All UK General Election Exit Polls from 1974-2017 - BBC

All exit polls from October 1974 to the present day.

[RBenjo]

You don't show the actual detail of the 2005 poll in this video, but it's worth noting that every single one the polls bar 1983 and 2010 overestimates Labour.

For those interested, the 2005 poll in full:

Share: LAB 37%, CON 32%, LD 22%, Oth 8%
Seats: LAB 356, CON 209, LD 53, Oth 28

[CaradhrasAiguo]

And I thought 1992 was a bad prediction both pre-election polling and exit poll. Look at Oct 1974 (61 seats too many for Labour) and 1987 (38 seats too little for Con)

[DFandV]

Wow what a majority for Labour in 1997. This will never happen unless May really screws up and people vote Labour but it's now split.

[TomBartram-bc]

If the whole of England had been a single constituency in 2005 Michael Howard would have won by 70k votes!

[gyrovague]

"We're able to predict what we think has happened tonight" Oh really!

[DaDez]

Who would pay to watch the party leader's reactions to the exit poll?

[sharmisarkar]

Wow, Labour nearly got half the vote in 1997. Talk about disestablishment...

[altelectionuk]

I don't know why pepole speculated so much in 1992, the exit poll in 1997, and 2005 was even less acurate!

[johnday]

The BBC seems to be showing an anti-Northern Irish bias by having the DUP lumped in with "other" while showing Plaid Cymru with 3 even though the DUP had more seats.

[martindavis]

If the BBC's exit poll in 1987 had been accurate and the Conservatives had won with such a small majority is it feasible the party would have lost that majority by 1992 with Kinnock then going on to win?

[adamfowle]

Who came up with the exit poll in 1987? Karl Marx?

[neventomicic]

It is useless to talk about prediction while you have First Past the Post system. It only matters in the proportional system, or only 2 candidates.

[SiVlog]

Just goes to show that the exit poll is not always accurate

[freakylocz]

1997 was also the beginning of 18 good years for the Lib Dems.