After the great success of the US site, Pollingreport, during the Presidential election the polling commentator, blogger and regular contributor to Politicalbetting, Anthony Wells, has launched his own version for the UK.
This will provide a one-stop table and other data of all the latest polling information and should be invaluable in the run-up to the General Election. We’ve included it as a fixed link on the list on the right hand side of the page.
This promises to be a great resource for political gamblers and all who are interested in the way public opinion is moving ahead of the election.
The expert analysis of polls is absolutelty critical for all of us who risk money on elections and Anthony has played a very big part in explaining their workings.
The spread markets remained unmoved by the Populus poll this morning. It’s still LAB 340-348: CON 200-208: LD 70-74.
The debate on the way pollsters weight or not by past voting has produced this remarkable explanation from Andrew Cooper of Populus.
We keep our past vote weighting methodology under constant review – and this whole aspect of polling involves a lot of educated guesswork and assumption as well as hard data and science. But if we weighted as ICM, for example, do we would have ended up with Labour a couple of per cent higher and the Tories a couple of per cent lower. At the moment Populus, ICM, NOP and YouGov apply past vote weights (all in different ways), while MORI, Communicate Research and System 3 do not.
So Andrew is virtually saying that an ICM poll with the same data would have the Labour-Tory margin 4% higher.
This is a huge gap and could make all the difference between Blair being returned with a three-figure majority and a hung parliament.
For punters election bets might come down to whether you agree with ICM or Populus in their approach to past vote weighting. We’ll have to wait until, presumably, May 6 2005 to find out which pollster has it right.
FULL ROUND-UP OF BETTING ODDS.