CQ-Contest
[Top] [All Lists]

Re: [CQ-Contest] QSO Penalties

To: "'Kenneth E. Harker'" <kenharker@kenharker.com>, K3BU@aol.com
Subject: Re: [CQ-Contest] QSO Penalties
From: "Georgens, Tom" <tom.georgens@engenio.com>
Date: Sun, 15 May 2005 19:08:18 -0700
List-post: <mailto:cq-contest@contesting.com>
As I see it, the question is whether the score reductions are intended to be a 
penalty or an estimated correction of the score to the true result (valid Q's 
times valid mults).  

Before the sophisticated cross checking we have today, the likelihood of 
uncovering a broken call was much lower.  Therefore, the issue was to estimate 
how many bad QSO's were going undiscovered for every one that was detected.  
CQWW historians would know better, but I suspect that estimate was the basis of 
the original 3X reduction.  If I remember correctly, the reduction used to 
accelerate to 10X when an error threshold was achieved.  It seems like a 3X 
penalty to equalize to the true score and a 10X penalty to punish the truly 
sloppy seems quite logical to me.

Now that the cross checking is much better (and we contesters owe a debt to 
those who made this possible), is the 3X penalty still appropriate to equalize 
the score, or has it become a true penalty?

I would suggest that, based on my imperfect operating, the 3X penalty is still 
likely to yield a score higher than the product of truly valid QSO points and 
Valid mults.  Looking at my logs, I had 9332 claimed Q's.  The U+B number 
totaled 145 reducing the "common" q's to 9187.  Most of the experts believe 
that almost all of these are truly busted calls but the committee does not dock 
us unless the error can be verified.  Add to this, about 25 lost NILs.  This 
number is limited by the fact that not all ops send in logs, otherwise it would 
be worse.

This means that my 9332 claimed Q's is likely to be overstated from my truly 
valid Q's by the number of U+B (145) and lost NIL's (25 and understated) for a 
likely true QSO total of 9162.  The actual score was reduced to 9254 Q's plus 
one lost mult for a total reduction of 3.5%.

My conclusion, pending further analysis by the experts, would lead me to 
believe that we have no reason to complain about the 3X penalty, our true 
result is probably worse.

As much as I find the UBN's humbling, I think they have contributed extensively 
to the quality of contesting.

73, Tom W2SC 8P1A

-----Original Message-----
From: Kenneth E. Harker [mailto:kenharker@kenharker.com]
Sent: Sunday, May 15, 2005 3:33 PM
To: K3BU@aol.com
Cc: cq-contest@contesting.com
Subject: [CQ-Contest] QSO Penalties


On Sun, May 15, 2005 at 12:31:05PM -0400, K3BU@aol.com wrote:
> 
> Time to make all logs public and remove the 3 QSO "PENALTY" rule - it is 
> insane, illogical!
> Like if you get speeding ticket, you would get three more so "you learn the 
> lesson?"
> Gimme a break! Let's stop silliness (or insanity) in ham radio!

The 3 QSO penalty rule makes perfect sense to me.  Anything else encourages 
guessing whenever there is uncertainty.  Consider the following options:

0 QSO penalty: If in doubt, it is _always_ to your advantage to guess at a 
  callsign or exchange element.  If you get it wrong, you are no worse off 
  then if you had (correctly) not logged the contact.  On the other hand, 
  there is always the chance that you can get it correct and gain points
  you did not truly earn.

1 QSO penalty: This is the break even point.  If 50% of the time you guess
  correctly, and 50% of the time you guess wrong, you break even.  On average,
  every time you are penalized is offset by some QSO where you guessed 
  correctly.

2 QSO penalty: Here is where taking guesses at callsigns or exchange elements 
  will more likely than not harm your score.  You have to guess right 2/3
  of the time to break even.  To put it another way, when in doubt, you
  have to feel at least 66% certain that your guess is correct before it 
  make sense (in terms of your score) to commit to your guess.

3 QSO penalty: Now you have to guess right 3/4 of the time to break even.
  This is a pretty high threshold to overcome, and I think it does a great
  job of discouraging guessing and sloppy operating, at least among those 
  who care about their score.  This penalty encourages getting it right 
  every time, and not logging those "almost" QSOs.

At the very least, there needs to be a 2 QSO penalty, but a 3 QSO penalty is 
better.

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

_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest

<Prev in Thread] Current Thread [Next in Thread>