The Tories could still be on track for a 100-seat landslide despite the shock exit poll.

Initial results appear to be less good for Labour than expected, with the party winning fewer seats from the Conservatives than the exit poll had predicted.

The exit poll – which is very rarely wrong – suggested that the Conservatives would be the biggest party but that they would fail to score a working majority in the House of Commons. That would result in a hung parliament, with any government having to rule as a minority or in a coalition.

But pollster Peter Kellner, the head of YouGov, said that initial results from the two seats that have declared – Sunderland and Newcastle – were far less good for Labour than had been expected.

If the exit poll was that wrong across the country and in the same way, then the Tories would be on track for a 100-seat landlord, he suggested.

In Sunderland, for instance, the exit poll had suggested that there would be a 7 per cent swing towards Labour. But that ended up to be only a 2 per cent swing – throwing the entire exit poll into doubt.

If the Tories did manage to gain such a huge landslide, it would be far more in line with polls ahead of the election.

The results may be wrong because the exit poll doesn't collect information for people who voted by postal vote, Mr Kellner said. Those people are expected to vote disproportionately for the Tories – potentially explaining why the poll's data appeared to suggest better results for Labour than in the general election.