Steve, I noticed the same pattern and not only from Gallup. One other possibility (of many) is the overwhelming number of ads working people are exposed to on the weekends.

Lorie Byrd noted something else in a note earlier after pubbed:

"I think one aspect you did not include is that many people just aren't home to answer pollsters on the weekends. This is something long believed to help Dems. Anytime there is a weekend poll it is looked at suspiciously."

She's right as well, and also couches it with a 'take it for what it is or isn't worth' caveat.

Food for thought in any event.

I go with the theory that there are more Dems answering the phone on weekends, for whatever reason. My question is why there is not a single step difference between weekends and weekdays. Is this because the graph you are showing is three-day averages of tracking polls? A more interesting look would be just the single-day results.

Interesting, but let me throw cold water on your assertion....

Gallup's daily tracking poll is a 3 day running poll... so in a way the poll is 3 days old by the time it is released.

O/T but worth checking out and PASSING ALONG to everyone you know:

hattip from Instapundit regarding LA Times refusal to let tape be released:

Obama said to rile audience up during toast. He congratulates Khalidi for his work saying "Israel has no God-given right to occupy Palestine" plus there's been "genocide against the Palestinian people by Israelis."

The LA Times doesn't think anyone has a right to know this...WE DO!! Please pass this along to everyone you can. MAYBE that will force their hand

Check with DJ as to the why's but polls generally skew more toward the democrat on week ends, guess Republicans have better things to be doing.

Interesting things is with the jump in McCain's numbers and the drop in Obama's in yesterday's today's could be very interesting since the last day in the average had to be heavy in favor of McCain, 50% range to have moved the 3 day averages that much from what was fairly stable for a week.

This is a pretty well known trend. Polls that include weekend polling days almost always are biased towards Democratic candidates.

Bob Krum recently wrote about this (via Instapundit):

http://www.bobkrumm.com/blog/?p=2034

but people who monkey around in poll analysis have seen this trend for many election cycles.

How did you guys feel about the polls when Sen. McCain was up?

No snark, just curious. In the archives there's a post of Kim's reporting that Sen. McCain was ahead post-convention, and I don't recall much skepticism. Why do you suppose that would be?

Parthenon, check this post of mine from 2004:

http://polipundit.com/index.php?p=4281

Critiquing polls is something I have done for YEARS.

In the archives there's a post of Kim's reporting that Sen. McCain was ahead post-convention, and I don't recall much skepticism. Why do you suppose that would be?

Because the polls that had McCain up by a point or two at the convention's end did not set off our bullshit alarms. However, other polls, namely the ones that had Obama up by double digits in traditionally red (or purple) states, did set off our bullshit alarms.

That, and also, we can see that most media outlets are totally in the tank for Obama. That sets off alarms, too.

Because the polls that had McCain up by a point or two at the convention's end did not set off our bullshit alarms.

Fair enough. But you understand my curiosity, no? Consistency is a difficult thing. I admit I try but don't always succeed.

Critiquing polls is something I have done for YEARS.

And keep up the good work. Just preferably with more paragraph breaks.

"No snark, just curious. In the archives there's a post of Kim's reporting that Sen. McCain was ahead post-convention, and I don't recall much skepticism. Why do you suppose that would be?"

Post-convention is always crazy and unpredictable - for both parties.

And just for the record, I do not rely on polls. I don't need a poll to tell me whether I'm right or wrong. I make my decisions and to hell with whether I end up as part of a larger or smaller group.

I've looked at polls in the past as simple measures of public opinion - their real and original intent. However, for many, many years I haven't believed that's so much the case. When pollsters of all stripes, large and small, realized they could actually drive public opinion they became little more than advertisers.

Consistency is a difficult thing. I admit I try but don't always succeed.

Which is why I avoided the practice of jumping up and down and gloating when McCain was ahead by a small margin. Live by the polls, die by the polls.

Also, the virtue of consistency is based on all other things being equal. My point was that we had good reason to believe that, at least in these circumstances, all other things are not equal.

But yeah, I see why eyebrows get raised.

Midweek Republicans don't have time to screw around with pollsters. Starting Friday night, they will usually talk.

Look at the delay - it starts with polls on Tuesday and then changes on Friday. I don't think that the poll results has anything to do with who will vote for who, when.