The complaint comes from the difference in results between Rasmussenand other national pollsters, such as Gallup. However, they use twodifferent sampling techniques: Rasmussen polls likely voters, while Gallup and others poll adultsuntil the final few weeks before an election. The former is much morepredictive for elections, while polling a general population of adultsis the least predictive sampling technique. And thatdifference gives Rasmussen an advantage that has already been seen thisyear. Rasmussen first detected the erosion of support for Obama andObamaCare in late June, an erosion that other pollsters corroborated inthe fall as discontent spread from the politically aware to the generalpopulation.