Wednesday, 18 August 2010

Sharing super brain's analysis

My old colleague "Superbrain - the man with a computer for a mind", as they called him on the London Sun back in the 1970s when I travelled with him to London to introduce ratings to that paper's racing pages, has dabbled over the years with betting on elections and had considerable success at it too as he does on all things where he chooses to have a wager. There aren't many people in this world who have survived paying taxes on gambling winnings for nigh on 50 years and I've certainly found his opinions profitably worth taking notice of so you might find his analysis of the markets on this Australian election of interest.
The starting point is the market on individual seats as quoted by the major bookmakers - on this occasion he has ignored the market on the Betfair exchange because for some reason there has been very little action - and taken as marginal all those where both sides, after taking out the bookmaker's margin, are given at least a 10% chance of victory. In the last few days the market has moved in this fashion:
18-08-2010 HOWTHEMARKETHASMOVEDAdd up the percentage chances for the 48 seats in this survey of the marginals and, with other seats staying with the same party as now, you end up with the Coalition gaining 4.4 seats to end up with 69.4, the ALP losing five to  come down to 77 with Greens and others having the balance of 3.7 between them. The text books suggest the standard deviation on this kind of analysis is 3.43 so the extra seats the Coalition requires to win is equal to 1.11 standard deviations. Convert this to a probability and you have Labor as an 86.7% chance of staying in Government with the Coalition as a 13.3% chance of winning.
All that depends, of course, on whether you think that the markets on individual seats provide us with a better guide than the market on the likely overall winner which is measured by the Crikey Election Indicator. This morning the Indicator gives us the following picture:
18-08-2010 crikeyelectionindicatorMy own experience suggests that as election day gets closer the Indicator tends to not get the favourite in short enough. Personally I am thus taking Superbrain's guide and having a dollar with my sponsor Sportingbet on a Labor victory. The $1.28 will do me.
Post a Comment