At about 6pm in the evening on election day in 2005 I got a text from a Cowley Street staffer who gave upbeat reports about a number of key seats. It was promising and all the specific seats turned out to be spot on. But I made the terrible mistake of extrapolating this into an overall national picture that was much better than actually happened.

The result was that I placed what proved to be the most costly political bet of my life. This involved ‘buying’ Lib Dem seats on the spread-betting markets for several hundred pounds a point. The ‘buy’ level was at 67 seats my hope was that I would get my stake back multiplied by the number of MPs the party go in excess of 67.

Alas it was not to be. The party came in at 62 seats so my loss was the difference between that number and 67 multiplied by my stake. Ouch.

Fortunately my other bets on the election came good and I ended the night a net winner.

But the lesson here was that you should be very careful about betting on things that you desperately want to happen. Your judgement gets distorted and you look only for evidence that supports your view and ignore that which suggests otherwise.

This is one reason why betting on politics can be quite profitable – provided you have a proper sense of detachment. There are many punters about, particularly in the closing days and hours of an election campaign, who are betting for their allegiance and cannot interpret things objectively. In other words they are mugs like I was. One factor about betting is that the more mugs there are taking part in a market the better the chance of profit.

Since May 5th 2005 I have imposed a fairly strict rule on myself never to bet on the Lib Dems. This proved costly at the time of the Dunfermline by election when all the indications were, as the votes were being counted, that a sensational victory was on the cards.

So today my betting is focused almost entirely on three area – how many seats the Tories will get at the general election, the fight for the Democratic nomination in the US and the London Mayoralty. In the first I feel completely detached because I don’t really care either way. I buy and sell seat on the spread markets and most of the time make money. I should explain that with this form of betting you can close down your position whenever you want and I’ve already pocketed several thousand pounds.

One factor that helps is that generally there are more optimistic Tory gamblers about than those of other parties. Normally that can create good value in the pricing.

The Ken-Boris battle is completely absorbing and people are starting to bet with a huge passion on either side. My money was on Boris when his price was near 2/1. That’s now tightened and it has become a very hard call.

My Democratic race betting is not looking as good as it was. I thought that Hillary would do better than she is doing. I do have a bet taken out in May 2005 at 50/1 that it will be President Obama. That looks pretty good and will have given me enormous satisfaction even if he does not do it.

* Mike Smithson founded and edits PoliticalBetting.com, the UK’s leading political discussion blog. He was a founder member of the Liberal Democrats, stood for Parliament at the 1992 General Election, and has served as both a county and borough councillor. This is the second in a regular series of monthly articles from Mike.