Analysing Exit Poll Outcomes and Their Accuracy Across Indian State Assembly Elections Through Descriptive Statistics

The period between polling and the final count of votes is one of the most suspenseful phenomena. Parties eagerly await results after months of campaigning and canvassing for votes. During this period, one often comes across news publications predicting the final results and seat tallies of the election through various surveys conducted by their survey teams on a sample of the electoral population. These surveys are published as exit polls, aiming to give the public and parties an idea of what the end outcomes of elections would look like. Historically, there are multiple examples of the exit polls getting it wrong. Who could forget the great headline of ‘Dewey defeats Truman’ published in the Chicago Daily Tribune in 1948? Even in the Indian general elections of 2004 and 2014, exit poll predictions failed to gauge the direction of public support. This research study investigates a specific period of exit polls and uses descriptive statistics to decode the accuracy of these polls in predicting the final outcomes of elections and the seat projections. Using the elections from the beginning of 2020 till the end of 2022, this study analyses all the State Assembly elections held in this time frame, by comparing the exit polls forecast by four different agencies (Poll of Polls, C-Voter, AXIS and Politique Marquer) in each of those elections and analysing them with the final results of the election by testing the accuracy of these polls through the means of descriptive statistics.