[CQ-Contest] [wrtc2014] WRTC Category Weighting Factor

Thu Aug 23 09:56:44 EDT 2012

As far as the statistics are concerned. All we need is the raw data.
Not everything, but just: callsign, claimed score, final score, entry
class. (That's what's already avialable today based on what a Cabrillo
file can provide).

Data elements like country, continent, CQ zone,  etc., can be derived
from the callsign.

It would be also nice to have though: Age, Total operating time, and
probably grid locator.

I'm including claimed score as a way to get statistics on logging accuracy.

If UBN reports were made public too, we could also produce statistics
for the most common logging errors on each mode and stuff like that.
It's quite interesting. I did that for our entries over the past five
years on a per Operator basis, and we were able to determine which
type of error was being made by each of the ops of the team.

We can really have a big universe to start our calculations for
several contests.

Data is actually available. Claimed score, final score, callsign and
entry class. Problem is formatting it in the way we need may take a
considerable ammount of time. I'm  sure contest orgranizers have it in
a more "tabular fashion" and probably normalized which helps speed up
the process of producing stats.

Vy 73.

Martin, LU5DX

On Thu, Aug 23, 2012 at 10:14 AM, Michael Adams <mda at n1en.org> wrote:
> I realize that I have an odd sense of what counts as "interesting" and
> "fun", but....it might be kind of fun and interesting if a collection of
> stats were made easily available for public review and analysis.
>
> So far, the discussion seems to support the assumptions I made when I
> mulled them over one night.  Besides, as a perpetual LP entrant with a very
> modest station, I'm just playing for the fun of it (and to fill out
> band-country charts).  WRTC qualification is a spectator sport for guys
> like me.
>
> It's clearly too late to modify the selection criteria for WRTC2014, but
> perhaps would-be WRTC2018 committee members are already thinking ahead to
> potential changes and improvements.
>
> I would argue that "the best" qualification scoring system would be the one
> that does the best job at predicting WRTC results.   With a body of
> publicly available data, different qualification schemes could be developed
> and tested against one another.  It becomes a simple predictive modeling
> exercise, really.
>
> I suspect that, at least among "reasonable" possible scoring systems, the
> list of qualifying entrants would be pretty similar.
>
> For whatever it's worth, I can't help but wonder what dynamic category
> weightings would do to qualification scoring (i.e., have the category
> weights for a particular qualifying event be based at least in part on the
> competitive index for that event).
>
> --
> *Michael D. Adams* (N1EN)
> Poquonock, Connecticut | mda at n1en.org
>
>
>
> On Thu, Aug 23, 2012 at 12:30 AM, Martin , LU5DX <lu5dx at lucg.com.ar> wrote:
>
>> On Wed, Aug 22, 2012 at 10:23 PM, Chris Plumblee
>> <chris.plumblee at gmail.com> wrote:
>> >
>> > Syl,
>> ...
>> >
>> > The reality of our hobby is that the best measure of who is the best
>> > operator is who can score best in the single op categories, and the
>> > category with the most serious entries and the most competition on a
>> > regional, national, and international basis is almost universally single
>> > op, high power. The scoring weight for single op low power was adjusted
>> > upward this year in an attempt to be more equitable, as Dan pointed out.
>> >
>> > ....
>>
>>
>> Hello Chris.
>> Your opinion is solely based in your own perception of reality.
>> But if you do the math (see below) your statement seems not to
>> accurately represent it (reality).
>> I did not do the numbers for LP in this case, because I wanted to show
>> the case of an even less rated entry category (in terms of WRTC
>> selection criteria), that is, SOAB(A) HP, which has a weighting factor
>> of 0.8
>> Raw data was taken from the great site http://www.pileup.ru
>> More precise calculations can be done using the organizer's score data
>> base with final numbers.
>>
>> It would be really interesting if Valery (pileup.ru) or the contest
>> organizers can provide the raw DB data to manipulate it as needed to
>> start a solid statistical analysis about competitiveness and other
>> aspects as well.
>>
>> 2011 CQ WW DX SSB total number of entries SOAB HP: 999
>> Total sum of Claimed scores for SOAB HP: 879,495,650
>> Average points per station in SOAB HP (Total sum of claimed points /
>> (total number of logs - checklogs) = 879,495,650 /  (999 - 41) =
>> 918,053.9144
>>
>>
>> 2011 CQ WW DX SSB total number of entries SOAB(A) HP: 805
>> Total sum of  Claimed scores for SOAB (A) HP: 751,417,601 points.
>> Average points per station in SOAB (A) HP (Total sum of claimed points
>> / (total number of logs - checklogs) = 751,417,601  / (805 - 49) =
>> 993.938.6257
>>
>> According to N0AX's formula for determining Competitiveness of a
>> category, that is,
>> Average of top ten scores / top score (in a given class)
>>
>> We can easily determine that:
>> In the 2011 CQ WW DX
>>
>> SSB SOAB HP has a Competitiveness index of: 0,740468333
>> whereas
>> SSB SOAB (A) HP has a Competitiveness index of: 0,672518322 (In this
>> case the competitiveness index is actually deviated by the score of
>> one station P40A with a big geographic and DX status advantage over
>> the rest of the top ten entrants. Most of them from Europe, two from
>> the States one from A6 and ST2AR who may have a higher DXCC status
>> ranking but he's using very simple antennas (singe tribander and
>> wires).
>>
>> In the 2012 CQ WW DX CW the competitiveness index favors SOAB(A) for
>> the top ten entrants (ballpark figures, since I don't remember exactly
>> thoug I did thte math):
>>
>> SOAB(A) HP 0.82
>> SOAB HP: 0.80
>>
>> This is not absolute the right perspective since it's done on a world
>> wide basis. This is just to demonstrate that in part, your statement
>> is not valid.
>>
>> Who are the best operators can only be determined under a very, almost
>> totally, better said, TOTALLY controlled environment. Which is not the
>> case of contests other than WRTC.
>>
>> Nevertheless, the criteria to determine who wins a place as a team
>> leader is based upon human opinions and perceptions rather than based
>> on statistical facts.
>>
>> Based on the numbers SOAB LP can never rank higher than SOAB(A) HP if
>> competitiveness is a factor. Neither can MS rank higher than SOAB(A).
>> Some would say, MS aligns more to the WRTC style of operation.
>> Not true. MS teams outside of WRTC can be formed by a high number of
>> operators, distributing the working hours by a bunch and making the
>> need for stamina, endurance, concentration by each operator a whole
>> lot less than a SOAB(A). In fact, MS can use packet or web clusters,
>> but they still get a weighting factor of 1!!
>>
>> Anyways, nice discussion.
>> Hope to meet you in W1 even if we go as visitors :-)
>>
>> Vy 73.
>>
>> Marti, LU5DX
>> _______________________________________________
>> wrtc2014 mailing list
>> wrtc2014 at lists.wrtc2014.org
>> http://lists.wrtc2014.org/mailman/listinfo/wrtc2014
>>
> _______________________________________________
> CQ-Contest mailing list
> CQ-Contest at contesting.com
> http://lists.contesting.com/mailman/listinfo/cq-contest