Six states to watch

While this election looks predictable in large swathes cent of the vote in 2019 in some states

How well-contested is this election going to be? If you are a Narendra Modi or Bharatiya Janata Party (BJP) supporter, you would say this is the least unpredictable election ever and 400 should be expected for the National Democratic Alliance (NDA). If you are an Opposition backer, you might say this election will turn out like 2004, where Vajpayee's NDA, the favourite, lost.

There's a curious buildup as the votes in the first phase are cast. There are so many "opinion polls" on TV channels whose methodology you know nothing about-predicting the same thing: A BJP tally bigger than in 2019. The most successful pollster lately, Pradeep Gupta (Axis My India) has not spoken yet. Or maybe he has. There was that tweet, apparently from him telling Moneycontrol, that the NDA was having a tough time in 13 states.

The tweet was deleted shortly afterwards. He did not claim that he had conducted an opinion poll.

When a champion pollster hedges his bets, the best a journalist can do is to merely "read" an election. Could it be therefore, that both sides are somewhat right on this and Mr Gupta, too, in what he said in that deleted tweet?

An Indian election, even with Mr Modi at his peak, is a net outcome of contests in different states. In the 25 years of coalitions (1989-2014) I had argued that an Indian general election is like a best-of-nine-sets tennis match. The side that wins five will grab this electoral grand slam.

The nine were: Uttar Pradesh, Bihar, Maharashtra, Andhra Pradesh (before division), Madhya Pradesh, Tamil Nadu, Rajasthan, Karnataka and Kerala. These added up to 351 seats. The side winning five would get close to 200, which for coalition-making was effectively 272. That ended in 2014.

For this election, we will revise that theory and look at six: Maharashtra, West Bengal, Bihar, Karnataka, Jharkhand and Odisha.