That's wierd. I would have figured a week-end sample would hurt Obama given his demographic strengths. I mean, a sizeable portion of his voters are out trying to hook-up on fridays and saturdays. Perhaps the poor economic conditions have made mixed drinks so expensive that we all stay home.

Seriously. 15 dollars for an 8 ounce White Russian with crap vodka in it? What is this world comming to.

i don't know how the scene progresses through the week in DC, but i would note that here in LA, the uber-cool kids i know go out early in the week and stay home fridays and saturdays b/c it's amature night at the bar. if you want a seat and some conversation, tuesdays are a pretty good deal.

maybe the whole paradigm is flipped: Hill's people are the friday night dinner and a movie crowd. Obama's are the monday/tuesday furtive-drinkers-on-a-schoolnight.

If you check over at pollster.com they note that this is not a general problem but one which is specific to Gallup.

http://www.pollster.com/blogs/day_of_week_effect_in_gallup_d.php

Rasmussen the Day of Week (DOW) is also observed at Rasmussen but the effect is totally different.

Gallup also makes a plausible argument that the problem is limited to a single month. This does make some sense as any given month only contains 4 of any particular day of the week. This is a very small sample and therefore false patterns could easily be observed. Gallup shows that when the number of weeks increase the effect appears to go away...

http://www.pollster.com/blogs/re_2_dayofweek_effect_in_gallu.php

There is yet more detail on this discussion over at pollster if one is interested...

Via TPM: Gallup got back to them and said there is no significant flaw in their polling averages.

http://tpmelectioncentral.talkingpointsmemo.com/2008/04/gallup_no_our_daily_tracking_p.php

This is certainly interesting stuff. Could it be that women are polled/at-home more during the week than the Obama types of supporters? It has me scratching my head in fascination.

For a bit of humor, take a look at these redactions from the Clinton Library:
http://ladycatherinebedamned.blogspot.com/2008/04/what-was-in-clinton-library-redactions.html

people feel more uptight early in the week. As the weekend comes,they get ready for CHANGE!

Maybe you're more likely to get women and retirees at home during the week?

Also, don't all of these polls significantly underrepresent young people b/c of cell phones anyway? So maybe that factor just drops out.

Well, if it's only a few percentage points switching during the week, maybe Gallup has simply stumbled upon a group of flighty voters that, when the weekend is here and they're excited and looking happily to the future, they want to vote for Obama, but when the work week begins and things seem like a grim sludge, they want the kitchen table comfort of Hillary?

Well, I guess I will be the official young-person spokesperson and note that I am far more likely to answer my phone during the weekend when plans are being made and there is fun to be had, while during the week I can usually bet that calls coming in are nothing important (and, thus, I am more likely to let them go to vm.)

Maybe more Obama supporters are in church on Sunday.

According to NPR, Obama supporters are just a bunch of latte-drinking "starbucks democrats". Maybe they were at their elite coffee establishments when the pollers called.