[CQ-Contest] Busted calls and Uniques

[Tree]

VR2002BG writes:

> However, expect them to put some extra effort to turn other Us into Bs
> as a result of your effort to convince them that some Bs were Us - 

and AA4GA comments:

> That doesn't seem like a fair application of the process.  Is this 
> what CQ says, or is it supposition based on human nature, i.e. "Oh 
> yeah, I'll show him"?

Well - which is more fair?

1. A non-automated process applied unevenly to a few logs. (This is the
old method of log checking before about 12 years ago).

2. An automated process that is evenly applied to all logs without any 
special consideration given to any log.

3. An automated process applied to all logs, and a few of them get special
consideration where each possible questionable judgement is challenged
without any possibility of an equal amount of attention given to possible
errors that could result in the score being reduced?

Most log checking processes I know of will default to give credit unless some 
criteria for busting the QSO is met.  The criteria should be sufficient to have
a high degree of confidence that a call is wrong.  In some cases, the wrong
decision might be made (when some number of improbable events line up). 

However, it should be even more likely that there will be callsigns that are
really bad - but the checking process was unable to determine what the bad
callsign was...  and therefore it was counted as a good QSO with a unique 
callsign.  Some manual work on the part of the log checker can probably 
figure out what the busted call is - but they don't have time to do this
for all logs - so it gets done for none (except maybe for a pair of logs
that are very close in the results).

For example, in the ARRL DX contest - the log checking really stinks.  I 
would estimate that only about 25 percent of errors made are detected by
the program.  Also, maybe one out of 100 of the bad calls is really good...
so the process isn't perfect.  However, I would argue that it is effective.

Trying to improve the error detection in the ARRL contest will result
in a significant increase in the false bust rate - which I think would
be too high to defend.  

I think that applying this process to all logs - without special consideration
is the most fair way to use it.  When someone comes up with a case where
the wrong decision was made...  I will look to see if a process change can
be made to address that specific case - which can then be used to re-run all
of the logs.

I think the important point to take away from this is that the first objective
of the log checking process is to make sure the order of finish is correct.
In other words, that the right person is winning the contest.  I think having
a totally defendable 100 percent correct log checking report is not the goal.

In the NBA - no ref gets all of the calls right.  However, they tend to get 
most of them right and the errors tend to average out (and this coming from a 
Portland Trailblazer fan no less).

It wouldn't be fair to let one of the teams challenge each call that goes
against them without the other team having the same opportunity.

As far as AA4GA's comment about log checkers not being willing to work
with him - I would sure like to know which contest that was for??  CQ WW?
What specifically was the problem?  I don't believe this is the normal
experience for people who approach them.

73 Tree N6TR
tree at kkn.net