I'm not sure who the "investigators" are over at geekosystem, but while they present a lot of data, they fell over when it came to analysis of the situation (as

I'm not sure who the "investigators" are over at geekosystem, but while they present a lot of data, they fell over when it came to analysis of the situation (as did numerous others). Scenario: For a time period, contestant B received votes at a rate of roughly 3 votes for every 2 that contestant A received. Contestant B had a different value for an unknown field that was combined as part of the cookie used in voting/vote confirmation. The extra value on contestant B's call is meaningless if they're estimating that the vote rate was determined by contestant A's votes. Based on the data reported of the two time periods: Shawn saw a 100% increase in votes/minute David saw a 66% increase in votes/minute Cheryl saw a 175% increase in votes/minute Maria saw a 28% increase in votes/minute Jasbina saw a 100% increase in votes/minute Jacqueline saw a 13% increase in votes/minute Kornelius saw a 10% increase in votes/minute Zach and Phyllis saw a decrease in votes/minute Total rate of increase in voting (excluding Zach and Phyllis) was 31% increase in voting rate for the non-top 2 positions. If we include Zach and Phyllis, there was a drop in voting rate to almost 50% of the previous number of votes per minute. Determination: eid probably played a large role in the cause of what appeared to be tampering of voting. There's no correlation between votes of any candidate and any other candidate. The most likely cause? The reported vote counts are not 100% live. The votes are processed on a server and displayed via a cache. Most likely EID is a way to force vote counting onto a specific system, which caused a backup of cast but un-counted votes for Phyllis. While the backup was being cleared, the rate of voting appeared distinctly high. Once the cause of the backup was identified and resolved, the backup was quickly cleared out and the rate of votes appearing took a sharp drop as there was no backlog for that candidate and the others still had backlogs to process.