In Depth Study By Radiosport Canada Raises Contest Cheating Questions Anew

By Jamie Dupree NS3T, radio-sport.net 

In a week punctuated by controversy surrounding the use of internet chat rooms in the CQ 160 CW test, two prominent contesters issued an in depth computer review of CQWW CW logs from 2006 which they say can help detect cheating in the future.

The effort by Sylvan Katz VE5ZX and Jose Nunes CT1BOH was publicly praised by CQWW Contest Director Bob Cox K3EST.

"This is a ground breaking attempt to try to "blindly" anticipate the category of an entrant," said Cox.

"This paper could lead to identification of questionabale entrants and more accurate certification of winners."

The mathematical breakdown produces a series of graphs that have a typical signature, or look to them, when evaluating contacts on a second radio.

Those stations that have a different "signature" can be reviewed more closely to see whether there are classification issues.

The study quickly found questions with the 2006 CW log of Jack Danielyan RW3QC, whose C4M operation in Cyprus was disqualified by the CQWW Contest Committee.

The study said there was a "notable difference" in the signature of the C4M log.

"It does not appear to be the signature that would be expected for a DX SOAB A (Assisted)" station, wrote Katz and Nunes.

"Log results deemed incompatible with category of entry," was the terse explanation in CQ Magazine as to why C4M was DQ'd.

The "Two Radio Event" study shed even more light on the C4M case, as the instances of second radio contacts put C4M on par with multi-operator stations in the Carribean like PJ2T and PJ4A.

"The article is more to get people who claim SO2R but in reality are using two signals on the air or are using more than one operator," CT1BOH told radio-sport.net.

Another log that raised red flags was another Assisted entrant, Serge Redkin UW8M, who finished 8th in the World in the 2006 CQWW CW test.

"On closer examination, the UW8M log appears to belong to a different classification," the report said.

"The log contained 60 percent more 2-Radio Events for 40 percent fewer QSO's than OM8A who won the MS (multi-single) class."

The study ended up tossing out the UW8M log because it skewed the data on Assisted stations from Europe.

The study authors admitted that their samples were small, but that did not worry CQWW Chief K3EST.

"The profiles generated in this paper lead to a consistently repeatable unique signature for SOAB, SOAB A and MS," said Cox. "Comparing a serious entrant's category to these profiles could help reveal outliers and further action could be pursued."

The study also urged other contests to adopt a public log policy, saying the move taken by the CQWW committee was a "bold step."

"It makes the adjudication process more transparent to the participants," the study said. "Hopefully other sponsors will follow suit and encourage the contest community to research and develop new ways to probe the art and practice of radio sports."

"An open log policy will increase peer scrutiny of the committee decisions as well as encourage the development of analytical techniques to help enforce the rules."

Other top operators agreed, saying they hoped it could be expanded in the future to cover even more logs.

"The profiling discussed in this paper could lead to identification of questionable entrants and more accurate certification of winners," said Scott Robbins W4PA, on his blog at w4pa.journalspace.com.