Saturday, July 21, 2012 | rightcoast
Yes, there is still four months to go, and an electoral map prediction seems ill-advised and premature... But what the heck...
I based my projections on a modified RCP poll average: I included only recent polls of likely voters, ignoring RV polls. I also used a more conservative version of the "incumbent rule" for the undecideds in these LV polls. Certain theory holds that 80% of undecided voters will break for the challenger. For this analysis, I was more conservative and assume that 65% of undecideds will ultimately vote for Romney.
Based on these criteria and assumptions, I've arrived at a prediction of exactly 270 votes for Romney, just enough for election. Two states remain very tight and too close to call: Colorado and Ohio.
Click here for my custom RCP map.
It's worth noting that money of the "close states" could end up breaking for Romney, especially in the Midwest; a "Rust Belt sweep" that includes Michigan and Pennsylvania, as well as the aforementioned Ohio, is still quite possible at this point.
