The different tactical voting sites all use different methodologies. BfB uses MRP, which is analysis based on a large sample of recent polling and has shown to be more accurate than "normal" polling. They make a recommendation based on which party has the best chance of winning.
Most of the "normal" polling services (YouGov, Survation, etc.) take polls but are not making tactical voting recommendations. People can make use of their results to make their own decision which party has the best chance of winning.
"Normal" polls may or may not be using MRP analysis (mostly they don't) so their constituency-by-constituency forecasts will have a higher margin of error (especially in constituencies where the sample size is small)
Some tactical voting sites make recommendations based on previous election results. This, in my view is simply wrong. 2019 isn't 2017, and it certainly isn't 2015 or 2010. The situation has changed, the candidates standing have changes. The party policies have changed.
Past-results sites are favoured by Labour because they are biased towards a two-party system and Labour were by far the biggest non-Tory party in the 2017 results - so in most cases the recommendation will be Labour. But this doesn't take certain things into account:
1. Labour in 2017 experienced a significant boost from Momentum - support which has died back somewhat over the last two years.
2. There was significant tactical voting in 2017 in favour of the Labour party by people who would usually vote LibDem. The Tories were denied a majority by LibDems voting Labour.
3. Many who voted Labour (whether tactically or unaligned centrist Remainers) are unhappy that Labour has not taken an unequivocal Remain position and will not lend them their vote again.
4. Moderate Tory voters (mainly Remainers, but also Leavers unhappy with Johnson's horrible Brexit plan, and also those undecided or uncaring of the Brexit outcome and just want it over) are more likely to tactically vote LibDem than they are Labour.
5. LibDems have recovered in the polls and in many places are becoming a viable option. Where in 2015 and 2017 voting LibDem would be effectively throwing away your vote, in many constituencies that is no longer the case,
All of the above mean that relying on the 2017 results to make recommendations for the 2019 General Election is simply wrong and may lead to disaster (also known as a Tory win).
The most important point is this:
While it seems counter-intuitive, Labour would probably win *more* seats by engaging with the other parties and following the tactical voting recommendations on getvoting.org than they would trying to go it alone.
While it seems counter-intuitive, Labour would probably win *more* seats by engaging with the other parties and following the tactical voting recommendations on getvoting.org than they would trying to go it alone.
Yes, it would mean standing down in seats where they think they may have a chance of beating the LibDems - but in a FPTP system, what good is getting 2nd place if the Tories win the seat?
Whether it's pride or principle that stops Labour from engaging with other parties, it's corrosive - and they need to wake up, pack it in and start thinking objectively. Labour *needs* to engage with the LibDems to stop the Tories from winning this GE.
Disclaimer: while the @BestForBritain tactical voting tool (getvoting.org) is currently the best analysis out there, it is not set in stone. The second round of analysis closer to the date of the election (after manifestos have been published) will be more definitive.
