[The final entry on this post can be accessed here]
I think most enjoy a casual prediction about the outcome of a match. Some braver individuals even stake their earnings on an opinion. Rarer still some individuals in the sports industry are paid to make predictions and preach about our sports. But how good are the professionals in predicting the outcome of a match?
One pundit who makes precise and weekly predictions about the Premier League's upcoming fixtures is Paul Merson of Sky Sports. Every Thursday or Friday he gives his verdict on what the results and score line will be for the weekend matches. Last season BBC's Mark Lawrenson's made 380 estimates, getting the score line right for 11% (approx. 42 matches), the result right for 43% (approx. 163 matches) and being wrong 46% of the time (approx. 175 matches).
I intend to follow Paul’s prediction this year and use his estimates to test how good the experts are at predicting the
outcome of football matches. Given that Paul will make 380 predictions over the course of the Premier League season (20/380 to date), we will have a rare opportunity to analyse the accuracy of a football pundit that systematically predicts.
'Paul’s Pie Chart' is below. After gameweek two and 20 matches he has predicted the correct results and score 16% of the time, the correct result (wrong score) 42% of the time and the wrong result 42% of the time.
I’ll update 'The Paul Pie Chart’ as the season progresses. After the first two weeks he has got over half of the results right, so not a bad start. Good Luck Merse!