Dangerous judging reforms are advocated!
Neil
judging "reforms" in the turbulent wake of the Canada-Russia
skating scandal. Some of the ideas were helpful and warranted:
disassociate judging from national organizations. Some were
perhaps OK, but would detract from some of the classic
associations: change from 0 - 6.0 system to grading on various
moves, etc. One concept was a very bad idea, however
well-intentioned: Use 14 judges, and discard 7 scores at random: ~
"to make it harder to fix the outcome." Well, that sounds good at
first, but there is a huge risk of throwing the result away from
what the actual average consensus was, due to statistical
fluctuation. Imagine a case where (using old rating system for
illustration) skater A really got 8 x 5.8 and 6 x 5.7, and skater
B really got 7 x 5.8 and 7 x 5.7, for a clearly lower score.
However, if the random throw took out five of skater A's 5.8s and
two of her 5.7s, she ends up with 3 x 5.8 + 4 x 5.7. If the throw
takes three of B's 5.8s and four of her 5.7s, she ends up with 4 x
5.8 and 3 x 5.7, to win over B.
I'm just chopping this out, but if you want to do the math, I
think you'll see that the chances of something like this happening
are quite high (< 50%, but too much risk, around the order of 0.1
or so.) To replace a small chance of some judges cheating with a
large chance of a good skater being edged out at random by a
poorer one is not good at all. The better solution is temporary
secret ballot, as I have argued elsewhere.
Neil Bates
Josh Drury
"Neil" <para...@mailcity.com> wrote in message
news:u73ftbj...@corp.supernews.com...
> One concept was a very bad idea, however
> well-intentioned: Use 14 judges, and discard 7 scores at random: ~
> "to make it harder to fix the outcome." Well, that sounds good at
> first, but there is a huge risk of throwing the result away from
> what the actual average consensus was, due to statistical
> fluctuation.
<mathematical example snipped>
>
> I'm just chopping this out, but if you want to do the math, I
> think you'll see that the chances of something like this happening
> are quite high (< 50%, but too much risk, around the order of 0.1
> or so.) To replace a small chance of some judges cheating with a
> large chance of a good skater being edged out at random by a
> poorer one is not good at all.
I don't see how this is worse than the current system, other than the sample
is now smaller than before (7 vs. 9). As it currently is, it's still the
luck of the draw as to who's judging, instead of having marks thrown out we
just don't get any more marks. Besides, if the judging system is truly
objective all judges should have similar marks, and poor skaters won't
flukily edge out better ones. I'll concede, that's a big if.
Josh Drury
William J. Pitcher
news:qVjc8.63775$A44.3...@news2.calgary.shaw.ca...
Some judges though mark high for everybody or low for everybody. This was
always said to be okay as long as they were consistent. If you got the
low scoring judge and the next skater didn't -- that wouldn't be fair. Has
anybody heard -- will the seven random judges stay the same for the entire
night or change after each skate?
Cheers, Bill.
Olsen
sample
> is now smaller than before (7 vs. 9).
Exactly. You could say the current 9 judge system is wrong a lot of the time
compared to what 18 judges would decide. They just need to make sure the
unused scores are never disclosed.
jaymeister
and snowboarding halfpipe both judge thneir events. In each case you have
specific judges marking specific aspects of the performance. And, as far as
I know, there has never been a real dispute with the results of these sports
in Olympic competition. The advantages to scoring this way are:
1. Each judge can can score each athlete on an absolute scale and not have
to worry about placement or "leaving room" for subsequent athletes. They can
let the computer take care of the final results.
2. It is more possible for the best athlete on that given day to win the
competition. How often in figure skating does the 21st ranked participant
win like Ms. Camplin did today in aerials? I realize that this is an apples
and oranges analogy, but clearly the lower ranked skaters often get screwed,
even when they skate exceptionally well for them, based on reputation and
the need to leave room at the top, so that a 21st ranked skater might end up
in 18th place instead of the10th that they would merit. (Or an 8th ranked
skater will end up 6th instead of 3rd, etc.)
3. Accountability. Since there are only a very few judges scoring each
element, a biased score sticks out like a sore thumb. You just don't see a
great disparity in the numbers among judges in these other sports.
Above all, the judges have to be separated from the national skating
federations. Figure skating purists have said for years that the current
system is best because it minimizes biases by requiring a consensus of the
judging panel, eliminating the impact of the renegade judges. If this is the
case, why does figure skating have, by far, the most judging disputes of any
judged sport? (Excluding boxing of course.) Politics definitely plays a
part, and the sport needs to be de-politicized.
Raptor
--
Lynn Wallace
http://www.xmission.com/~lawall
Windows ME sucked so bad, I ditched Micro$oft for Linux.
Ruthann Biel
<wil...@pitchergroup.com> wrote:
>Some judges though mark high for everybody or low for everybody. This was
>always said to be okay as long as they were consistent. If you got the
>low scoring judge and the next skater didn't -- that wouldn't be fair. Has
>anybody heard -- will the seven random judges stay the same for the entire
>night or change after each skate?
That was a concern of mine also. It would seem very UNfair if the
judges changed during the event. One skater might get a judge that
scored high, and another a judge that scored low....and their two
performances might be very close despite what the scores might
show.
--
Ruthann Biel | Mother, Unschooler, Stitcher, Violinist.
r...@woozle.Emp.Unify.Com |----------------------------------------
+1 916 381 4205 | Sacramento, California USA
Ruthann Biel
the spins, footwork, and in-between-stuff, and 3 judges score the
overall impression. and remove the judges from the federations and
put them under an international umbrella.
Fiona McQuarrie
judges themselves and/or their federations are corrupt, someone is going
to find a way to put the skaters they want first into first. It's not the
judging system that is the problem, it's the judges.
Fiona
Bill Taylor
|> You can tinker with the marking system all you want, but as long as the
|> judges themselves and/or their federations are corrupt, someone is going
|> to find a way to put the skaters they want first into first.
Yep; judges with axes to grind or causes to support are going to make
a mockery of most subjectively-judged events, which I tend to agree have
no place in Olympic competitions.
|> It's not the judging system that is the problem, it's the judges.
Well said indeed, Fiona!
------------------------------------------------------------------------------
Bill Taylor W.Ta...@math.canterbury.ac.nz
------------------------------------------------------------------------------
We need drug testing for figure-skating judges!
------------------------------------------------------------------------------
galaxicon2000
judges who assigned the highest and lowest scores are
revealed, but no other judges' identities are revealed?
Then you would count all the judges' scores except the
highest and lowest.
Also I must say that the social problems invloved with
generating a number that everyone agrees is random could
be daunting. We're not dealilng with cryptographers here,
you know.
"Neil" <para...@mailcity.com> wrote in message news:<u73ftbj...@corp.supernews.com>...
Anonymous
On 27 Feb 2002 08:40:29 -0800, galaxi...@yahoo.com (galaxicon2000)
wrote: