BANGALORE: While many experts are surprised at NDA’s thumping victory in Indian Elections 2014, analysis of pre-poll online conversations had predicted a massive victory for the Narendra Modi-led opposition.
As we posted on our blog yesterday (read full post here), NDA was expected to win 310-342 seats.
This was more accurate than most exit polls in the media, which gave NDA around 276 -290 seats.
Apart from the national scenario, we were fairly accurate about many state/city results as well.
For example, we predicted that BJP will win 6 seats in Delhi. It ended up winning all the 7 Lok Sabha seats.
Similarly, the BJP-Shiv Sena combine was expected to win 5 of the 6 seats in Mumbai. The final tally was 6.
In Bangalore, our prediction of 3 out of 4 seats for the BJP was spot on, as was our forecast of just 1 seat for the Left Front in West Bengal.
Among regional parties, our analysis showed DMK and BSP winning around 1 seat each in Tamil Nadu and Uttar Pradesh respectively. They did even worse than that by not opening their accounts at all.
In Punjab, Twitter data showed AAP winning around 3 seats. They ended up by winning 4.
Similarly in Bihar, BJP + LJP’s tally of 29 was close to our prediction of 27.
Areas where Twitter got it wrong
It’s not as if social media conversations were a perfect reflection of final results.
For example, online data predicted a massive wave for AAP, giving it more than 80 seats. Similarly for Congress, the figure was expected to be 98-127.
For AAP, it will be interesting to see their vote share once the counting is over. The party, being nascent and not consolidated by regions, might not be able to convert votes share to seat. The social media mandate and the vote share may have paved a great future path for Kejriwal and team.
Generally speaking, either Twitter data was spot on about regional parties, or way off the mark.