Have the Gallup pollsters lost their minds?*

Today's tracking poll has Romney ahead by six points (51-45), which is five points better than the Republican-leaning Rasmussen survey. The latest results are based on responses before Tuesday night's debate, so presumably those numbers will narrow. But even putting Tuesday night aside, a six-point difference doesn't make a whole lot of sense if you look elsewhere on the Gallup website.

U.S. unemployment, as measured by Gallup without seasonal adjustment, is 7.3% in mid-October, down considerably from 7.9% at the end of September and at a new low since Gallup began collecting employment data in January 2010. Gallup's seasonally adjusted unemployment rate is 7.7%, also down from September. October's adjusted mid-month measure is also more than a percentage point lower than October 2011.

This follows another Gallup survey from Tuesday:

Americans' economic confidence, on a weekly basis, is now at the highest level Gallup has recorded since late May. The Economic Confidence Index score for the week ending Oct. 14 is -17, up from -19 the week prior and -21 the week prior to that. The current data are the first to include a full week of interviewing after the generally positive Oct. 5 jobs report.

How do you have an incumbent down by such a large margin on one side of your house when surveys on the other side show that the economic metrics keep improving? The NYT's Nate Silver notes that "even the best polls will give an inexact estimate of the condition of the race." Or put another way, some numbers are simply illogical on their face.

*Silver tweets:

The first rule of poll analysis is that if a poll looks like an outlier, it usually is.

Earlier: Two Gallup surveys that make absolutely no sense