FiveThirtyEight's in-house election forecasting model shows Brown as a 74 percent favorite to beat Martha Coakley, with all the latest polls from ARG, Research 2000 and InsiderAdvantage taken into account.
As other analysts and oddsmakers have warned, there is considerably more room for error in a special election, especially one as dynamic as this one. And a lot of polls were wrong about NY-23. But FiveThirtyEight's model, as Nate Silver is quick to note, correctly predicted the outcomes of all 35 Senate races in 2008.
Silver says:
Coakley's odds are substantially worse than they appeared to be 24 hours ago, when there were fewer credible polls to evaluate and there appeared to be some chance that her numbers were bottoming out and perhaps reversing. However, the ARG and Research 2000 polls both show clear and recent trends against her. Indeed the model, which was optimized for regular rather than special elections, may be too slow to incorporate new information and may understate the magnitude of the trend toward Brown.
Interestingly, Silver's model compensates for the tendency of polling data to underestimate dominant-party turnout (in this case, Democratic turnout) in the face of rigorous challenges. The observed difference is a net of +2.3 points for the dominant-party candidate, and given that "bonus" Coakley's chance of victory would be up at about 40 percent. With no correction at all, the model gives Brown a prohibitive 85 percent chance of winning. The 74 percent figure comes from Silver essentially splitting the difference between underestimating Democratic turnout and underestimating the wind at Brown's back.
So just to be clear, the most skeptical interpretation of the most accurate Senate election forecasting model out there makes Republican Scott Brown a 60/40 favorite to replace Ted Kennedy in the United States Senate. Just thought that needed to be said aloud.
corner.nationalreview.com |
