I spent 15 minutes after lunch drawing up a couple of tables containing some improvised calculations to hopefully give an indication as to how voting may result on Thursday night.
I've used data collected by Pits n Pots on their poll (click here to have a look).
The poll has had just over 1000 votes since becoming active last Sunday so I thought now would be a good point to do a quick analysis.
Firstly, I'll show you my maths and then explain it and some of the factors making it unreliable after.
In my first table, I have simply taken the number of votes on the poll and divided it by the number of fielded candidates to give 'support per candidate'. This comes about because there are x amount of candidates per party across x amount of wards. The more votes for fewer candidates suggests a stronger amount of support for that party overall and perhaps more chance of scoring top at the polls.
As you can see, TUSAC and the Conservatives come top in the poll suggesting strong support overall with a score of 10 and 9.4 repectively.
Labour, the Independents and UKIP score the mid-points suggesting very good support.
And the Lib Dems lead the lower half of the board with a score of 2.4.
The nationalist parties all place near the bottom.
Labour, the Independents and UKIP score the mid-points suggesting very good support.
And the Lib Dems lead the lower half of the board with a score of 2.4.
The nationalist parties all place near the bottom.
My second table uses percentages and is a bit more complex.
Here, I have calculated each party's % of total candidates first.
Next, I noted each party's % of votes on the PitsnPots poll.
Then I calculated the difference between % support and % of candidates.
This gave a score comparing the support for the party and the number of seats they're contesting. Essentially the same as above but calculated on % and not number of votes.
Next, I noted each party's % of votes on the PitsnPots poll.
Then I calculated the difference between % support and % of candidates.
This gave a score comparing the support for the party and the number of seats they're contesting. Essentially the same as above but calculated on % and not number of votes.
I believe the second table is more accurate in showing support for parties overall but I'm not a statistician, a mathmetician or a trained political commentator so I have no idea which (if either) is more accurate but they make interesting reading I think!
This table almost reflects the results of the previous except a few key differences.
Most notable are the opposite ends. It appears that the Conservatives are likely to have great support for their % of seats contested. On the other hand, the Lib Dems make up 22% of candidates but have only 10% support giving a score of -12.
Most notable are the opposite ends. It appears that the Conservatives are likely to have great support for their % of seats contested. On the other hand, the Lib Dems make up 22% of candidates but have only 10% support giving a score of -12.
The nationalist parties fare slightly better but this may be because they have a low number of fielded candidates.
Overall, the Conservatives (10), TUSAC (3), Labour (3) and Independents (2) are the only parties to have more % support than % of total candidates, suggesting they are the most likely to win seats on Thursday.
But remember...
The poll is open on the web but of course website users do not reflect the whole electorate for example, senior citizens are a large section of the voting community and often believed to be the least online. The data may only relfect how internet users who have voted on the pits n pots poll would vote but 1000 votes is a good sample to look at some trends.
Also, some users may vote more than once (how undemocratic!).
Also, some users may vote more than once (how undemocratic!).
And then there's the wards issue. Who knows, the 1000 voters may only come from 5 wards across Stoke-on-Trent. It's impossible for the 1022 (to be exact) strong poll is ever going to be a fair reflection of the real result but the tables have certainly highlighted some trend and topics for discussion!
