Pollster.com

This race is kind of frozen. Nobody but Newsweek found a sudden movement among the electorate since their previous poll. I really think that people will start paying attention after the conventions. Thank God the conventions start at the end of August. This is becoming kind of boring, don't you think?

This is right in line with average of other non-tracking polls over the past three months even though the Gallup and Rasmussen trackers show a tighter race. It's also in line with overall trend in state polls.

Despite claims by some of a tightening race in the past couple of weeks, nothing has changed in this race for the past three months. And nothing will unless there is some dramatic event that shakes up the race, which is highly unlikely.

Obama got a good boost from his tour. Media is doing a good job on Obama's campaign this year.

It is too early to tell if Obama is getting a bounce out of his overseas trip. Could the race have tightened and Obama is starting to get a bounce? The only way we will know is if they release the day-by-day numbers. I suspect the race was tightening and he is starting to get some traction. The right-track/wrong-track number is very, very, very bad news for McCain.

Todd made the point not to over-analyze the four-way horserace question because it was conducted on a half-sample, and thus the sample size is smaller. However, it occurred to me that the four-way horserace results may be related to the two-way results for the half-sample. If in the two-way race Obama's margin over McCain is higher in the half-sample, then that margin would be reflected in the four-way race. Perhaps Mark, you could dig up the two-way results for the half-sample? It did not seem to be reported in the results pdf.

On another note, the gabfest on MSNBC's Morning Joe this morning made fun of Barr's support being below the margin of error of the poll. Of course, as we know, the margin of error for Barr's support is calculated from his sample proportion, 2%. My back of the envelope margin of error for Barr's support is +/- 1.2% (using a 95% confidence interval and assuming a simple random sample of 500, which would be about the size of the half-sample). So Barr's support is likely not actually below zero, which makes sense since we know there were at a minimum a few respondents to the survey that support Barr (in the unlikely event that his only supporters in the entire country were in the sample).