The final voting for the India’s general election concludes at 6:30 PM local time on Monday, May 12 and, “as the voting closes in India, no one knows what the results will be (opens in new tab).” There are total of 814 million eligible voters (opens in new tab) in this massive display of democracy, which is why voting is spread out over more than a month (not enough polling machine and election workers, so they need to move around the country). It will take an additional four days to count the final vote tallies, due to be released on Friday, May 16, although we expect exit polls to be released as the voting closes.

While readers from the United States are accustomed to intense and accurate (opens in new tab) coverage of the election polls through Election Day (the US is already inundated with 2016 election polls (opens in new tab)), India bans polling from being released during the voting portion of the campaign. Microsoft Research saw this as an opportunity to test a new crowdsourcing prediction game (opens in new tab) to see how well we could do at predicting the unpredictable.

The Indian Elections Game (opens in new tab), joint work between Microsoft Research India (opens in new tab), including Shipra Agrawal (opens in new tab) and Nikhil R. Devanur (opens in new tab), and Microsoft Research New York City (opens in new tab), including Miro Dudik (opens in new tab) and myself (opens in new tab), was released just before voting began on April 7 and ran through today’s final voting day. Nearly 4,000 users created their predictions on the number of seats that any given party, or combination of parties, would get in any of the 35 voting states and union territories. For example, a user may look at Delhi’s seven seats and predict that BJP would get between 4 and 5 seats while Congress (INC) would get between 0 and 1 seats.

(opens in new tab)

As the crowd makes predictions on the outcome of all 35 voting states, the national parties, and national alliances, the back end of the game examines the crowd’s predictions and creates the probability of any given outcome occurring. In this example, since game’s aggregation of the crowd’s predictions currently concludes there was just less than 50% likelihood of this combination occurring, a prediction that wagered 1 point on BJP getting 4 or 5 seats and Congress getting 0 or 1 seat would yield 2.28 points if correct and 0 points if incorrect.

(opens in new tab)

Computing the probability was not trivial, as the game had to learn how the different parties interacted within a given district and between districts. And, unlike the United States where there are just two major parties, there is large number of viable parties, some national and some regional, competing for the 543 seats. The potential number of outcomes is huge! In future blog posts, over the next few days, we will go into more depth about how all of that was constructed.

But, right now, just as the election closes, and the exit polls are released, we wanted to share with you the final probable outcomes that our game generated from the approximately 4,000 users who played over the last month.

We expect the opposition party BJP to carry 228 of the 543 seats and their national alliance, the NDA, to carry 270 seats. The party currently in power, Congress is expected to carry just 87 seats and its alliance, the UPA just 110. The Third Front alliance is expected to carry 80 seats. That leaves 83 seats to smaller parties that are not formally aligned with any of the three major alliances. The national party predictions have a margin of error plus or minus 7% of seats and the national alliances plus or minus 10% of seats.

This prediction is generated by considering each voting states independently and then considering how they aggregate up together. Below we list each of the 35 states and our individual predictions:

(opens in new tab)

Please check back over the next few days we discuss this game and its predictions in more detail!

The post Microsoft Research’s Indian Election Game appeared first on Microsoft Research.

(opens in new tab)

The inaugural year of Microsoft Research New York City (opens in new tab) has been stupendous. All of us at Microsoft are thrilled with our newest lab.

The lab officially opened on May 3, 2012, with the announcement of a group of 15 founding researchers: David Pennock (opens in new tab), Sébastien Lahaie (opens in new tab), Justin Rao (opens in new tab), David Rothschild (opens in new tab), and Giro Cavallo (opens in new tab) in algorithmic, computational, and empirical economics; Duncan Watts (opens in new tab), Dan Goldstein (opens in new tab), Sharad Goel (opens in new tab), Sid Suri (opens in new tab), and Jake Hofman (opens in new tab) in computational and behavioral social science; John Langford (opens in new tab), Miro Dudik (opens in new tab), and Alekh Agarwal (opens in new tab) in machine learning; and Fernando Diaz (opens in new tab) and Elad Yom-Tov (opens in new tab) in information retrieval. Together, these researchers bring a deeply original and phenomenally productive approach to data science, particularly in the domains of economics and the social sciences.

In the fall, the group was joined by one more member, Jenn Wortman Vaughan (opens in new tab), who has done research in machine learning, algorithmic economics, and social science, and who, therefore, was a great match for the lab.

(opens in new tab)Our New York City researchers immediately dived in and engaged deeply with product groups, other Microsoft Research labs, and the local community. Rothschild spearheaded an effort with the Xbox LIVE (opens in new tab) team to put together a tremendously successful interactive channel for the U.S. election. Polls conducted on the channel during events such as the presidential and vice-presidential debates and election night received millions of responses—producing phenomenal data for academic studies and substantially increasing user engagement. On election night, Rothschild’s prediction engine, done in collaboration with Pennock and Dudik, correctly predicted all state outcomes except Florida. Other deep product-group engagements include Lahaie’s engagement with Microsoft’s Online Services Division, Langford’s engagement with the company’s Server and Tools Business, and many more.

Our researchers ran conferences to bring together both the academic and internal communities. Goel and Hofman organized the Computational and Online Social Science Workshop at Columbia. Langford was co-chair of the program committee of ICML, one of the major worldwide conferences in machine learning. He also organized an internal Microsoft machine-learning class attended by hundreds of people from Microsoft product groups.

The lab also has received many honors. Vaughan won the Presidential Early Career Award for Scientists and Engineers (opens in new tab) from President Obama. Goldstein was named a Fellow at both the Center for Measurable Marketing (opens in new tab) at NYU’s Stern School of Business and the Applied Statistics Center at Columbia University’s Institute for Social and Economic Research and Policy. Goldstein and Suri, along with Preston McAfee of Google, won the Best Paper Award during the 13th ACM Conference on Electronic Commerce.

Given their amazing start, I can’t wait to see what the New Year brings for our Microsoft Research New York City researchers!

The post Microsoft Research New York City: 2012 in Review appeared first on Microsoft Research.