[CQ-Contest] Domestic Contest Spots/Cheating

[Kenneth E. Harker]

On Fri, Oct 14, 2005 at 08:42:46AM -0500, Bob Naumann - W5OV wrote:
> N7MAL said:
> > Here is the maximum you are going to get from me. STOP asking for 
> more because it's not going to happen under any circumstances. <
> 
> Mal,
> 
> With all due respect, if you make a claim that there is rampant 
> cheating, and cannot back it up, I think it is reasonable for one to 
> conclude that you are not being honest.
> 
> Your presumption that any single op that works you within minutes of a 
> packet spot therefore must be cheating is simply not valid.  

Mal never said that he could or would conclude that any particular station
was cheating.  He merely said that the number of stations submitting logs 
as unassisted who worked him during periods immediately after he was spotted
seemed higher than reasonable.

Let's consider what's reasonable...  Let's say that after being spotted, 
your rate doubles for 10 minutes.  For agument's sake, let's say pre-spot
rate was 48/hr and post-spot rate was around 90/hr.  So, in those ten minutes, 
you work roughly 15 stations instead of 8, for 7 additional stations.  Now,
of those 7 additional stations, at least a few probably also found you 
by S&P, not by packet spot.  Why?  If you're only working stations at 
48 per hour, you're calling CQ a lot and depending on your CQing style, 
where you're beaming, etc., some station might be passing you by.  But at 
90/hr, there's less dead air, and stations from more places are likely 
working you, making it that much easier for someone to figure out that 
there's a station on frequency.  So, let's say 2 of those 7 were S&Pers,
for arguments' sake, and that the remaining 5 were attracted to you through
the packet spot.  If those five submit logs, they should be submitting 
them as either multis or assisted stations.  So 5 out of the 15 you work
in that time period is 33%.  Since not everyone submits logs, what you 
need to do is look at those of the 15 who did submit logs and see if 33% 
of them submitted as multis/assisted or not.  (And that's a little 
conservative, as probably some of the stations that find you through 
tuning are multis/assisted anyway.  The real percentage might be more
like 40% or more - but leave it at 33% for argument's sake.)

If, over time, and being spotted enough times to make this kind of analysis 
meaningful, you notice that fewer than 33% of those station who work you
within 10 minutes of a spot and submit logs in the contest are submitting 
as multis/assisted, then you might conclude that there's a problem.  In 
particular, if instead of 33%, the percentage was close to the same as the 
percentage in general for the entire contest, that would also be telling 
evidence of widespread cheating.  You'd still never be able to identify a 
particular cheater, but you could infer evidence of a problem.

I'd love to see Mal work through his logs in this manner and provide 
some quantitative data, rather than just qualitative observations.  I think
it would also be more interesting if it was done with logs from a station
that gets spotted more often than Mal, but not so often that they are 
being spotted constantly.

-- 
Kenneth E. Harker WM5R
kenharker at kenharker.com
http://www.kenharker.com/