CQ-Contest
[Top] [All Lists]

## [CQ-Contest] Crosschecking, Penalties, Accuracy, Log Massage

 To: cq-contest@contesting.com [CQ-Contest] Crosschecking, Penalties, Accuracy, Log Massage "José Nunes CT1BOH" Mon, 6 Mar 2006 12:14:04 +0000
 ```CROSSCHECKING It is obvious to the majority of contest entrants that the level of log adjudicating has been increasing every year. Notably the CQWW Contest has been the leader in this task, and it is without a doubt one if not the most well adjudicated contest in the world. The number of electronic logs sent to the contest administrators has been growing every year and because of this, the level of cross checking is very high compared to what it was some years ago. The entrants, thanks to the penalty system into their final scores, feel the impact of this high level of crosschecking the QSOs in a very hard way. Lets look at some real numbers: Imagine the following performance by a contest entrant in the CQWW Contest: 7000 QSOs 700 multipliers 3 points per QSO for all the QSOs QSO error rate (bad Calls and not-in logs) = 3% The entrant would post the following claimed score: 7000*700*3 = 14.700.000 points Now let's imagine three "Cross-checking" scenarios from the Contest adjudicator: 30%, 60% and 90% of all QSO's are crosschecked. Take a note that in a big contest like the CQWW Contest crosschecking is around 60% (http://www.qsl.net/ct1boh/accuracy.htm) for a top SOAB entrant. After the work of the contest adjudicators the final score will be: Scenario: 30% Cross-checking Claimed score = 7000*3*700 = 14,700,000 points (Bad Calls and Not in Logs) = 0.30 * 7000 * 0.03 = 63 QSOs Penalties = 63 + 63 * 3 QSOs (CQWW penalty) = 252 QSOs Final score = (7000 – 252 ) * 3 * 700 = 14,170,800 points Score reduction = 529,200 points (3.6 %) Scenario: 60% Cross-checking Claimed score = 14,700,000 points Final score = 13,641,600 points Score reduction = 1,058,400 points (7.2 %) Scenario: 90% Cross-checking Claimed score = 14,700,000 Final score = 13,112,400 Score reduction = 1,587,600 (10.8 %) The more electronic logs a contest sponsor receives, more crosschecking is possible, better adjudicated the scores are and more score reduction is inflicted in the logs for the same accuracy level. PENALTIES Lets now assume we have three different entrants with exactly the same contest performance, but different accuracy rates - 1% error rate, 3% error rate and 5% error rate (assume 60% crosschecking of QSO's): Entrant 1 Claimed score = 7000 * 700 * 3 = 14,700,000 points Accuracy = 99%; QSO error rate (bad calls and not in Logs) = 1% Final score = 6832 * 700 * 3 = 14, 347, 200 points Score reduction = 352,800 points (2.4%) Entrant 2 Claimed score = 700 * 700 * 3 = 14,700,000 points Accuracy = 97%; QSO error rate (bad calls and not in logs) = 3% Final score= 6496 * 700 * 3 =13,641,600 points Score reduction= 1,058,400 points (7.2%) Entrant 3 Claimed score = 700 Accuracy = 95%; QSO error rate (bad calls and not in logs) = 5% Final score = 6160 * 700 * 3 = 12,936,000 points Score reduction= 1,764,000 (12%) With this score reduction numbers example ranging from 352,000 points to 1,7 million points it is amazing some contest entrants do not realize that accuracy is the best point-value asset in today's contesting performance. Because of the high level of penalties for errors (bad calls and not in logs) it is absolutely important to log accurately: If not absolutely sure of a call ask for a repeat. If not absolutely sure of a QSO, don't log it. ACCURACY If accuracy is so important what can an entrant do to improve his error rate? There is no straight answer to this but it evolves around being a better operator: Experience Carefully hearing Logging only when absolutely sure Knowing what can cause an error Log only exact timed QSOs … There are so many little things and they all add up, but being aware of the impact of careless operating is a good starting point to improve accuracy. LOG MASSAGE There are some log accuracy tools to improve performance during the contest. The most widespread one is SCP (super check partial). If used correctly and not to help guess calls it is a great tool. Also some logging software provide N+1 calls, which are calls that differ from the copied call by one character, against a call sign database and can alert the entrant to a possible mistake. There are some log accuracy tools to improve performance after the contest and before the log deadline. This is a very tricky area to say the least, but with 30 days deadline to send logs, anything goes: Apparently it is OK to change a call if it is an obvious typo. Imagine after the contest you find in your log P4oE. Most entrants would surely change it to P40E (CQWW annotated rules). Apparently it is OK to remove a call from your log if you think it is a bad call (CQWW annotated rules). It is not OK to change calls after the contest (annotated rules). How can an entrant identify a bad call in order to remove it from the log? Until now only the most sophisticated users had the tools to do this, and the procedure was very limited but nowadays anyone can do it. SH5 (http://rescab.nm.ru/) is a free contest analysis software available for download that reveals potential mistakes in the reception of calls, checking against a data base after the contest is over. Because the accuracy score penalties are high there is the risk of "not so skilled" contest entrants removing from their logs potential bad calls (good calls after all + real bad calls), hoping that the outcome of this is better than leaving the potential bad calls and facing the removal of the real bad calls plus the 3 times bad calls penalty. The problem with this action is it will have a great impact in others scores, because by removing potential bad calls from his log the entrant will cause a NIL in the logs of the good calls he worked in the contest but removed after the contest based in his potential bad calls list from the post contest log checking tool. IT IS A VERY PERVERSE SIDE EFFECT - in order to try to improve his accuracy the entrant will cause inaccuracy in others logs. And there is no point in reducing the 30 days deadline to submit logs, because all these actions can be executed in a couple of minutes although it is true that reducing the 30 day deadline to submit logs would definitely reduce cheating combining the use of contest recording audio and these after the contest tools to change calls. I appreciate that WRTC06 organizers require recording of the contest in their rules for the championship. This is the way to avoid tampering of logs after the contest is over at 23:59. May be contesting software could generate a log/time stamp code to avoid any changes to the contest log after 23:59. It is interesting to see that the good efforts by contest adjudicators to penalize inaccuracy and the ever increasing score reduction penalties may be backfiring with increased tampering of logs by the entrants made possible by the widespread availability of post contesting log checking tools. 73 José Nunes CT1BOH, CT3NT www.qsl.net/ct1boh _______________________________________________ CQ-Contest mailing list CQ-Contest@contesting.com http://lists.contesting.com/mailman/listinfo/cq-contest ```
 Current Thread [CQ-Contest] Crosschecking, Penalties, Accuracy, Log Massage, José Nunes CT1BOH <= Re: [CQ-Contest] Crosschecking, Penalties, Accuracy, Log Massage, K3BU Re: [CQ-Contest] Crosschecking, Penalties, Accuracy, Log Massage, Marijan Miletic [CQ-Contest] Better to submit a half log or no log??, Don Moman VE6JY Re: [CQ-Contest] Better to submit a half log or no log??, Mike McCarthy, W1NR