University students use technology to predict election results

By Amanda Liberatore

Fifty years ago, people would have never expected that a computer could predict the results of a presidential election.

However, students at the University are playing a part in the process of educating and informing people of election prediction.

“Roll…to the White House: A Prediction Model for the 2008 United States Presidential Election” is a student and faculty-run Web site focusing on using models in order to predict the outcome of November’s election.

The Web site was created by computer science and political science students at the University, including Sheldon Jacobson, a computer science professor at the University.

The site is an advanced computer model that predicts the outcome of the 2008 Election.

Get The Daily Illini in your inbox!

  • Catch the latest on University of Illinois news, sports, and more. Delivered every weekday.
  • Stay up to date on all things Illini sports. Delivered every Monday.
Thank you for subscribing!

William Kormos, senior in Engineering, is the site’s webmaster and said that he believes predicting the results is beneficial.

“I think it is beneficial on some level,” Kormos said. “Campaigns need to make informed decisions on where they should campaign heavily and spend their ad money. Everything is a time crunch, and there is a (large) budget.”

Kormos, along with various other political science and computer science students, have been working alongside Jacobson in order to interest students.

The site has been getting plenty of hits since it was put up.

“As of today, we have had over 40,000 views since the end of August,” Jacobson said.

Jacobson said the model is employed by Bayesian estimators who use available state poll results to determine the probability of each candidate winning each of the states.

These calculations are used to determine a probability distribution for the number of Electoral College votes that each candidate will win in the 2008 presidential election.

“The results of the prediction models inform voters of the political race and who is winning so they can get a better idea of what will most likely happen in the election,” Jacobson said.

As technology advances, the more advanced the process of prediction may get.

“Forecasting elections isn’t a new thing, and people have been working on statistical models that are solved by computer for decades,” said Brian Gaines, professor of Political Science. “However, the models can be more and more complicated as powerful computers get more affordable.”

While some people may think that these models are a good idea, other voters feel that it is not necessary.

“No, I think it’s entirely detrimental to the process,” said Zack Sasnow, sophomore in Engineering.

“Instead of allowing voters to weigh the candidates based on their platforms, it causes them to feel inclined to go along with the flow of the model.”

Either way, Kormos said technology is key in this year’s election.

“The better someone can model the election, the better the information that is available to those who need to make those decisions,” he said.