Wang misses mark on Senate predictions, but not by much

After Republicans won the Senate majority in Tuesday’s midterm elections, associate molecular biology professor Samuel Wang’s prediction that Democrats would maintain control has been proven false.

Control of the Senate hinged on the results of races for 36 seats,15 of which belonged to Republicans and 21 to Democrats. A Republican Senate majority would have required that Republicans win six of the previously Democratic seats.

AfterTuesday’selections, seven Senate seats in Arkansas, Colorado, Iowa, Montana, North Carolina, South Dakota and West Virginiachanged parties.Republicans now hold 52 seats, while Democrats hold 45.

In an essay published on his blog, thePrinceton Election Consortium, onWednesday, Wang noted the closeness of his polls-only snapshot to the elections’ final outcome.

“Midterm polling is less accurate than presidential-year polling, probably because turnout is hard to predict,” Wang said in an interview with The Daily Princetonian. “So, in fact, a one- or two-seat erroris to be expected. We’re waiting for results from Alaska and Louisiana.”

Wang explained that long-range political forecasting is challenging and the races were so close that the statistical uncertainty of one or two seats in his model made his majority prediction incorrect.

“Based on polls, the likely Senate outcome was close to the magic threshold between 50 and 51 Republican seats, and an exceptionally large number of races was uncertain,” Wang said. “For these reasons, I gave a probability that was in an uncertain range.”

He said he is working on developing his model so that it will be more accurate in the future.

“The right thing to ask now is: What could have been improved? That is the way to get better,” Wang said.

In October, a rivalry between Wang and well-known election forecaster Nate Silver developed in a series of exchanges over Twitter and their respective websites. Silver was one of many forecasters who predicted a Republican majority and attacked Wang's methodology on his website.

“I don’t like to call out other forecasters by name unless I have something positive to say about them — and we think most of the other models out there are pretty great,” Silver wrote on his website. “But one is in so much perceived disagreement with FiveThirtyEight’s that it requires some attention. That’s the model put together by Sam Wang.”

Wang’s model, which relies on aggregating data from polls as a direct measure of public opinion, predicted midterm election outcomes with Democrats in control of 50 Senate seats.

Get the best of ‘the Prince’ delivered straight to your inbox. Subscribe now »

His forecast, a September version of which was published by The New Yorker, showed in October only a 42 percent chance that Republicans would win the majority.FiveThirtyEight, however, said Republicans had a 76.2 percent chance of winning the majority in the Senate.

In the same week, The New York Times Upshot gave Republicans a 60 percent chance and The Washington Post Election Lab gave the Republicans a 78 percent chance. Most other statisticians predicted a Republican majority as more than likely.

Wang’s prediction was controversial as one of the few that did not forecast a Republican majority.