General Election PoliForecast 2019
So this is ambitious!
This is a forecast of how everyone is going to vote on December 12th. Including Northern Ireland, including independents and taking into account the standing down of MPs etc, “Remain Alliance”. If there is no Green standing, there shouldn't be a Green on the forecast. So what we *should* have here is a forecast of how many people are going to vote for each party in all 650 constituencies. Not just some wishy washy vague percentage – but the actual number of people.
This is based on Conservatives on 42%, Labour on 30%, Lib Dems on 16% and BXP I have on 3%. Clearly as the polls widen or narrow in the coming weeks this will change.
Clearly it can be used to drive tactical voting - currently still predicting BXP with zero seats, but with a significant impact on the election result nonetheless.
On election night, when Houghton and Sunderland South is declared at about 11.15 you can actually see how the result compares to the expected result.
Clearly, I don’t know what 40 million people are going to do on December 12th - I'm not even 100% sure what I'm going to do. So it's going to be wrong. It’s just a model. It’s based on a lot of assumptions. Susie Quatermass is standing for the Motherworld Party in Wells. I don’t know what her policies are, but according to the assumptions in the model she isn’t going to get all that many votes. The assumptions could be wrong. I’ve also made some, shall we say, “anecdotal adjustments” – I’m not proud of myself, I felt a little dirty afterwards, but I was just trying to take into account known factors and giving them a value.
Some of it doesn’t have much to go on. Buckingham will now be a competitive seat for the first time in quite a while – so data for when the seat was competitive is out of date. Chorley, conversely, is no longer a competitive seat – but needless to say the Speaker will get a sizeable majority. Most of this doesn’t matter a great deal – I may have misjudged the popularity of Motherworld by a factor of 5 – Wells should still be a Tory seat on December 13th.
Where it may make a difference is where the independent candidate is, say, Dominic Grieve. So if you see a large vote share in the “Others” category, it may be a mistake, but it may also be an “anecdotal adjustment”.
Look out for constituencies you know and look for any anomalies. I shall improve the data as we get closer to December 12th, and also work on the presentation, but it should already give us some idea of what to expect. Any glaring errors, please let me know!