(CBS News) The new CBS News/YouGov Electoral Vote Tracker - a computer simulation model of the Presidential race - shows that if President Obama were to win the national popular vote by three points (which is his current edge today in national polls) that would translate into 332 electoral votes for Mr. Obama and 206 for Mitt Romney on election night.
The model shows how the votes would most likely be distributed across all fifty states, if Mr. Obama's three-point lead held up on November 6th and that reveals which states would ultimately tilt to Mr. Obama, and which go to Romney.
The president would pick up enough swing voters from across the battleground states to take almost all of them. He'd win Pennsylvania, Ohio, Nevada, New Hampshire, Iowa, Colorado, Virginia; he'd win a close race in Florida, and win a squeaker in Wisconsin, while North Carolina would go to Romney.
Importantly, it isn't a forecast - it is today's score. We've got six-plus weeks to go, and this can change. We'll re-run the simulation again in the coming weeks.
This simulation is based entirely on data collected by CBS News and YouGov, a Silicon Valley research company working with CBS News on this project. The national lead of +3 is based on the findings of the latest CBS News/New York Times national poll. We then use over 36,000 additional interviews fielded by YouGov, matched to voter demographics across the states, and run the simulation. It's designed as a 50-state snapshot meant to show where things stand if we translate a national lead into the only score that really matters - the Electoral College vote.
How the CBS News/ YouGov Estimate works:
We know that every state contributes to the national vote. The trick is to pin down which states are changing as the national vote moves, and by how much.
Some of the national lead we see invariably comes from deep-blue Democratic or deep-red Republican states like California and Texas, but we already know where those states' electoral votes are headed. What we're really after is the portion of the national lead that's coming from tight battlegrounds like Ohio or Iowa. Those places start out close to 50/50 - that's why they're battlegrounds - and even small shifts in the vote within those states will tilt them.
We start by making state-level estimates that are, importantly, all in sync with the national vote. They're taken from interviews fielded at the same time as the national poll and they account for both battleground and non-battleground states.
We combined CBS News's national estimate with YouGov's online sample across 27 states, including all the battlegrounds, for over 36,000 added interviews. The remaining noncompetitive states are estimated using past and current vote percents by region. The online sample was weighted to match the demographics and vote intentions in CBS News's national survey.
Together, this gives a full picture of the race, along with detail at the state level that is sufficient to build an accurate model of the Electoral College votes each candidate would receive, if the election were held today.
Next, we ran a computer simulation model that considers all the possible outcomes that would lead to a national three-point Obama win, taking into account those estimates, and the range of potential swing that could happen in every state.
With the simulation, we also compute the candidates' mathematical chance to win each state. (As their lead shrinks, the chances of winning reduce.) Of all the states he would be estimated to win in this scenario, Mr. Obama's chances of hanging on to Wisconsin are most precarious, at .53. His chance to win Florida is much higher, at .63. That means 63 times out of 100 we ran the simulation, the vote distribution to make a three-point national lead tips Florida to Obama. By comparison, his chances of winning Iowa are stronger, at .77.
Anthony Salvanto (@salvantocbs) is CBS News' Elections Director
Doug Rivers is Professor of Political Science at Stanford University, Chief Innovations Officer at YouGov and a CBS News Consultant