— Oliver Cooper (@OliverCooper) November 1, 2014
CON leads amongst those who voted for parties on the ballot at GE10
UKIP’s Survation lead based on NON GE10 voters the most flaky of all
Why betting AGAINST UKIP might be value
Overnight I’ve been laying (betting against) a UKIP victory on Betfair at odds which put the purples at a 92-94% chance.
Bets, as we know, are assessments of value and NOT predictions and my reading is that at current levels UKIP’s chances are being rated too highly. Just look at the Survation tables above to see where the purples might have problems.
Just 56% of the Mark Reckless support is from those who voted for parties that were on the ballot at GE10. That compares with 85.2% for the CON candidate. A complication is that 42 people told the pollster that they voted UKIP last time when, as we all know, the party did not contest the seat. Clearly some of them might have lived elsewhere but my guess is that some of their memory of what they did or did not do four years ago is based on their current voting intention.
We saw at Heywood how non-voters last time included in headline figures can distort things. At that election the polls had more of LAB’s support being of non-GE10 voters than UKIP support.
The key factors in Rochester are that more of the the 2010 CON vote is going to the Tories than UKIP which is having to rely on non-GE10 voters and those who supported LAB and LD.
This by election is tighter than it looks.