

 Step 3: Having established the
ranking category it is then a
straightforward task to allocate
ranking points" to each of the
competing bands. This allocation is
shown in Table I.
I should point out that the extract
shown (Table 3) applies only when
there are 10 or more competing
bands. When fewer than 10 bands
compete, the "ranking points"
awarded area scaled down accordingly.

 Step 4: 1 mentioned earlier that
bands are ranked by the weighted
average of their ranking points. This
feature is quite complex but basically
the weights are designed such that the
average weight in months 13-24 is
three times greater than the average
weight in months 1-12. Similarly, the
weights in months 25-36 are three
times greater on average than the
weights in months 13-24. Each
month's weight is also higher than that
for the previous month.

I won't bore you with any more of
the detail, suffice to say that the
average weights for each 12-month
period are 1/2, 1'/2 and 4'/2
respectively and the weight for
month 36, February 1996 in our
example, is 6.63. The weighted
ranking points (i.e. rankings point x
6.63) are shown in Table I.

 Step 5: We have now reached the
final stage in the process. The total of
all weighted ranking points awarded
over the 36-month period is simply
divided by the number of contests
played over the same period to give
the average ranking points as
published.
 (NB. The number of contests
played, the so-called 'divisor', is
subject to a minimum of 12).

 We can examine the progression
of average points from month to
month by looking at the effect that
the South Somerset contest had on
Tredegar's ranking.

Total ranking points over the 36-
month period to 1/2/96 = 388.80*
 Number of contests played - 27
 .'. Points average = 14.40
 Consider now the 36-month
 period to 1/3/96:
 Total ranking points (excluding the
South Somerset contest) = 355.32*
 *As Tredegar did not compete in
February 1993, these totals are
derived from the some 27 contests.
The reduction illustrates the effect of
the weights and the rolling 36-month
period.
 .'. Points average at 113/96 (now
including the South Somerset contest)
is (355.32+79.56)\28 = 15.53
 So there we have it, not
straightforward I know. Fortunately,
my trusty computer is programmed
such that I don't need to think about
the mechanics of the system too often.
 I hope bandspeople will continue
to enjoy the BBW rankings (they are
meant to be a form of practical help
and entertainment - not to be taken
too seriously!), especially now that
they have a greater understanding
both of what constitutes a "ranking
event" and the system in general.
Furthermore. we know that the BBW computer
3.4 rankings have frequently assisted bands in their negotiations for
sponsorships and other forms of support and recognition.
So, in that sense, the BBW computer rankings are helping bands to help
themselves. Lastly, a plea for your support. If you know that your band
has competed in a "ranking event", the result of which hasn't appeared in
BBW, then please let the magazine know the full result as soon as possible
after the event. Only then can it be guaranteed that "ranking points" have
been correctly allocated for this contest.

			COLIN ARCHIBALD

South Somerset Entertainment Contest - 5th February 1996

			Rank				Ratings
Ranking Weighted
					    @ 112196	Points    Points
Ranking Points

Ist Tredegar	    15				  45	12.0
79.56
2nd Camborne	    43				  1	 9.4
62.32
3rd DVTYorkrlmpr	12				  54	7.3
4_8.40
4th Cor/		21				  28	5.7
37.79
5th Woodfalls	   25				  20	4.4
29.17
6th WmDavisConC	 28				  14	3.5
23.2 1
7th Bournemouth Con     55					    2.7
17.90
8th BTM		 22				  26	2.1
13.92
9th Cwmaman	     37				  4	 1.6
10.6 1
IOth Hendon	     40				  2	 1.3
8.62
Ilth Sovereign Brass    42				  2	 1.0
6.63
12th Friary Cuildford   134				 -	 0.8
5.30
13th Regent Brass       67				  -	 0.6
3.98
196

						    Table I
Ratings Points by Pank

      100    26    18
2     95     27    16
3     90     28    14
4     85     29    12
5     80     30    10
6     76     31    9
7     72     32    8
8     68     33    7
9     64     34    6
10    60     35    5
11    57     36    4
12    54     37    4
13    51     38    3
14    48     39    3
15    45     40    2
16    42     41    2
17    39     42    2
18    36     43
19    33     44
20    30     45
21    28
22    26
23    24
24    22
25    20
Table 2


--
unsubscribe or receive the list in digest form, mail a message of 'help' to
listserv@xxxxxxxxxxxxxxxxxxx

[Services] [Contact Us] [Advertise with us] [About] [Tell a friend about us] [Copyright © 2016 NJH Music]