Year 2016: Not Your Grandpa’s Election
Sweet Sixteen–it has been 16 years since we started this project, engaging students in the political process – to be educated and to be involved. Current participants in the project include both undergraduate and graduate students. Scroll to the bottom of the page to see how this project began in 2000.
This is an election that defies all conventional wisdom. It will be the subject of studies in political science, sociology, history, statistics, polling science and psychology for decades or even centuries; in this country and perhaps all over the world. As a country, I hope that we will collectively reflect to move forward in a positive way. We shall begin with a victory concession speech by Mr. Trump, the Republican Nominee, to begin the reconstruction process to build a better future for America and for the world.
In 2016’s edition of Stanford Predicts, the undergraduate students are first-year Stanford students who took part in SSEA: Stanford Summer Engineering Academy. Matthew Chen was a SSEA alumnus and he participated in the project in the 2012 election cycle.
As they engage in a current and exciting event, students within the project also learn model building: how to build a bigger model from smaller components and how to make reasonable assumptions about that model.
For a list of participants this year, please visit the team section of this website.
For more information about Stanford Predicts, including inquiries about our methodology or other press inquiries, please contact Matthew Chen.
Year 2012: Reborn, with a New Website
Participants in the 2012 edition of the project included both undergraduate and graduate students for the first time in its history. You can view its original website at this link: Stanford Predicts 2012.
In 2012′s edition of Stanford Predicts, the undergraduate students were first year Stanford students who took part in SSEA: Stanford Summer Engineering Academy. Graduate students were from MS&E 220, the course in Probabilistic Analysis taught by Professor Samuel Chiu.
Year 2008: Sleeping with Eyes Wide Open
The ‘Stanford Predicts’ website went dormant for the 2008 election cycle, but the tracking aspect of the project did not, as Professor Chiu gave students in MS&E 220 periodic updates as election day neared. On November 3rd, 2008, one day before the election, the prediction engine calculated the most likely outcome of the election as an Obama victory—more specifically, as an Obama victory with him receiving 364 electoral college votes. The next day, Obama won 365 electoral college votes—the one vote difference coming from Nebraska, which split its electoral college votes by giving 4 to McCain and 1 to Obama. In 2008, the prediction engine treated every state as giving either all or none of its votes to one candidate, but that was an estimate which foiled the project, as both Nebraska and Maine split their electoral votes among candidates depending on the number of votes they receive. Having been so very close to correctly predicting the result of the election, the prediction engine is being modified in 2012 to account for Nebraska and Maine’s ability to split their electoral college votes.
Year 2004: Growing up, with Growing Pains
In 2002, Professor Chiu still taught MS&E 220, but his previous students had long since graduated. With a new set of students, the project attempted to track and predict the results of mid-term congressional elections, and it achieved moderate success. Leading up to the 2004 election cycle, a website was created with the aid of an MS&E student, Can Sar, and the project went public under the name “Stanford Predicts”. That website still exists and can be accessed at this link: Stanford Predicts 2004.
As the days and months passed leading up to the 2004 presidential election, there were ups and downs within the project, both poll wise and work wise, but the project never ceased to be fun. Under the drive and dedication of two other MS&E students, Rajiv Dulepet and Somik Raha, ‘Stanford Predicts’ took flight and began tracking the state-by-state polls and attempting to predict the presidential election. The tracking was made easier by an innovative computer program which automatically updated poll results as they appeared in the news media. Despite the continuous updates, the prediction was fairly uneventful until October 31, the Sunday forty-eight hours before election day.
On that day, the members of the project were uncharacteristically unquestioning when they factored in a new poll into their calculations: the Harris Internet Poll. This poll, which polled very large samples in the critical swing states of Ohio, Florida, and Pennsylvania, proved to be fatal for the project’s goal of correctly predicting the presidential election. The project expected Kerry to win the election, but instead Bush won reelection as U.S. president. After having learned a deadly lesson about using internet polls, the members of the project performed a postmortem analysis in the document Wednesday Morning Quarterback. The website went dormant, and the members of the project finished their time at Stanford. Can Sar started a company called Apture, Rajiv Dulepet started a company called WiseWindow, and Somik Raha became an associate at SmartOrg.
Year 2000: The Beginning
This presidential election prediction endeavor started out over a decade ago as a project within a probability class. The class, MS&E 220: Introductory Probabilistic Analysis, was and still is taught by Professor Samuel S. Chiu in the Management Science and Engineering Department at Stanford University, and the project had three main objectives.
Firstly, the project uses state-by-state polling data to calculate a win/loss probability for each state for the two primary presidential candidates, who, in 2000, were George W. Bush and Al Gore. Secondly, the project combines these win/loss probabilities into a distribution corresponding to the possible totals of electoral college votes that each candidate could win. Thirdly, it uses this distribution to calculate the probability of a candidate winning the entire election. Because there are a total of 538 electoral votes, the probability of a candidate becoming president is essentially equal to the probability of that candidate winning 270 or more electoral votes.
Within the class, students tracked the progress of the election right up until November 7th, the day votes were cast. The students and teacher learned a significant amount about the art and science of predicting an election. After George W. Bush was declared to have won the election, the project stayed dormant until the next presidential election, four years later.