UK elections and update on Austria

In Austria, it looks like a Conservative-Green coalition will be formed; I will continue to provide update until a new government has been formed.

In the UK we have an interesting poll; one that shows the Tories at 40%, Labour at 30%, and the LibDems at 15%. Given current trends, this poll is not unreasonable, and I wish to use it, in part because of its nice 'round' numbers, to make a number of projections.

Instead of using any of my own math, which in the past has shown mixed results in the UK, I'm going to use the following websites:

http://ukpollingreport.co.uk/
https://www.electoralcalculus.co.uk/homepage.html
http://principalfish.co.uk/electionmaps/?map=predictit
http://www.electionpolling.co.uk/swingometer/uk-parliament
https://icon.cat/util/elections

You'll notice that last time isn't a UK-projection website, instead, its a proportional representation calculator. Using the aforementioned topline numbers, I added 5% for the SNP, 1% for PC, 3% for the Greens, and 6% for Brexit, to produce the following:

253 - C
190 - L
95 - LD
31 - SNP
19 - G
6 - PC
38 - BX
18 - NI (northern ireland parties)

This can help you compare the other projections to a truly proportional result.

Speaking of which, starting with UK Polling Report, the results are as follows:

351 - C
214 - L
41 - SNP
21 - LD
3 - PC
1 - G
1 - Speaker*
18 - NI

*Note that this website, as with many others, uses the 2017 speaker results, IE bercow's seat, which is solidly tory. As such, from here on in, I will be adding a seat to the tories and removing and speaker seat a projection gives me where said seat is Bercow's seat and not Hoyle's seat.

Electoral Calculus produces the following:

375 - C
195 - L
35 - SNP
23 - LD
3 - PC
1 - G
18 - NI

Principal fish, which at the same time gets a #1 mark from me for being the easiest to use, gets a near-auto-fail for not providing a place to put in UK/GB wide polls, demanding things be split into England and Scotland/Wales/etc. Using their own nowcast projection*, I've been able to produce the following:

343 - C
213 - L
48 - SNP
23 - LD
4 - PC
1 - G
x - NI

*Their GB-wide split is 37C-25L-17LD while their England split is 39C-26L-18LD, as such, I went with 42C-31L-16LD for my England split. I also decided to use real polls for Scotland (39S/21C/19L/13LD) and Wales (29L/28C/12P/12LD/3G/15B)

Finally, Election Polling, gives us the following:

343 - C
220 - L
41 - SNP
23 - LD
3 - PC
1 - G
18 - NI


It would seem that all but Electoral Calculus give similar results; the latter being the only one to not use the uniform national swing method. Adding in some tactical voting, weighted in a way I feel is reasonable, and adding in a specific scotland projection, gets me the following:

391 - C
173 - L
45 - SNP
19 - LD
3 - PC
1 - G
18 - NI

Election Polling had the best Labour results. Note that there is a near 50 seat swing from Labour to the Conservatives between the two.

It is this difference in models which will make the election interesting to watch.