Here are 8 factors to watch out for, says psephologist Yogendra Yadav.
The final round of pre-election surveys, including the much-awaited pre-poll round of National Election Study by Lokniti-CSDS team, has confirmed what was widely felt by Indian political observers: At the starting point of the race, it's advantage NDA.
Various surveys have made different seat projections.
The Lokniti team projects 222 to 232 seats for the BJP (263 to 283 for the NDA) and 74 to 84 seats for the Congress (115 to 135, including existing Congress allies).
The Times Now-VMR poll projects 279 seats for the NDA and 149 for the UPA.
The real question is: Are these numbers for the BJP likely to travel northwards or southwards? A reduction of 20 to 30 seats for the BJP compared to these projections could open up the game, while the addition of the same number could mean no challenge to Mr Modi.
Here are 8 factors to watch out for.
I will use the NES data because it has placed detailed tables, the entire methodology and question-wording in the public domain.
Will the Balakot-effect cool off?
The NES data shows while Pulwama-Balakot or national security is not by itself an election issue, it may have helped shore up the image of PM Modi and his government in a big way.
As many as 79% had heard about the Balakot air strikes and nearly half the respondents give the Modi government some credit for it.
Importantly, those who had heard were much more likely to prefer Modi for PM and give his government another chance.
Will this effect persist? As many as 61% agreed that the BJP was trying to make electoral gains from the air strikes, although half of them were all for giving this government another chance.
Will PM Modi's repeated attempt at invoking Balakot be counter-productive?
Will NYAY hot up as an electoral issue?
Although the formal announcement of the Congress's minimum income guarantee scheme took place in the middle of this survey, as many as 48% had heard about it.
This number is bound to go up after the campaign.
There are two problems for the Congress here.
One, the poorest who might benefit from it know less about it than the better off.
Two, awareness of NYAY leads to a small gain for the Congress (reduces the popularity gap between Narendra Damodardas Modi and Rahul Gandhi by 9%), nothing compared to the way Balakot worked for the BJP.
Will the young vote favour the BJP despite anxiety on unemployment?
The survey throws two contrary findings.
When people were asked to name an issue that will matter most to their vote, unemployment came up at the top.
Young and educated were most likely to hold this view.
Yet the BJP seems to be getting higher than average support among young voters.
Clearly, many young voters do not blame the government for joblessness.
Will this remain so, if the Opposition runs an aggressive campaign targeting Modi regime for its record on employment?
Will state level pro-incumbency counter central incumbency?
The NES shows that more voters are going to think of the central government's performance while voting in the Lok Sabha election compared to the previous election.
In six Opposition-ruled states (West Bengal, Odisha, Karnataka, Chhattisgarh, Madhya Pradesh and Rajasthan) there could be a clash: The Modi government, as well as the non-BJP state governments, enjoy pro-incumbency in these places and the voters put some weight on both these.
Which of these two considerations will trump?
Can the BJP neutralise strong anti-incumbency sentiments against its candidates?
In five states (Uttar Pradesh, Bihar, Madhya Pradesh, Maharashtra and Delhi), which accounted for a majority of the BJP's MPs in the dissolved House, there is a strong anti-incumbency sentiment.
Will the voters's anger against them affect the ruling party?
How much of a presidential battle will it become?
The BJP has tried very hard to make this race into a Modi vs Rahul contest.
Among those who vote on the basis of candidates, the NDA has a mere 2 percentage point lead.
Among those who vote on the basis of the party, the UPA has a 3 percentage point lead.
But among the one-fifth voters who vote on the basis of their preferred PM, the NDA has a 51 percentage point lead! It seems Balakot may have influenced this election by increasing the salience of the PM choice.
How much will the poor coalition damage the opposition?
So far, the mahagathbandhan has proved to be a string of loosely coordinated, imperfect local alliances.
The Congress is unlikely to win many seats in Uttar Pradesh, but its votes could hurt the SP-BSP tally seriously.
The same is true of the Congress in Odisha, West Bengal or Delhi and the BSP in Madhya Pradesh, Rajasthan and Chhattisgarh.
Will the non-BJP parties work out some last-minute arrangement to avoid division of votes?
Will there be a lower turnout among anti-Modi voters?
The most significant finding of this round of the NES study is the phenomenon of 'active pro-incumbency' vs 'passive anti-incumbency'.
Simply put, while Modi backers are enthusiastic about voting, those who are unhappy are less likely to turn out to vote.
The proportion of reluctant voters is higher among Muslims.
About one-tenth of the respondents said they were unlikely to vote.
Of them, there are more UPA than NDA voters.
If all of them do not come out to vote on D-day, this will boost the NDA's vote lead over the UPA by as much as 3 percentage points.
This could tilt the scales for the NDA in more than 30 seats.
So, watch out for the turnout data.
By special arrangement with The Print
Yogendra Yadav is the national president of Swaraj India.