Thanks Matt for excellent idea.
Yes, we analyzed several ALL.TXT files as well as LOGs.
Beside missing RR73, common problems is also "Two QSOs in parallel":
I call stn A, then B. While QSO with B is in progress, A comes back and I try
to complete QSO with A.
Asking people to upload ALL.TXT instead of cabrillo LOG was discussed as well,
maybe even real time. It is not yet time.
Internal LOG is needed to keep track of duplicates and new multipliers.
This might be problem with X-QSO record after RR73 is send.
Good practice is to keep repeating RR73 till some other activity from called
station is heard,
like CQ or calling somebody else.
Best 73, mni DX
Iztok, S52D
Date: Wed, 26 Feb 2020 16:29:13 -0500
From: Matt Power <mhpo...@mit.edu>
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] FT4 and FT8 Contesting
Message-ID: <202002262129.01qltdcq004...@outgoing.mit.edu>
Log a QSO when you send RR73 if you are reasonably confident it will be copied.
I'm also active in FT4/FT8 contesting (https://ww-digi.com/scores.htm
etc.), and have worked on design of a different (not yet active)
FT4/FT8 contest for a different major sponsor. I feel that this
guideline should be revised slightly. I'd like contest rules to add:
"If you send RR73 or RRR to a station, and decode no further
transmissions directed at you from that station, then a submitted log
must include that on either a QSO or an X-QSO line. If adjudication
concludes that you sent RR73 or RRR, but then decided to omit that QSO
from your log, you will be disqualified for unsportsmanlike conduct."
The intent of this rule is not to minimize the overall NIL rate, but
to make logging entitlement more predictable. If Station B receives an
RR73 or RRR from Station A, then Station B should always be entitled
to log a QSO and be credited with its QSO points. This entitlement
should not depend on Operator A's opinion of whether reception was
likely. As far as I know, this has always been the case for similar
QSO protocols, e,g., documents as old as
https://physics.princeton.edu/pulsar/K1JT/doc/wsjt/ state "If you have
received RRR???-???that is, a definite acknowledgment of all of your
information???- the QSO is complete." If Station A were allowed to log
selectively (after sending RR73 or RRR), competition becomes less
fair. A strong station can, in effect, push NILs into the results of
many weaker stations by declining to log them unless the weak signal
continues to be decodable for a 73. The strong station may be
motivated to do this if the contest has a substantial NIL penalty, or
if the strong station wishes to have a Golden Log.
There has been an analogous CW situation. A well-known 160m CW
contester once made an announcement that, after faintly copying a
report from a caller, he would send "TU EE" instead of the usual
TU+callsign. Unless the faint caller sent EE back, that caller would
not be logged. My opinion is that the faint caller achieved logging
entitlement upon receiving the TU.
FT4/FT8 contesting is different enough from CW/SSB/RTTY contesting
that having no NIL penalty is a correct design over the long term:
1. Adjudication is inherently easier. Thousands of stations around the
world are storing every decoded FT4/FT8 message into an ALL.TXT file.
Contest sponsors can accept voluntary ALL.TXT submissions from anyone,
and use this crowdsourcing to automatically detect many types of fake
QSO claims. By contrast, CW/SSB/RTTY adjudication is -- to some
extent -- limited to matching one log against one (forward or reverse)
log, or using humans for manual review of wideband decoding.
2. Survival of weak signals across multiple messages is, in some ways,
inherently less likely for FT4/FT8. One reason is that FT4/FT8 have
all-or-none decoding. Another reason is that new intense QRM often
arises in the middle of an FT4/FT8 QSO, but less often in the middle
of a CW/SSB/RTTY QSO (both because FT4/FT8 transmissions are sometimes
longer, and because FT4/FT8 operators sometimes have great contention
for a small set of usable transmit frequencies). To put this another
way: the engineering of the FT4/FT8 decoders is superb enough that a
weak station will intermittently be decoded with near 100% certainty,
even if that station usually is not decodable.
3. Omitting the NIL penalty, and/or accepting that the NIL rate may be
five percent forever, does not eliminate incentives for strategy
improvements. Although (in my proposal) an operator must log a QSO
after sending RR73 once, the operator will sometimes find it best to
send RR73 multiple times. This can be based on understanding
propagation conditions, knowing one's own transmitting effectiveness,
or even past experiences with the receiving effectiveness of specific
QSO partners.
4. There's nothing inherently wrong with a contest in which entrants
often lack "full confidence that a QSO is complete and will be logged
at both ends." Yes, it might be unsettling to know that you have a
minuscule chance of a Golden Log even if you try very hard. However,
it does add another strategy element. A successful entrant may need to
keep track of QSOs where RR73 reception seemed iffy, and try to rework
those stations later (stations are "in between" a dupe and a non-dupe).
Matt, KA1R
[http://psn.sdn.si/ts/TS_Banner_NEO_oktober_2019_250x170.jpg]
<https://neo.io/info?utm_source=mail_podpis&utm_medium=mail_podpis&utm_campaign=mail_podpis>
Pravni pogoji / Legal disclaimer
Telekom Slovenije, d.d., Ljubljana <http://www.telekom.si/disclaimer>
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
