Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search
 

Scuba

(53,475 posts)
Tue May 6, 2014, 01:21 PM May 2014

New Election Model Machinery: A Look Under the Hood

http://www.nytimes.com/2014/05/06/upshot/new-election-model-machinery-a-look-under-the-hood.html?_r=0

John Sides, Ben Highton and Eric McGhee of The Monkey Cage have started Election Lab, a Senate forecasting model at The Washington Post. Their model gives Republicans more than an 80 percent chance of retaking the Senate; that’s much more than the 54 percent chance that Leo, The Upshot’s Senate model, gives the G.O.P. It’s also better odds than many Republican strategists give their party. In a recent survey by Reid J. Epstein of The Wall Street Journal, Kevin Madden, a former aide to Mitt Romney, gave the party a 65 percent chance of taking the Senate. John Brabender, Rick Santorum’s former campaign manager, put the odds at 75 percent.

As we’ve written before, we plan to tell you not only about the assumptions behind The Upshot’s Senate model but also to explain why and how it differs from others. So what explains the difference between Election Lab and Leo? Some of the biggest discrepancies are in less obviously competitive races, where The Post tends to give Republicans a far greater chance of victory. The Post gives the G.O.P. a 12 to 24 percent chance of defeating Democratic incumbents in Delaware, Massachusetts, New Mexico, New Jersey, Oregon and Minnesota. Leo, on the other hand, gives Republicans a one to six percent chance of winning in these states.

...

The Post is also more optimistic about the G.O.P.'s chances in New Hampshire, where the former Massachusetts senator Scott Brown is an impressive challenger on paper to Jeanne Shaheen, the incumbent senator, but trails by a significant margin in public opinion polls. Similarly, The Post is more optimistic about Republicans’ odds in Iowa. In a message on Twitter, the Monkey Cage modelers said that they had not yet incorporated fund-raising data there, but that when they add such data, the Democrats’ chances will appear to be stronger.

Leo and The Post also take different approaches in dealing with races in which the primaries have yet to take place. Instead of incorporating information about the candidates who are running and estimating each candidate’s odds of winning the nomination, the Post model assumes that a candidate of “typical” quality will emerge.



Monkey cage? Maybe it's to honor the chimpster.
Latest Discussions»General Discussion»New Election Model Machin...