“Totally wrong”: Newspaper columnist criticised for “misleading” article about tactical voting websites

A Guardian newspaper columnist has been heavily criticised by polling experts for writing a “misleading” article about tactical voting.

In her piece for the newspaper earlier today, Ellie Mae O’Hagan tells her readers to “beware of dodgy [anti-Brexit] tactical voting sites”, which she says are being used by “centrists” to manipulate Labour voters into plumping for the Lib Dems.


Her critique in centred on the MRP predictive method that has been used in Best for Britain’s country-wide tactical voting website.

O’Hagan suggests that MRP is being falsely used to produce constituency-level predictions, when actually it’s designed for national results. She says:

“MRP is usually used to predict national results, but in the case of GetVoting.org, it is being used to produce constituency projections…

“But this is misleading: predicting constituency-level results is a different ball game to predicting national ones.”

However, according to several leading pollsters, the complete opposite is true.

Self-professed number cruncher Matt Singh says that O’Hagan’s theory “is completely, totally wrong,” adding that “small area estimates are *precisely* what MRP is usually used for in this context.”

He has been backed up by Philip Cowley, Queen Mary University of London, who said that the piece is “just plain wrong” and that O’Hagan has “completely misunderstood” the MRP voting method.

Ben W. Ansell, Professor at Oxford University, also pointed out it’s ironic that O’Hagan accuses the tactical voting apps of misleading people, while “doing just that.”

There will be too much fake news pushed out by right-wing politicians and journalists during this election. We don’t need it from the left as well.

2 responses to ““Totally wrong”: Newspaper columnist criticised for “misleading” article about tactical voting websites

  1. There will be an error margin (e.g. +/-2%) associated with the models used for voter profiles in MRP analysis combined with a confidence probability of falling within that margin (e.g. 95%) expressed as “credibility interval” when applying down from national polling to constituency level. With 650 constituencies there will likely be some that are wrongly called – that’s unavoidable probability.

  2. Clearly the election is corrupt.
    People should not vote.
    Tactical voting and election pacts are bullying coercion and should not be tolerated. Communicate policies and let people decide and detach. Dont try and manipulate the result for such is a foul crime.

    So in a nutshell POLITICS is not the answer because it is normally devoid of SPIRITUALITY.

Leave a Reply

Your email address will not be published. Required fields are marked *