Somewhere in Toledo, OH, a plumber is filing to rename his small business.

My thoughts this time aren't quite as organized.  I tried keeping a log of all the questions, but I realized early on that this debate was going to be all about style over substance.  Sure, meaningful views were articulated, but those views were never going to drive the narrative.

Oddly, I came to this conclusion long before negative campaigning was mentioned; long before Ayers and ACORN entered the debate.  I came to this conclusion when McCain failed to give coherent answers to the first two questions.  This problem continued throughout the night.

There's more...

We're back with another liveblog!

New contributor Student Redux here, and it's time for my post-debate rundown.  I watched the debate on MSNBC, so I have a bit different view than the CNN crowd here.  But before I get started, let me give a short digression about dial groups.

I have to admit, I've got a lot of sympathy with Republican strategist Mike Murphy on this issue.  Dial groups are very questionable metrics, and easily overinterpreted.  Following CNN's lead, MSNBC has also started running dial groups for each debate (although MSNBC doesn't provide realtime dial tracking during the debate).  Let's take a moment to consider the nature of these dial groups, though.

There's more...

We're back, and tonight's post is about CNN's dial groups.  Dial groups are a form of focus group in which people are sampled by strata and given a handset. This handset has an analog dial that allows people to indicate their responses to stimuli, favorable to unfavorable, usually graduated 1 to 100.  The output from these handsets are then aggregated by strata and averaged to produce a function of time. Here's an example:

CNN has put together a group of 99 people from the state of Ohio. This group is 33 Republicans, 33 Democrats and 33 Undecided/Independent voters.  All the output from the each group's handsets will be added together and divided by 33, and then that number will be graphed as the y variable, with time as the x variable.  Each stratum will be given its own function.  That way, you can see how people are reacting in real time to the speeches or events.

There's more...