Early Draw - Urban Myth or Fact?

Discussion in 'The Adjudicators' Comments' started by DaveBBb, Sep 9, 2014.

  1. DaveBBb

    DaveBBb Member

    It's often been said that there is an early draw disadvantage in contesting. I have often wondered if there is any truth in this urban myth. To explore the question I have done a preliminary analysis of the 'British Open Brass Band' contest results back to 1953 (some 60 years) using 'Brass Band Results' data. In that time only one band (Munn and Feltons, 1954) came 1st off draw number one. Taking together all of the results the best Draw to get is between 12 and 17. There you have a 43% chance of being in the top 5. You have a 10% chance coming in the top 5 places if you draw between 1 and 5. A 31% chance of being in the top 5 if you draw between 6 and 11. A 43% chance of being in the top 5 if you draw between 12 and 17 and a 16% chance of being in the top 5 if you draw between 18 and 23. So there you have it, regardless of the band, the conductor, the test piece, the adjudicator(s), if you want to stand roughly half a chance of being in the top 5 places hope that you are drawn anywhere between 12 and 17. I'll look at the Nationals next. DaveBBb
  2. Adamskied

    Adamskied Member

    Not sure if I am right but I think Friary Guildford band were the only qualifier for the Albert Hall of the number 1 draw. Please correct me if I am wrong.
  3. Ianroberts

    Ianroberts Well-Known Member

    For poorer bands I can understand them not wanting number 1, over the years bands I have played with have won many contests off number 1, Ifton won the midlands area last year, played number 1, played the national anthem like a big bame band, that rendition I am sure was in the adjudicators mind as well as the test piece.
  4. Draw number one is bad. Everyone knows that. You can win, but the odds are stacked against. The real question is why? Is it the band plays it safe, and lacks a bit of winning punch? Or are the ajudicators still waking up? Or does the national anthem upset all the republican bandsmen? Why, why, why!!
  5. Ianroberts

    Ianroberts Well-Known Member

    And that is why you are still on baritone ! :D:D
  6. DocFox

    DocFox Retired

    Judges are human. The first band to play MAY not get the highest score to "allow room" for the other bands to play.

    I think it was two years ago at NABBA (North American Brass Band Association) that the Number 1 won the Championship Section by a narrow margin. I believe if you draw an early number no mistakes will be overlooked. IF you play the best you will still win.

    I also believe a late numbered band can play at the same level (this is so objective -- who is really to say?) as the number 1 and the later number will finish higher. If you draw an early number, although you already are playing at peak level, you may have to be mistake free and unbelievable in areas of dynamics and intonation. In other words, if you draw an early number can, you turn it up even another notch (asking a lot at a contest) and make an unforgettable impression?

  7. MoominDave

    MoominDave Well-Known Member

    I suspect that the main difference comes down to human psychology in the minds of the judge(s). It is perceived as a special thing deserving particular attention when judging to determine which band has won - and so the question of which band is currently in the lead will always be in their minds. As each band submits its performance, it is compared against the current ranking, and if it is a good performance, the specific question must be asked: "Is this a good enough performance to displace the current first place?". A band that plays first must overcome the challenge of every subsequent band, whereas a band that plays last must only overcome the challenge of the current leader - and further, the early drawn band must survive through the inevitable tendency of the mind to favour its most recent experiences - who after listening to 15 bands does not find the performance of band number 1 fading in the memory, vivid though it was at the time?

    4BR conducted some analysis on this a number of years ago, which I can't currently lay my hands on online (anyone have the link?). As I recall they used British Open data (possibly combined with Nationals data?), and came to the conclusion that there is indeed an effect - if I recall correctly, the best positions to play were 2 or 3 before the end (not 100% sure on this?), while the worst was 2nd, not 1st (but this I do recall clearly). 3rd I think may also have been worse then 1st. Which suggests that it's a combination of factors - perhaps in addition to the reason I suggest above, judges are more alert for the first band than for the ones that follow, giving the first one a little boost above the where one would expect it to finish?

