Pete N4ZR wondered how the error rate is defined. I don't have the
official definition, but I feel fairly sure that it is NOT the
percentage reduction in score.
>From my report and one or two others, I believe the error rate is the
percentage of non-dupe QSOs that were busted (in any of the three
categories -- busted callsign; miscopied precedence, check, or
section; or miscopied QSO number). I think the report lists the dupes
at the start mainly for completeness; the dupes are already marked in
our computer-based logs and have been counted as 0 points in our
claimed scores.
The percentage reduction in score would include the penalty for busted
callsigns: three QSOs deducted from the total (after the busted QSOs
have been removed) for each QSO with a busted callsign. An analysis
of the reductions would be of interest, too.
I would like to see a discussion of the choice of three as the number
of additional QSOs to deduct for each busted callsign. (Is this
covered in the forthcoming NCJ article?) Perhaps it was borrowed from
the current practice in the CQWW DX Contest. Given the origin of the
SS exchange in the preamble of messages in traffic handling, three
doesn't strike me as an obvious choice. The parts of the preamble
aren't quite of equal importance, in the impact of garbling one of
them. Perhaps getting the precedence wrong would have the most impact
(say, by changing Priority to Routine or vice versa), followed by the
check. Garbling the callsign would make it harder to reply to or
service the message, but it wouldn't usually prevent delivery. What
do others think about this?
73,
Dave, K1HT dave_hoaglin@abtassoc.com
--
CQ-Contest on WWW: http://www.contesting.com/_cq-contest/
Administrative requests: cq-contest-REQUEST@contesting.com
|