It’s been a while! The polls haven’t been moving much, which is largely why I haven’t been updating, but I have been doing a bit more work on Electobot. The main bit of function I’ve just added is the ability to export a summary of how ofen each consituency has been won by each party, which allows you to use Excel (other spreadsheeting programs are available, I personally use LibreOffice Calc) to easily calculate each party’s target seats.
What does that look like with the polls as they are? Well, with the current UKPR Polling Average (Con 33, Lab 34, LD 9, UKIP 15, Green 5), the Conservative’s knife-edge seats look like this (numbers show how many times the specified party won the seat out of 1000 simulated elections):
Constituency | Conservative | Labour | Lib-Dem | SNP | PC | Green | BNP | UKIP | Other |
Castle Point | 489 | 155 | 0 | 0 | 0 | 0 | 0 | 27 | 329 |
Carmarthen West and Pembrokeshire South | 491 | 508 | 0 | 0 | 0 | 0 | 0 | 1 | 0 |
Colne Valley | 493 | 460 | 30 | 0 | 0 | 0 | 0 | 17 | 0 |
Torbay | 499 | 16 | 469 | 0 | 0 | 0 | 0 | 16 | 0 |
Norwich North | 500 | 465 | 0 | 0 | 0 | 0 | 0 | 35 | 0 |
Dover | 505 | 479 | 0 | 0 | 0 | 0 | 0 | 16 | 0 |
Taunton Deane | 505 | 5 | 478 | 0 | 0 | 0 | 0 | 12 | 0 |
Cheltenham | 506 | 12 | 471 | 0 | 0 | 0 | 0 | 11 | 0 |
Cornwall North | 508 | 11 | 453 | 0 | 0 | 0 | 0 | 28 | 0 |
Great Yarmouth | 509 | 471 | 0 | 0 | 0 | 0 | 0 | 20 | 0 |
If the Conservatives win more than half of those, they’re doing better than Electobot thinks they will. How about Labour?
Constituency | Conservative | Labour | Lib-Dem | SNP | PC | Green | BNP | UKIP | Other |
Bermondsey and Old Southwark | 22 | 474 | 493 | 0 | 0 | 0 | 0 | 11 | 0 |
Rossendale and Darwen | 510 | 476 | 0 | 0 | 0 | 0 | 0 | 14 | 0 |
Watford | 425 | 477 | 94 | 0 | 0 | 0 | 0 | 4 | 0 |
Dover | 505 | 479 | 0 | 0 | 0 | 0 | 0 | 16 | 0 |
Dundee East | 0 | 488 | 0 | 511 | 0 | 0 | 0 | 1 | 0 |
Stevenage | 487 | 503 | 0 | 0 | 0 | 0 | 0 | 10 | 0 |
Cannock Chase | 472 | 504 | 0 | 0 | 0 | 0 | 0 | 24 | 0 |
Warwick and Leamington | 487 | 505 | 0 | 0 | 0 | 0 | 0 | 8 | 0 |
Carmarthen West and Pembrokeshire South | 491 | 508 | 0 | 0 | 0 | 0 | 0 | 1 | 0 |
Elmet and Rothwell | 479 | 515 | 0 | 0 | 0 | 0 | 0 | 6 | 0 |
A couple of the same seats appear in both lists, indicating that Carmathen West and Dover seem to be the most uncertain Con-Lab marginals with the polls as they are right now. Also interesting to note that the Tories have a lot more LD marginals in their list than Labour do (4 to Labour’s 1, although Labour also having to contend with the SNP in Dundee East takes the shine off that a bit).
All the usual caveats apply – I’m more sure than ever from looking at this that Electobot is over-estimating UKIP’s level of support and under-estimating the SNP’s – but it’s certainly a new way of seeing where parties should be concentrating their strength.