    Of course, as already noted, if a band puts down an unarguable marker anywhere, it will win... Provided the judge notices it!
  8. DS2014

    DS2014 Active Member

    Check out the link below on the 4BR website; it was the live updates feed for the Open day. Towards the bottom of that page, there is an interview with Steven Mead. In it, he says that being drawn first or last ought not make a difference. The judges, he says, have the test piece for some time before the competition, and they should use that time to go through the piece and create a grading criteria sheet for all the key parts of the piece. Each band should then be given marks against each of the criteria as they play. So, for example: a flugelhorn solo at, say, letter B would be one criterion, with marks awarded out of 10 for solo playing and ensemble accompaniment; by the time the band hits letter C, the adjudicator should be able to jot down a mark and then focus on the next bit, and so on. This should allow an adjudication right when the band plays, and mean there is no need to rely on memory. I guess it's still not an objective science, but it sounds better than the idea that some suggest that adjudicators rank the bands only once they've heard them all. The data analysed above in the first posting is still a rather small data-set from which to claim statistical significance. Additionally, adjudication styles will have changed greatly. Finally, we would have to make sure that we are not comparing apples with oranges. Were there the same number of bands competing in every competition? Which rankings closely align with statistical probabilities, and so on? Finally, which bands were drawn where in each of the years? I mean, if Dyke, Brighouse and Grimethorpe were consistently drawn first, second and third, then I bet there would be a much higher frequency of those draws resulting in victories rather than the bands drawn later. Each year, there are only three or four bands with a realistic chance of winning, so, if they are all drawn in a cluster, then the likelihood is that the winner will come from that cluster.

  9. midlandman

    midlandman Member

    It's a medical fact that you can concentrate for around 20mins before your mind wonders and there after the concentration time falls a little each time. So, for example, on Saturday band number 1 would get full concentration, however by number 17, even taking into account the short breaks, there is little chance of a full concentration throughout the piece. This is not derogatory to the top quality adjudicators but merely a mental impossibility. So there may be an advantage in a late draw when there is a large number of bands.
    However I do agree that there is still room for an outstanding performance at any time. But the key word is outstanding.
  10. DaveBBb

    DaveBBb Member

    I have completed a second statistical analysis and although the results are not as clear cut as the first they support the hypothesis that there may be an early draw disadvantage. This time I looked at the results of the National Final Championships held in the RAH from 1945 to 2013, source Brass Band Results. They show that there is a 14% chance of being in the top 4 if you are drawn 1 - 5, a 31% chance if drawn between 6 - 11, 33% if drawn 12 - 17 and a 22% chance if drawn 18 or above. Being drawn somewhere in the middle seems to offer the best chance of coming on the podium. Interestingly no band from 1945 to 2013 has come first from the band drawn number one. The best result, in terms of this statistical analysis was Besses who came second off spot one in 1978. At this time I offer no theories or psychological explanations for these observations. More research is needed before any firm conclusions can be reached.
  11. iancwilx

    iancwilx Well-Known Member

    Slightly off topic, Yorkshire Imps won the British Open on Pageantry in 1970 off a no. 13 Draw ~ Lucky for some eh ?

    ~ Mr Wilx
  12. MoominDave

    MoominDave Well-Known Member

  13. iancwilx

    iancwilx Well-Known Member

    Yes I've just spotted that Dave. No need for Triskaidekaphobia in banding eh ?

    ~ Mr Wilx
  14. JR

    JR Member

    not quite...

