I have no idea what Peter Robinson is talking about when he asks if I can read timestamps.

He suggests I’m somehow being unfair and that last night, January 17th, shortly after 11 p.m., he wrote that “all the polls show Fred trending sharply upward.”

I appreciate the solicitude, reply by assuring you that I can indeed read numbers, and ask, I’m afraid, if you can read timestamps. I put up that post last night, when the polls listed on RealClearPolitics did indeed show Thompson having gained several points—or some 30 to 40 percent—for the week. The time-space continuum being as unforgiving as it is, I could not have known, last night, that that pleasing set of polls would yield, today, to more mixed result, including a couple from which I would rather avert my gaze altogether.

But every one of the polls that I cited in my post were completed on the 16th. All of them were released on the 17th. None of the polls that I cited came out after his posting. When he says, “all,” I guess he means, “two of the four that were released today.”

I also think he’s stretching things when he declares ”Thompson gained 30 to 40 percent for the week.” Two polls completed on the 16th showed a jump:

Rasmussen, 1/16: Thompson 16 percent. Preceding a week, completed 1/9, Thompson 12 percent. (In between, Rasmussen completed a poll 1/13 that put Thompson at 16 percent.)

Survey USA, 1/16: Thompson 17 percent. Preceding week, 1/6, Thompson 11.

So yes, Rasmussen showed a jump of four percent, a bit outside the margin of error on his nice, large-sample polls. I guess when you start at 12 percent, it’s fairly easy to get that 40 percent jump. (Also note that Rasmussen had Thompson at 18 percent in early December.)

As for Survey USA, yes Thompson increased six percent in a week… in polls where the margin of error is 4.6 percent.

Has Thompson gained ground? Sure. But he more or less hit bottom in the past weeks. He actually dropped below 10 percent in the RCP average right after Iowa.

I’d like to see Thompson surging into first place, or even a strong second, but the polls don’t show anything like that right now. I think we do folks a disservice by suggesting that they’re showing something that they’re not showing.