Over recent weeks, several betting houses have been pooling several polls into one, in an attempt to reduce their standard errors. Now Simon Jackman (who wrote a paper after the 2004 election called “pooling the polls”) has crunched the data for 2007, pooling all the Galaxy, Nielsen, Morgan and Newspoll polls into a single estimate. This is handy if you think that some pollsters have “house biases”, causing them to systematically favour one side or the other.*
Simon’s bottom line: the pooled poll shows a very strong likelihood of a Labor victory (much stronger than the implied probability from the betting markets). His mean estimate of the ALP 2pp voteshare is 54%, with a 95% confidence interval running from 52-56%.
OK, now here’s a challenge for the betting agencies: why don’t one of you run a betting market on the two-party preferred vote? That would give us a direct head-to-head way of comparing polls and markets.
* Assuming the polls “house biases” offset each other, that is.
“… Simon’s bottom line: the pooled poll shows a very strong likelihood of a Howard victory… ”
Whoah, don’t you mean “… very strong likelihod of a RUDD victory …”
Looks like the ALP’s “Rudd = new Howard” strategy has worked on you too!!
Thanks – corrected!
A market on the national seat count would be great as well.
Yes, wouldn’t it? If only you and I ran betting shops…
OK, OK people quit your griping!!!
Total Labor seat market now available – we’ll be keeping it pretty low key (say, odds to $1000 per bet), but if you want an interest…
https://www.portlandbet.com/index.php?cPath=3156&event_id=ALL&market_type_id=-#go_
Good luck!
Eben Peart
General Manager – PortlandBet
Pingback: Andrew Leigh » Blog Archive » Betting Markets vs Polls: It’s On