Actual seats won: CON 306, LAB 257, LD 57

One thing that really annoys me is when people start suggesting that betting prices are the best guide to what is going to happen.

If this were the case then favourites would always win. They don’t. In the two TV debates during this campaign the betting markets made Nigel Farage favourite to be judged the winner in post debate polling. He wasn’t.

But a better example of the shallowness of the perception is 2010. The above panel was published on PB just as the polling stations were opening. As can be seen the markets over-stated the Tories and LDs and over-stated understated LAB.

So we had the extraordinary position that the spreads had CON with a 100 seat lead – compared with the 49 seat one that they actually achieved.

I don’t bet to provide a prediction tool for journalists who can’t be arsed. I bet to try to win money. My bets are not predictions but personal assessments of value. Are the chances of something happening better in my view than the odds being offered. Sometimes I get it right – sometimes I don’t.

Betting prices are NOT a good indicator of political outcomes

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble

For latest polling and political betting news follow @MSmithsonPB