[CQ-Contest] Unqiues

[Donald Field]

I've been following the discussion with interest, as I'm about to embark on the adjudication of the IOTA contest (some 1220 logs here so far, about 400k QSOs in the database).

First observation is that unqiues only apply to QSOs which cannot be verified by cross-checking of logs received. In other words, if my log has HC8K as a unique, but HC8N sends in a log, it pretty soon shows up that this wasn't a unique at all, but a broken call. Interestingly, for this contest (I don't have the stats for CQWW or ARRL), I can cross-check 80% or more of the QSOs in the logs from "casual" entrants, because they will have tended to work the more serious entrants, who will have sent in a log. Indeed, with the ease of sending a log these days, the percentage of players who send in a log of some sort (even if only as a check log) is increasing every year. I went one step further this time - for those island entrants who had made a reasonable number of QSOs, but hadn't sent in a log by the deadline, I tried to contact them to invite them to send in a check log. This has been amazingly successful - I was able to find e-mail addresses for something like 75% of the stations concerned, and probably some 75% of those responded to my pleas for a log. In other words, an overall success rate of something like 50%. I know other contest organisers (e.g. EU Sprint) have been doing something similar. All of this helps to reduce the issue of uniques.

So we are then left with a minority of QSOs in logs, where no cross-checking can be undertaken. In the past it was hard to do anything about these, with the result that a number of regular entrants started to pad their logs - this is why it became an issue. For some of these offenders, it actually made a significant difference to their contest placing; hence why contest organisers started to take the issue seriously. Nevertheless, I have been surprised to read that some organisers (e.g. LZ contest) appear to penalise uniques as a matter of course. The golden rule should always been "innocent until proven guilty". But there is quite a lot a contest adjudicator can do to determine whether a call really is a unique. For example, if I found "G2XTT" in a log, I'd bust it - not because it's a unique, as such, but because I know such a call doesn't and cannot exist. There are many similar examples (N6AA has done a lot of work on this as, I am sure, have others). Equally, in IOTA I have the benefit that it's a serial number contest.. If a unique appears, and had given a serial of 003, then it may indeed be a casual IOTA chaser who gave out a handful of contacts while chasing some new islands. I have the recourse, should I so wish, of e-mailing the station concerned, to ask whether the QSO was valid. If he says he didn't go near the radio that weekend, then it's a busted call .. If the received serial was 1349 (for example!) then there's something very wrong - anyone who had been that active wouldn't appear as a unique, so there is some other explanation, and I can start looking for it (an obvious broken call, perhaps?).

Which, I think, all goes to say that uniques should not be busted simply because they are unique, but that contest adjudicators have other means to determine, in at least some of the cases, whether "innocent" or "guilty". Of course, these methods can be time-consuming, so are probably only worth pursuing where there is a close call for one of the coveted certificates or trophies. For the most part, it really isn't worth the effort! What I do try to do, though, is to indicate in contestants' logs where a call is unique, even where no penalty has been applied. I believe this exactly analagous to the CQWW UBNs. It can be useful feedback for the contestants, and also demonstrates that the adjudicators are on the ball.

Incidentally, worse than uniques are those calls which appear maybe 20 or 30 times in the logs, and which therefore have some sort of spurious validity, but which turn out to be broken spots on the Cluster. A download from DX Summit soons enables me to spot these, and it's then quite interesting to note those supposed "non-assisted" entrants who happened  to "work" that broken call!

73 Don G3XTT (g3xtt at lineone.net)