Rasmussen Reports, under fire for some questionable poll numbers that keep showing Republicans doing too well, insists its methodology is fine when it comes to measuring favorability and unfavoraability of candidates -- which also are a measurement of name recognition. Here's something we overlooked in their Feb. 18 poll:


The poll was done on Feb. 17. We have no idea who paid for it, which makes it even a little more suspect. But, be that as it may, here's the thing:


On Feb. 17, Rassmussen asked favorable/unfavorable ratings on the two Republican candidates for US Senate (the two who are actually running, not the guy supposedly thinking about it. They found that Terrence Wall's favorable-unfavorable numbers were 34-35, with 31% having no opinion.


The numbers for David Westlake, the other GOP candidate, were almost identical, and within the margin of error, at 33-31, with 36% having no opinion.


But Wall had been on statewide television for nine days before the survey was taken. And Westlake, who's an unknown who has not run any television, is equal with him?  Not likely.  TV doesn't necessarily work miracles overnight, but there should be a measurable difference, especially when one candidate's the only one on the air.


So: Hold page one, we've got a scoop buried in the numbers. Political television advertising doesn't work.


Wait'll that gets around. Candidates will be back to making speeches on the street corner, like Fighting Bob LaFollette (pictured) used to do. And campaigns will be really cheap.


Wanna bet?