KQ3M writes:
> K3ZO posted comments and queries on how the ARRL log-checks certain
> types of "unstable" exchanges. I am not sure that they were ever
> addressed or responded to and I would like a clear and definite
> explanation of how certain situations are and will be handled by the
> log-checkers.
>
> I too have been very frustrated by exchanges that change in the middle
> of a contest and I believe that all of us have lost many points because
> we have CORRECTLY copied the "new" exchange.
What data do you base that statement on? I don't believe this is
true. Do you have checking reports from your own efforts to back
this up?
> 1) How do the log-checkers know that PA0XYZ used 100 watts on 10 + 15
> and 150 watts on 80?
This would be very easy to figure out - because PA0XYZ would show up
in all (or most) of the logs with 150 watts when worked on 80 meters.
> Or that LYxx used 200 watts on 15 + 20 but KW on 40?
Again, the data would be there. You can even collect this information
without having the log from the station.
> In the computer software that most of us use, there is no way to "note"
> that a particular station changed his power for contacts #361-#514. So
> if they don't note their power change in the log (or forget to) how will
> the log-checkers know? And why should we be penalized for it?
You shouldn't be. It is up to the log checking software to detect
these cases and "do the right thing".
> Let's use another example....
>
> 2) I work F5XYZ in the contest who gives me #40. I have copied
> correctly #40 and should be rewarded for it. Is it my responsibiliity
> to assume that the number DOES NOT represent power and inquire further
> as to how many watts the guy is running or assume that hw is giving me a
> qso # instead of power? What if I don't speak French and he doesn't
> understand English? Do the log-checkers penalize me for copying an
> "incorrect" exchange? I have lost lots of points and qso's for this in
> the past. I feel strongly that this is UNFAIR and penalizes me for
> accuracy rather then rewarding it!
Again - it really doesn't matter what the number means. The log
checking software should detect do one of two things based upon
the received data:
1. Determine that F5XYZ appears in > 80 percent of the logs with #40
and thusly conclude that is what was being sent consistently.
or
2. Determine that F5XYZ was sending some random information that can't
be used for log checking purposes (i.e., a QSO number).
> 3) I work DL8XYZ. We have a good contact. Later in the contest he
> calls me on the same band. I tell him it's a dupe and he says no it is
> not. I work him again for an "insurance" contact and in the log mark it
> a dupe for 0 points. If in fact he copied my call wrong the first time
> (I would not know this since he never sent my call!) do I get penalized
> for the first qso since I am not in his log? What if I logged the first
> qso but did not log the second qso? A dupesheet crosscheck would show
> us in each others logs but at different times. I would have logged the
> first qso and he would have logged the second. Do we both get credit
> for the qso's or both lose credit since we show the qso at different
> times? Or is the only way to get credit by showing the first qso and the
> "dupe" qso?
This is again something that the log checking software should be smart
enough to detect and "do the right thing". I believe I have detailed
how the new SS log checking software handles this type of situation.
This is also a feature of the CALLTEST program that is used for the
CQ WW and more recently in the IARU contest.
> 4) I work G3XYZ and he gives me #3. Later I work him on a different
> band and he gives me 100. I copied correctly BOTH times but clearly the
> second time he is giving his power (or he changed from QRP to 100
> watts).
> Are the log-checkers going to penalize me for getting a number instead
> of power on the first qso? If so, then do I have to go back into my log
> and "correct" the #3 to a 100 to get credit? What happens if I do that
> but then the log-checkers see that he sent #3 in the first qso and
> penalize me for receiving 100? How will the log-checkers know when the
> guy stopped sending serial numbers and started sending power?
You should never correct this type of information. Once a pattern has
been established (in the logs we have on hand) - it could be used
against you.
For example - if we can determine that he sent 3 (as in 3 watts) for his
first 57 QSOs, and then switched on his amplifer after QSO #57, a really
smart log checking program will sync up to this and bust the QSO where
you changed it to 100. This would be "doing the correct thing" in my
book.
> I for one would like to know EXACTLY how the log-checkers are handling
> these situations. I want full credit for the qso's I make correctly and
> I DON'T want to lose credit for qso's where I copy correctly but the
> log-checking practices might be unrealistic and unfair.
I have outlined how "it should be done". Not all log checking processes
are up to the above task. In general, they should either improve
themselves - or work in a limited way so as to not remove information
that can't be confirmmed. As I write the programs I use - I use the
"IUPG" principle - Innocent until proven guilty.
> In the 1500 qso's I made there were more than 20!! exchange changes.
> This is a VERY COMMON problem!! In about 10 qso's there was a language
> barrier and the station either couldn't understand the question or
> didn't want to answer. Of the remaining 11, I had to use my "broken"
> high-school Spanish FOUR times to get an answer (if I spoke French and
> German too I would have had more answers Hi Hi).
Again - I think you should just put down what they say - and not worry
about it.
> 1 station went qro because he fixed his amp (changed from 100 to 500)
> 1 station went from 100 to 150 becuase on 10 his swr was too high and on
> 15 it was lower!
>
> 2 stations went from 300 or more watts to 100 because of TVI
> 4 stations sent serial numbers the first time and power the next time
> 2 stations were using different power on different bands the WHOLE
> contest! (ie. 500 watts on 20 and 100 watts on 10 ALL CONTEST!)
This data isn't surprising. I saw many similar things happen in the SS
where people would run at one power level for awhile and then switch
to another.
> I can't even tell you the number of times that I stopped my run and took
> the time to explain to the station I was working that the correct
> exchange was the "power" not the serial number or CQ Zone. I am not
> even counting these in my 11 exchange changes! Since other stations had
> changed from serial number to power the next time I worked them, clearly
> I was not the only station "educating" others on the correct exchange in
> the contest.
Good.
> So I would appreciate clear asnwers to all my questions and an
> explanation as to how the log-checkers handle these very common
> situations. This will go a long way to "de-mystifying" the log-checking
> process and improve the quality of the checking as well as our
> understanding of the fairness and/or weaknesses of the process.
I hope my answers have helped. There is still a lot of work remaining
to be done. I have put a lot of my time into improving the process
for the SS contest and am sorry to say that K1EA's suggestion hasn't
caught on enough for me to quit my day job and do the same kind of
thing for the ARRL contest.
> 100% accuracy is the goal of any consummate professional or contester
> but in the absence of clear answers and guidance it is unattainable and
> creates otherwise unnecessary frustration and perceptions of unfairness.
I think this effect is amplified by ignoring the positive steps that have
taken place during the past years and focusing on a small percentage of
the exceptions. Any log checking process is going to have some holes
in it - and we are doing our best to make them as small as possible.
But if the errors created by the small holes get blown up out of
proportion to the overall results, then people will continue to
be frustrated.
73 Tree N6TR
tree@contesting.com
--
CQ-Contest on WWW: http://www.contesting.com/_cq-contest/
Administrative requests: cq-contest-REQUEST@contesting.com
|