David,
What kind of post-contest analysis are you looking to do? I can think of a
number of valuable statistics, such as frequency of errors related to rate,
CQ vs S&P, first or second day, time of day, time without a break, etc. It
would be particularly interesting to correlate -N penalties with large
pileups, band conditions, antennas, etc.
It occurs to me, however, that the ability to import UBN data will
ultimately escalate the computer's role in determining correct calls
*during* a contest. Using UBN data from one or more past contests, it would
be relatively easy to compute a probability that a particular entry is
incorrect and suggest the most likely correct call. While not 100%
foolproof, at a minimum such a warning might prompt the operator to ask for
a fill. Of course, many lazy ops would simply choose the computer's
suggestion and gamble on its accuracy vs slowing the rate.
I have mixed feelings about this sort of thing. I most certainly use SCP,
because I know just about everyone else uses it, and we all know it's not
100% accurate -- you take your chances if you choose one of several similar
calls suggested by SCP without asking for a fill. But I've been concerned
about the long-term trend of relying more and more on the computer versus
the operator, and the possible negative impact this may be having on
developing operator skill over time.
For example, all the popular loggers now display the exchange data from
previously logged contacts. I use this feature, too, on the theory that I
copied the exchange the first time around and should be able to use what I
worked to obtain (and I can't figure out how to turn of this feature
anyway!) Like SCP, it's not a good idea to rely on previous exchange data,
especially in contests where power is used in the exchange. But there's been
escalation beyond this. Writelog has a feature that will display exchange
data from a previous contest log when a call is entered. This can be
particularly useful in contests with complex exchanges, like ARRL November
Sweepstakes. (Thankfully, it's not sophisticated enough to check multiple
logs and display the exchange with the highest probability of being
correct.) I, for one, don't use this feature because it seems to me that the
issue is how well I'm copying in this contest, not a different contest, and
I think the feature works against developing better copying skills. I have
similar feelings about using Writelog's built-in CW decoder.
Frankly, I have reservations about all the computer-aided copy features. I
realize they can make the contest more fun, and there's no way to prevent
people from using them, but I wonder to what extent are they working against
gradual improvement of operating skills. I'm curious what others think about
all this, and in particular dynamic use of past UBN data during a contest.
73, Dick WC1M
> -----Original Message-----
> From: David Robbins K1TTT [mailto:k1ttt@arrl.net]
> Sent: Sunday, May 22, 2005 10:34 AM
> To: reflector cq-contest
> Subject: [CQ-Contest] Logger writers everywhere UNITE!
>
>
> I have been looking at automatic importing of ubn data into
> the n1mm logger. I have come to one big realization here.
> All logging programs (for ARRL and CQ contests at least) have
> to produce Cabrillo log files for submission. But there is no
> standard format for reports coming back from the log check
> process. I would think that it would be good to develop a
> standard report format that is designed for easy import back
> into the logging programs for post contest analysis. Logging
> program writers everywhere UNITE! Its time to demand
> something back for all the work put into generating Cabrillo!
>
>
> David Robbins K1TTT
> e-mail: mailto:k1ttt@arrl.net
> web: http://www.k1ttt.net
> AR-Cluster node: 145.69MHz or telnet://dxc.k1ttt.net
>
>
>
>
>
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
|