1/ I've been following the debate about #TacticalVoting in UK #GeneralElection2019, and particularly the controversy about @BestForBritain's site getvoting.org, which is accused of giving "bogus" advice based on a fairly opaque methodology theguardian.com/politics/2019/…
2/ So... I decided to get the constituency-by-constituency data from this site and a couple of others (tactical.vote & tacticalvote.co.uk) to see how they compare, and whether they pass a basic "plausibility" test based on % voting swing needed from 2017 results
3/ I narrowed down my analysis to the 317 seats won by @Conservatives at the 2017 General election, since these are the core focus for tactical voting, whether to get the #ToriesOut or #StopBrexit. This chart shows which party came second in each of these constituencies in 2017 
4/ I narrowed again to look at % swing needed for @LibDems or @UKLabour to win in each constituency. The baseline tactical strategy would be to vote for whichever party came closest to beating the #Tories in 2017, and this is almost exactly what tactical.vote advises
5/ The same plot for getvoting.org shows something very different. Even on the assumption of a huge swing from @UKLabour to @LibDems on 12 December, there's no plausible "tactical" case for voting for Lib Dems in clear Tory/Labour marginals like #Watford or #Hendon
6/ Meanwhile, tacticalvote.co.uk takes a more cautious approach by reserving judgement on many constituencies until later in the election campaign. As the 2017 General Election and 2019 #EU Election showed, there can be a huge shift in public opinion during this short period
7/ "Tactical voting" is worthless if voters who want the same thing follow contradictory advice. Whatever the "science" behind getvoting.org, its advice contradicts the others and doesn't pass a basic plausibility test, so I would suggest to view it with extreme caution
8/ I compiled all of the data for this analysis in a spreadsheet (incl. 2017 voting data from @BritainElects). There is a link to the data at the bottom of my article on Medium (you're also welcome to download and reuse the charts) medium.com/@bothness/an-a…
@britainelects So... for anyone interested in learning more about @BestForBritain's methodology, they shared this thread with me from the @FT's @jburnmurdoch... Interesting stuff, but will they update their advice as polling inevitably shifts between now and 12 December?
9/ Ps. This whole contradictory "tactical voting" advice fiasco makes a massive case for #ProportionalRepresentation in UK general elections. Some people are already calling it #TacticalVoteGate... And there's more to come next week
10/ Pps. @chrishanretty has done a wonderfully geeky job of extracting @BestForBritain's #MRP voting projections for anyone who wants to look at their numbers. Not the same as looking inside their model (survey Qs, weighting etc), but better than nothing!
@chrishanretty @BestForBritain 11/ UPDATE! @TacticalVote just added voting advice for a bunch more constituencies. Here's how it affects their Labour vs Lib Dem horse race. I added a dotted line for a 12% swing based on current @BritainElects & @guardian polling averages
