CQ-Contest
[Top] [All Lists]

[CQ-Contest] ARRL 10M Contest - The REAL Story {necessarily long}

Subject: [CQ-Contest] ARRL 10M Contest - The REAL Story {necessarily long}
From: k8cc@ix.netcom.com (David A. Pruett)
Date: Wed Aug 11 21:55:00 1999

I've had computer problems for the past three days, however friends have
been calling me asking about the brewing brouhaha about the 10M Contest.  I
drag out the laptop to find 200 messages waiting from three days.  Egads! 
For those who recently (in the past three days or so) have e-mailed for
their entrant report, please be patient while I get my computer situation
straightened out here.  At the moment, e-mail is on one computer and the
logchecking results on another.  Until I can get the e-mail on the desktop
straightened out, it will be a slower process transferring the reports
between computers with floppy disks.  Yeah, I know I need a network...

I've read the comments about the 10M Contest log checking, some of which
raise good points, but some of which were just plain ignorant.  Here is the
true story.

Nobody, repeat NOBODY, got a NIL because a log for the claimed station was
not received.  Let me explain how the log checking worked, at least for the
10M Contest.

There are two passes in the log checking process.  In the first pass, each
QSO in every entrant's log for which we have a log for the "other" station
is checked.  If a QSO with the entrant's call does not appear in the log,
it is marked as NIL, but that's not the end of the process.  Once all the
QSOs which do match up are checked, a second pass goes thru all of the
NILs, looking for a callsign in the "other" station's log which could have
been mistaken for the entrant.  If a similar call is found, the NIL is
cleared and the entrant gets credit for the QSO.

To better explain this, let's use ZL1ANJ's example.  While checking
ZL1ANJ's log, we find a QSO with K1TTT.  Looking thru K1TTT's log, we don't
find ZL1ANJ so the QSO is marked NIL in ZL1ANJ's log.  On the second pass,
we go looking again in K1TTT's log to find a callsign similar to ZL1ANJ
which has not already been confirmed with someone else.  By this, if we
found ZL1GQ for example, but that QSO had already been confirmed, it would
not be considered as a possible candidate for the ZL1ANJ QSO.  Since we did
not find another callsign in K1TTT's log similar to ZL1ANJ that had not
been confirmed elsewhere, the NIL was allowed to stand.

The fundamental question here is how do we define a callsign which is
"similar" to another?  Some years ago, the term used was a "plus 1" check,
which I interpreted to mean that two callsigns were within one digit of
being identical.  I recently posed this question to Dick Norton, N6AA from
the CQWW team and he said that the definition had evolved to mean "two
callsigns which could be confused for one another".  I agree with this
definition, but how do you describe this in real terms.

The ARRL 10M logs were checked using "plus 1" rules to judge a similar
callsign.  With this method, if K1TTT had logged Martin as ZL2ANJ, or
ZL1AN* or even ZL1AN, if would have been judged "similar" and he would have
gotten credit for the QSO.

Improving the "similar" or "confused" call algorithm is one of my goals for
next time.  For example, what if two letters get switched - ZL1AJN for
ZL1ANJ - this seems like a common error that could be reasonable to catch,
and not result in a NIL.

There is another side to the similar call question which we did not check
for in this years results, and that is when the entrant busts a call.  For
example, suppose ZL1ANJ logged K1TTT as K2TTT.  If we have no log for
K2TTT, Martin get's credit even though in truth, its a busted QSO.  If we
had done a "similar call" check on K2TTT, K1TTT would have popped up and
the computer would have searched the K1TTT and found the QSO, which would
have been scored as a "busted" call.  We plan to add this capability next
year.

In closing on this particular topic, we aren't out to "get" anybody.  Since
the logs are all scored by computer, everybody is treated fairly. 
Everybody gets the benefit of the doubt, unless hard evidence indicates an
error.

With regards to the K2BA QSOs in ZL1ANJ's log, I believe the removal is
legitimate.  NIL means "NOT IN LOG".  The QSO was not in K2BA's log. 
Speaking for the 10M checking process, I would be glad to find a way to
accomodate "non-scored" QSOs.  Its my understanding that in CQWW, when
someone wants to submit an all-band log for club points, but only compete
on a single band, for example, they still must submit all QSOs with a note
on the summary sheet to the effect of their single band status.  This
allows the CQWW checkers the ability to check all the QSOs, but categorize
the score as the entrant desires.  N6AA or N6TR, correct me if I'm wrong.

We saw a number of strange QSOs where the QSO was found in both logs, but
one station claimed it was on CW and the other on SSB!  Who do you believe?
 Some of this appeared to be related to SO2R.  After much agonizing over
this, we decided to resolve the QSO in the following manner.  If the
received exchange at one end (RST + state/province/QSOnr) exactly matched
the send exchange from the other end (RST + state/province/QSOnr) then the
mode was assumed to be the mode indicated by the RST.  To put it another
way, either the RST and state/province/QSOnr must match, or the mode and
the state/province/QSOnr must match to get credit for the QSO.  If a QSO
fails both of tests, then it is marked as "logged incorrectly" and credit
is removed.  However, this is not a NIL, and no penalties were assessed.

I hope this clarifies how the ARRL 10M logs were checked.

73,

Dave Pruett, K8CC


--
CQ-Contest on WWW:        http://www.contesting.com/_cq-contest/
Administrative requests:  cq-contest-REQUEST@contesting.com


<Prev in Thread] Current Thread [Next in Thread>