ICM

[jamie]

Well, they’re not herding at least.

[The Bishop]

It would be one of the ironies of this election if ICM retrieved their once high polling reputation by being the "outliers" who were actually correct.

[Old Fashioned Leftie]

Dec 9, 2019 15:10:34 GMT Andrew_S said:

Equivalent ICM poll from 2017

Con 45

Lab 34

LD 8

[andrewp]

Dec 9, 2019 20:08:51 GMT Old Fashioned Leftie said:

Dec 9, 2019 15:10:34 GMT Andrew_S said:

Equivalent ICM poll from 2017

Con 45

Lab 34

LD 8

It all depends on whether pollsters have corrected their methodology since 2017. They are confident they have, we’ll find out on Friday.

[justin124]

Dec 9, 2019 20:10:28 GMT andrewp said:

Dec 9, 2019 20:08:51 GMT Old Fashioned Leftie said:

Equivalent ICM poll from 2017

Con 45

Lab 34

LD 8

It all depends on whether pollsters have corrected their methodology since 2017. They are confident they have, we’ll find out on Friday.

ICM certainly did change its methodology post 2017 having accepted that it had overcompensated for the 2015 polling debacle. I believe that is also true of Comres. For that reason, it is false comfort to use the 2017 data on a like for like basis.

[Forfarshire Conservative]

Thread. This is halfway down it as it provides an overview.

[Robert Waller]

Does anyone know if we're expecting a 'final' ICM poll today?

[curiousliberal]

Dec 12, 2019 10:08:08 GMT Robert Waller said:

Does anyone know if we're expecting a 'final' ICM poll today?

If we have no more, an average of the last 6 prior to close of polls projects a 10.5% lead for the Conservatives, exactly 2 points higher than the one they had (according to pollsters who've since changed their methodology) in 2017.