Let's start with a couple assumptions:

1. The McCain and Obama campaigns will fend off attacks from one another.  Our simplistic assumption is that they will be roughly evenly matched, so no McCain attacks on Obama will stick better than Obama attacks on McCain.

2. The primary way we can help Obama get elected is by volunteering to register new voters (giving Obama a +1 net) or by flipping McCain voters (giving Obama a +2 net).

What we're up against.

Let's dive into the math.  My goal here isn't to do anything complicated, and the numbers we have to work with are very rough, but the goal is to provide a conservative estimate of how much work we need to do.

1. Impact of media bias = 2.5%.  To quantify this, I'm going to use the national polling averages from 2004, as seen here.  Using the following graph, we can see when different events shaped the polls (here shown using EVs): I'd argue that the media's bias isn't quantifiable in a positive direction but we can see the effect of Kerry overcoming media bias: the debates provided a way for Kerry to push back against the media's repetition of Bush talking points.  Thus, the bounce in early October from 43.5% to 46.0% is what we'll consider the anti-media bias bounce.

2. Impact of 527s = 3.0%.  Using the same approach as above, from the start of the SBVT 527 ads to the end, Kerry went from 48.0% to 45.0% in the national polls.

3. Impact of voting machines = 1.1%.  Voting machines are the last piece of this little math problem.  Verified Voting has an amazing map showing which states use good voting machines and which don't.  (In their map, yellow and red indicate trouble.)  They estimate that modern voting machines have a failure rate of 1.1%, so we'll use that figure.

Total margin we need to provide: 6.6%.  In other words, using this rough calculation, we need to bump Obama by 6.6% of the national vote to ensure a win despite media bias, 527s, and bad voting machines.

How do we do it?

1. Each volunteer = +50 votes.  Let's figure out how effective each volunteer can be.  There are roughly 100 days left until the election.  Let's say that on average, a dedicated volunteer is one that can get a +1 for Obama every other day.  (Most likely in bursts.)  That means registering one new Obama voter who will actually show up on election day, every other day.

2. Total 2008 turnout = 150,000,000.  There were 121 million voters in 2004, which was 20 million more than in 2000.  Let's say there'll be 150 million voters this year, give or take.

3. Total new Obama volunteers needed = 198,000.  We need to bump Obama by 6.6% of 150,000,000, which means we need 9,900,000 new Obama votes.  If each volunteer gets us 50 votes, we need 198,000 new volunteers.

Summary

The takeaway from this diary should be threefold:

1. Everyone, and I mean everyone needs to volunteer.  Go read how to do voter registration and get started today.  Also check out the Obama campaign's action guide.  The 198,000 new volunteers we need are ones in addition to the volunteers the Obama campaign already has.  We need to not just donate or blog - we need to get on the ground, knock on doors in communities that have never seen the attention of a presidential campaign, and register voters.

2. We need to be smart about where to target voters.  Target areas of town where the population is in flux more - where folks may not have registered yet.  (For example, in apartment buildings.)  Some states have high population turnover (Nevada) and are ripe for voter registration efforts.  Show up at naturalization ceremonies, where there will be thousands of new eager citizens who I'd bet would be excited to vote for the first son of an immigrant for president.  You can find your nearest Citizenship and Immigration office here and ask them for a list of naturalization ceremonies.

3. We need to do more than fight fire with fire.  Pushing back directly or complaining about media bias, about 527s, and about bad voting machines might be cathartic, but it doesn't overcome the impact of those forces to ensure a win.  We need to do an end run around these negative forces by appealing directly to the voters.

Also cross-posted at Daily Kos.