[CQ-Contest] Log Checking Software

[Ward Silver]

> >Are there more important things?  You bet there are. Catching cheaters,
> >reducing reporting turnaround time, making it easier for people to submit
> >logs, extending log checking to more contests, etc. etc. etc.  Not
worrying
> >about sub-1% minutia that can't even be defined except on a QSO-by-QSO
> >basis.
>
> Ward, I was reading the ARRL DX SSB results in QST this month, and I
> believe the USA/VE Multi-single category was won with about a 0.4% margin
> of victory. I can imagine the station that won (or came in 2nd) vehemently
> arguing for that 1%.

That argument confuses accuracy for precision. There will always be races
that are determined by margins smaller than the margin of error of the log
checking process.  Always.  Why?  The smallest possible margin between
scores is equal to the QSO points for one QSO.  Since the smallest change
that log checking can make is to remove one QSO, to be better than that the
log checking error must be less than one QSO - that is, perfect.  Attempting
to reduce the log checking error to a level so small that it can not affect
the order-of-finish in any race is quixotic and impossible.

Analogy time - professional athletes do not expect umpires and referees to
make every call correctly.  They expect ENOUGH calls to be correct and
consistent that over the course of a season, the margin of error is small
enough that everyone has confidence that team performance determines the
final standings.  Yes, individual games will be decided by individual blown
calls, but at the end of the season, the right teams finish first and last.
(If anyone wants to discuss John Denkinger's blown call at first base in the
1985 World Series with St. Louis vs. Kansas City, you'll find a sympathetic
ear here...but I digress.)

The contest community has already endorsed the decision that unique calls
should not be removed from logs if they can't be positively identified as
bad calls.  This lets MANY bad calls into the logs.  It is inconsistent, at
best, to then focus on these semi-random errors deep in the noise when other
far more significant and addressable sources of error are present.

In a world of limited resources, one has to make decisions about where to
expend them.  Where would we rather see effort expended - going through LCR
reports with tweezers or finding packet cheaters and self-spotters?  Would
we be willing to wait another two or three months for the results to be
polished?  Are there enough volunteers to do it?

AD1C does us all a great service by maintaining the CTY file as best he can,
gratis.  Does Jim screw up once in a while?  Yes.  Do people feed him bogus
information from time to time?  Yes.  Is there conflicting information out
there in the various CTY editions - i.e. K2G?  Yes.  On the whole, are the
CTY files of sufficient quality to use as a reference?  Yes.  Even with the
imperfections, yes, and we are grateful for them.

Perhaps if the amount of work that goes on behind the scenes on our behalf
was more visible, there would be a greater appreciation of what we seem to
regularly take for granted.  On reflection, it's likely that most of us do
appreciate the work involved and the quality of the results.  Unfortunately,
the Internet - great leveller that it is - gives equal time to unreasonable
and reasonable viewpoints alike.

In my own Busted Q Department, it has been noted that log checkers and
contest administrators do receive a nominal payment for doing the checking.
It's hardly market rate, but it is not Foxtrot Radio Echo Echo, so I must
retract that particular statement.  Nobody is getting paid nearly enough,
however, for delivering quality and being rewarded with unrealistic
expectations in return.

73, Ward N0AX