Saturday, October 24, 2009

RAMPANT CHEATING? NAH. NOT REALLY...

A very prominent local triathlete told MTN recently that he doesn't vote for Race of the Year because he suspects rampant cheating. "Ballot Box Stuffing," were his words. And yes, he spoke in capitals.

Guess what?

That triathlete was right!

Well, sorta. Cheating happens, though on what we consider to be a "sub-rampant" scale.

What that triathlete didn't know until it was explained to him, was that the obvious cheating is detected and those votes are deleted and suspicious votes are "Red Flagged." In the end, we do in fact know which event deserves the award.

Here are some of the safeguards we employ:

* Finding & eliminating multiple e-addresses. Example: "TH" had voted three times, using her accounts with gmail, yahoo and hotmail. Two of her votes were deleted.

* Scrutinizing those events for whom more than 40% of their finishers vote for that event. Historically, ROYs have received votes from 35-40% of their finishers. In one case, 71% of the equivalent number of finishers of an event voted for that event. A closer look revealed lots of duplicates, multi-addresses and curious addresses (from people unassociated with the event or "goofy" addresses). This event was eliminated from consideration for ROY.

* Goofy Addresses. Actual Example: booger@nosepicker.com. We send test e-mails to these addresses, which invariably bounce back as undeliverable. They are then deleted.

* Curious Addresses. Example: A few years back a certain race got 60 votes from a certain municipal address. We red flagged these, knowing that not everyone who works for the City of St. Louis Park raced at the _______ Triathlon.

At the conclusion of voting, we subtract the number each race's red flags from that event's voting total. If doing so affects a potential winner, we scrutinize the addresses even further and contact the race director and send e-mails to addresses that don't seem to link up with names in the results. We have only had to do this on two occasions. Every other time, the number of possible cheaters, have not been large enough to, once subtracted, dislodge the event from its position among the top vote-getters.