UK General Election 2019 - Election Day and Results Thread

[AMOLAK MANN]

The initial BBC projections haven't always been right on, but their track record is pretty good overall:

2017 - projected Tories short by 12 (actually short by 8 )
2015 - projected Tories short by 10 (actually a majority of 12)
2010 - projected Tories short by 19 (correct)
2005 - projected a Labour majority of 66 (correct)
2001 - projected a Labour majority of c.160 (actually a majority of 167)
1997 - projected a Labour majority of c.175 (actually a majority of 179)
1992 - projected Tories short by 25 (actually a majority of 21)
1987 - projected a Tory majority of 26 (actually a majority of 102)
1983 - projected a Tory majority of 146 (actually a majority of 144)

(They didn't provide an exact figure in 1997 & 2001, giving just a rough estimate.)

They were rather off in 1987 & 1992, but otherwise were pretty close on all the rest (though you can debate if their 2015 exit poll was 'close' or not, I suppose).

1997 the BBC didn't broadcast the forecast but it was for a 209 LAbour majority and in 2001 they did broadcast seat numbers and the precise majority forecast was 157.I can give details of ITN if anyone likes and details for both channels prior to 1983

[AMOLAK MANN]

I'm not sure also if the overestimate was again the shy Tory phenomenon-the unweighted figures and LAb 49%, Con 27%.

BBC:
October 1974-projected Labour majority of 125(actually a majority of 3)
1979-projected(implicititly) a Tory majority of 19 (actually a majority of 43).

ITN-same as BBC from 2005 but before that:

2001 - projected a Labour majority of 175 (actually a majority of 167)
1997 - projected a Labour majority of 159 (actually a majority of 179)
1992 - projected Tories short by 21 (actually a majority of 21)
1987 - projected a Tory majority of 68 (actually a majority of 102)
1983 - projected a Tory majority of 116 (actually a majority of 144)
1979- projected a Tory majority of 63  (actually a majority of 43).

Another exit poll for The Sun newspaper in 1992 projected Labour as the largest party short by 19(with Tories short by 24)
October 1974-projected Labour majority of 15(actually a majority of 3)

[AMOLAK MANN]

It is maybe understandable that some Tory voters were a bit "shy" in 1997......

It really hasn't been a thing in more recent GEs though (2015 was for other reasons)

Given that the exit poll supposedly focuses on marginals, it's not too surprising that the popular vote figures were off - the Labour swing was stronger in their target seats than in the country as a whole, and the effectiveness & efficiency of anti-Tory tactical voting made them make even more gains.

A lead of 47% to 29% would mean a 13% swing to Labour; the swing in Tory-held marginals was 12%, and the 'effective' swing (that is, one nominally necessary to make 146 gains) was 12.4%. Given that the rough number at which Peter Snow's arrow initially pointed when he was explaining the exit poll figures was a Labour majority of around 170-180 (he didn't actually specify a number until some results had actually come in, and then forecast a majority of 171), which was pretty consistent with the final result, I think that this may be why the popular vote numbers exaggerated the Labour lead.

This was true of the ITN poll which accurately forecast the marginal would swing more and so it's national vote share projection from the changes in the marginal exaggerated the LAbour lead-46 to 30% but the BBC poll was national so the reason for the exaggeration may still be a shy Tory effect as in 1992