    What about Desford? - they won in 1991 on Energy off number 3

  15. DS2014

    DS2014 Active Member

    I still think that you would need to factor-in form to your analysis in order to derive a more robust conclusion. For example, for each year, if you compiled a form-weighting derived from how each band did in their respective areas, the nationals, and the open, then you would know which bands were in form for that year, or, perhaps more reliably, over a period of 3 years previous to the year under analysis. This would allow you, and I am guessing here, to see which bands had a realistic prospect of winning the competition in question. So, if, for example, Fodens had won their areas and won the Open (or done really well in both), but then had come way down the field in the Nationals having been given an early draw, it would lend more weight to the hypothesis that an early draw is a disadvantage.
  16. DaveBBb

    DaveBBb Member

    The Besses result was was from the British Open and they came first from draw number one. Where as the Desford result you refer to was when they came first from the number three spot at the Nationals. All the results from both contests - the Open and the National Championships, are accounted for in the ranges 1-5, 6-11 and so on and are expressed in terms of percentages. As I said this is a statistical analysis without any reference to the band, the contest, the adjudicator or the test piece. It is to explore the notion that there may be an early draw disadvantage in contests.
  17. DS2014

    DS2014 Active Member

    I appreciate all of this, but your analysis is predicated on the assumption that each of the 17 (or so) bands has an equal chance to win before the draw is made (given that you discount from your calculations the variable of the identity/form of the bands in question). My observation, and I know I am not explaining it very clearly, is that these bands are not like lottery balls in a drum. There are very good ones, good ones, and ones that are just about OK (relative to the others). Therefore, to draw robust conclusions, you would need to take into account the identity/form of the bands. To extend the lottery metaphor: if all of the bands are balls in a drum, then the best bands would be represented by more balls in the drum than the less good bands, giving them a more likely chance of winning.

    Another thing to consider is whether there is a statistical variance in the amount of times the best bands get an early draw compared to the weaker bands.

    I know, this is a load of balls :)
  18. MoominDave

    MoominDave Well-Known Member

    If you've got enough of a sample, the random draw will make all this come out in the wash anyhow
  19. DaveBBb

    DaveBBb Member

    Ermmm....... There could be a PhD is this......or at least an MSc ..... Any statisticians out there? Lol
  20. DocFox

    DocFox Retired

    Well, when you start combining statistics and psychology, it really hits my spots. My doctorate was in Ed Psych and Research (a change of one class and I could have earned a Ph.D. in Applied Statistics). Some very interesting analysis and observations have been made. It leads me to a question, how reliable are the Top 100 ranked list (usually posted on 4BR and/or WOB)?

    Are the rankings BEFORE the competitions, or a result of the competitions? IF, and I mean IF, the rankings were deemed reliable BEFORE a competition (which is often true in sports) you could have a great study. Stick with football for the sake of discussion. I am a Yank and therefore not very knowledgeable about club football, which keeps me from being emotionally attached. According to recent rankings Manchester United is a better club than Burnley. This could be tested statistically and would be kind of fun to do. It doesn't mean Burnley could not beat Manchester United, it would just be unlikely.

    If the Brass Band Rankings were reliable (rankings most of the time are subjective - but can still be reliable) then some very interesting statistical analysis could be done in regards to place in the Finals and the place in the draw. I would have a ball doing such a study.

    From what I have observed from the other side of the pond is that the draw is a factor. Adjudicators are human. I have been asked many times to be on a panel of judges for state band or orchestra contests at the High School level (the last 3 grades in school before graduating and going to a college or university for higher education in the US) for the state championship.

    That first band sets a standard. It is human nature to make the first band the leader until they are beaten. In giving scores, an adjudicator would have to leave room for other bands to do better (at least in scoring). I believe if you looked into it, if a band won from an early draw the scores, scores across the board would be lower. It is human nature.

    Competing in the arts is always dependent on objective scoring. I believe the draw does have a small effect on the scores. I also think that sometimes the judges blow it, but that is rare. Most often, within a degree of human error, adjudicators do a good job. Therefore, I believe that the best performing bands often score the highest.


    Adjudicators are human, but in my humble opinion, they for the most part, do a terrific job.