iGB Affiliate 50 AprMay | Page 35

ELECTION SPECIAL “The most unusual wager on our books is for an MP hoping to hold on to his seat who has backed himself to the tune of £50 at 10,000/1 ever to become Prime Minister. He will cost William Hill £500,000 should he achieve the highest office.” Graham Sharpe, William Hill 2015 election interest is that the outcome looks too close to call, and hence the prices are very attractive to political bettors. The view that betting markets are more accurate predictor of political outcomes than opinion polls has been gaining credence of late. How does this tally with your past experiences of offering markets on political events? Ed Fulton: Polls are snapshots. Many people, including pundits and the press corps who should know better, forget this and shape a narrative around a single poll. Not all polls ask the same questions in the same way, or use the same methods to contact people. If a poll uses only calls to landlines to canvass opinion, how many 18-35 year olds will they reach? Not many. Sporting Index’s political markets were very accurate in the 2010 GE. Our seats market implied 36.8 per cent support for the Conservative Party, which was just .2 per cent off the final tally. Closer than any opinion poll. Matthew Shaddick: The last UK-wide election was the Euro elections in 2014, and our odds on the party vote shares got closer than any of the final polls, a subject on which I wrote a short blog piece. In the Scottish referendum last year, the media were constantly telling us the result was on a knife-edge; the betting markets always had ‘No’ as the reasonably strong favourite. In essence, polls are just snapshots of opinion, not forecasts, unlike betting markets. The most notable market of late was the Scottish referendum, where all markets were predicted pretty accurately by the betting markets. The general betting markets correctly predicted from the start that ‘No’ was an 85% chance to be the final outcome, that the percentage of those voting ‘No’ would be 55% and even the voter turnout was predicted to within 0.5%. Graham Sharpe: Opinion polls are a snapshot of opinion on the day they are carried out. The betting markets are predicting what will happen on a specific day. There is no useful comparison between the two in terms of forecasting accuracy. What is an accurate forecast, and at what stage do you regard the odds as a forecast? You could argue that an accurate forecast is one successfully predicting an outcome, but a bookmaker’s odds only really reflect their opinion of outcome on the day they are released, after which they reflect money gambled. How many markets are you offering on the UK election, where has most activity been concentrated, and what is the largest or most unusual trade/wager you have received to date? Ed Fulton: We allow punters to back and lay on over 500 General Election markets. Our General Election seats market (betting on the precise number of seats won by each party) has seen a phenomenal amount of activity since the markets went live in November 2014. We have also fielded huge interest in our ‘Name the Government’ markets, as well as predicted turnout and even size of smallest winning majority. Dale Tempest: Political betting markets have been very accurate in recent times and our traders consider them as the key indicator, much stronger than the polls. “Sporting Index’s political markets were very accurate in the 2010 GE. Our seats market implied 36.8% support for the Conservative Party, just .2% off the final tally. Closer than any opinion poll.” Ed Fulton, Sporting Index Matthew Shaddick: Hundreds of markets, as we have odds on all 650 UK constituencies, although the big one is probably going to come down to which of the Conservatives and Labour will win more seats. Based on what we’ve taken so far, the Tories will be big losers for us on that measure and vice-versa. We are also seeing a lot of money on the make-up of the next government, with a Labour iGB Affiliate Issue 50 APR/MAY 2015 33