But there is one exception and that is the Delhi-based “Today’s Chanakya”, a family-run group that has conducted interview-based poll survey and came accurately identical to the final results.
Its CEO VK Bajaj attributes the accuracy to the quality of their sample than the size of the sample. “In India, when any poll is released, everyone talks about the sample” numbers and not about the quality of the sample, he told Reuters. “Every state has some caste combinations prevailing … we tried to make the sample as representative as we could.”
With 815 million voters, the mega-election was spread over several weeks in the country, and the exit polls were released as per the Supreme Court verdict on the last day of voting, May 12, and the results were declared on May 16.
Today’s Chanakya has predicted that the BJP-led coalition would win 340 of the 543 seats, with an error margin of 14 seats less or more. It said BJP on its own would get 291 seats.
A week later, the results showed the NDA winning 336 of the 543 seats, just 4 seats less than the main prediction and the BJP won 282 seats, just 9 short of the prediction. Taking the margin into account, the survey was almost as accurate as was the final outcome, making the statistical survey reflecting the reality.
In fact, all other post-poll analyses forecast less than 300 seats for the NDA.
Today’s Chanakya, the two-decades-old research firm interviewed 38,984 voters and employed about 250 field representatives and 175 head office team to conduct the survey and prepare the outcome.
The next best was by Hansa Research for the NDTV news channel that predicted BJP and its allies to win 275 seats.