The steady increase in Romney's percent of the vote and steady decline in Santorum's represents a statistical anomaly. In this case, the anomaly is amazing according to the researchers. They argue that the probability of this happening by chance alone is so small it exceeds the capability of statistical packages to handle. Their software says Romney's share of the vote, increasing with precinct size has zero probability of occurring by chance alone.
The significance of the Wisconsin analysis is of grave concern. Presuming the use of appropriate statistical measures and analysis, human intervention is the most likely available explanation.
Vote flipping gave Romney a 57,000-vote victory over Santorum in Wisconsin. Absent vote flipping, Santorum would have won over Romney by about 54,000 according the group's analysis.
Was Wisconsin the only state where Romney's share of the vote increased in this way as precinct size increased?
There were eleven states that showed this amazing anomaly, Romney gaining in votes and margins as precinct size increased. The chart below shows the estimated vote flipping for eleven of the fifty states analyzed by the group using precinct-level data.
Estimated Votes Lost/Gained in 11 State Primaries - 2012 - Vote Flipping