Gap in AMA finals rules - resend

[Atwood, Mark]

Having run a LOT of scores through a spread sheet, and then "biasing" one set by 2 points per maneuver ( both up and down ), dropping highs and lows almost NEVER fixes the problem.  It also didn't solve the problem for just biasing high K-factor maneuvers.

I'm going to over simplify, but here's the general jist...

Take this scenario...   5 judges, and 5 pilots who all get virtually the same scores...  One low judge, 3 neutral judges and one easy judge.

Judge 1)  6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 

Judge 2)  7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7

Judge 3)  7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7

Judge 4)  7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7

Judge 5)  9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9

Let's say all 5 pilots get these identical scores.   5 way tie regardless of whether we drop highs or lows correct??   Now lets say Judge #2 is biased...and he breaks the trend and gives me (he's my buddy) all 10's   

Without dropping Highs and Lows...I win the contest.  
If we DO drop Highs and Lows...the other four flyers get straight 7's as judge #1 and Judge #5 are dropped.   But for me, my 10's are dropped, Judge # 2...but instead I keep Judge #5 with his 9's and I still win

If my buddy was more subtle and only gave me 8's...Judge 5 would be dropped...but I'd still win keeping my 8's.  

Bottom line...dropping the highs and lows doesn't help things when there is a disparity of scores.

FYI, the same effect occurs if you zero a few maneuvers...I will always lose, because I'm keeping a lower score, even though I'm not keeping the zero.

We HAVE a legitimate problem...but dropping Highs and Lows does NOT consistently...or even remotely improve the accuracy of the outcome until there are a large number of judges.

-Mark
-----Original Message-----
From: discussion-request at nsrca.org [mailto:discussion-request at nsrca.org] On Behalf Of Grow Pattern
Sent: Tuesday, October 25, 2005 5:27 PM
To: discussion at nsrca.org
Subject: Gap in AMA finals rules - resend

From: Grow Pattern
To: Rick Wallace ; rcpattern at comcast.net ; atwoodm at paragon-inc.com ; 
discussion at nsrca.org
Sent: Tuesday, October 25, 2005 11:34 AM
Subject: Creating rules where none exist


FYI- This is what FAI says about scoring. The AMA does not use TBL .(See 
note 2 on discards).

Most folks are writing about this from the perspective of "gut-feel".

It is true that one "bad" judge can spoil it for others, the logic of 
"discarding does" not address the issue. It only deals with one special case 
of scoring.

Eric.


Scores for all rounds, preliminary, semi-finals and finals, will be computed 
using the Tarasov-Bauer-
Long (TBL) statistical averaging scoring system. Only computer tabulation 
systems containing the
TBL algorithm and judge analysis programs and approved by the CIAM Bureau 
can be used at
World and Continental Championships. All scores for each round, preliminary, 
semi-final and finals,
will then be normalised as follows. When all competitors have flown in front 
of a particular group of
judges (i.e. a round) the highest score shall be awarded 1000 points. The 
remaining scores for that
group of judges are then normalised to a percentage of the 1000 points in 
the ratio of actual score
over winner's score.

PointsX =
SX ´1000

SW

PointsX = points awarded to competitor
X
SX = score of competitor
X
SW = score of winner of round.


Note 1: Final and semi-final flights to determine the individual winner are 
only required for World and
Continental Championships. For smaller contests the total of the three best 
preliminary flights may
be used to determine the individual winner and team placing.

Note 2: The TBL system can only be applied for events with at least 10 
competitors and 5 judges.
For those smaller events that are not scored with the TBL system, the high 
and low scores for each
manoeuvre will be discarded if four or more judges are used. 

=================================================
If you want your reply email to go to the list, you must Cc: the list!

To access the email archives for this list, go to
http://lists.f3a.us/pipermail/nsrca-discussion/
To be removed from this list, go to http://www.nsrca.org/discussionA.htm
and follow the instructions.

List members email returned for mailbox full will be removed from the list.

If you want your reply email to go to the list, you must Cc: the list!

To access the email archives for this list, go to
http://lists.f3a.us/pipermail/nsrca-discussion/
To be removed from this list, go to http://www.nsrca.org/discussionA.htm
and follow the instructions.

List members email returned for mailbox full will be removed from the list.