For CON comparisons we should use the LAB/LD/GRN aggregate
Whenever new polls come out we tend to look for just one figure and that is the Labour lead over the Conservatives.I would suggest that we are being too simplistic.
The experience of the by-elections last week suggests that the next general election is going to see quite a bit more tactical voting against the Conservatives.
Let us recall that Labour which had come second at the general election in the Tiverton and Honiton seat was reduced to losing its deposit. In Wakefield on the other hand LAB went up with the Tories seeing their share down even further on what happened at the 2019 general election. Without Corbyn as Labour leader it is going to be harder warning of the “danger” of the main opposition party
We also know from academic analysis of the last few elections that a fair slice of those saying Green before the campaign end up voting Labour.
So for all these reasons I suggest that whenever we analyse a poll we compare the Tory figure against the aggregate of the Labour, Lib Dem, and Green shares.