Some thoughts on an old chestnut - Lay the Draw

One of the best things about betting exchanges is that they provide you with the opportunity to lay (oppose, or bet against) an outcome. Effectively you are providing someone's potential winnings, whilst scooping up their stake if their selection is wrong!

In a 15 horse race it is easier to pick a horse that won't win than it is to pick a winner, surely?? So why not just lay an outsider in every race and after a year retire to Barbados a multi millionaire? Sounds like a plan, doesn't it? The answer, of course, is that your potential liabilities are high. If you lay a horse at 14 / 1 for a £100 stake you will need to pay out £1300 if it wins. Ouch!

With a football match there are only three potential outcomes, and the odds on offer reflect this. One of the oldest trading strategies on the exchanges is to lay the draw. This is effectively the same as backing both teams to win. In the premier league just under a third of the matches are drawn, and unsurprisingly the draw odds at the start of a match are usually between 3.5 and 5 on Betfair. Blindly laying the draw is a guaranteed loser, but if we are smart and trade the draw it can be very profitable. The idea is to trade out of the match after a goal. This a) locks in a profit (or small loss) and b) frees your money up for another trade.

There are two major potential banana skins awaiting the draw layer. The first, and by a long way the worst, is the dreaded 0-0 result. The other is that the underdog scores first and the draw price doesn't move sufficiently to yield a profit, and in many cases will actually shorten as the market expects an equaliser.

As an academic exercise I have been paper trading laying the draw in a totally blanket fashion over the last few days. My brief was to hedge out once the market had settled as soon as the first goal was scored - simple as that. If no goal was forthcoming then I'd back the draw at odds of 1.8 for the same amount as I'd staked the lay for. If the match remained 0-0 there would be a heavy loss on the trade, but should a late winner arrive it would be a scratch trade. I conducted NO pre-match analysis and did no other trading during the game - so the results you see on the screenshots are really simplistic, idiot proof, trades.

This shot shows a summary of the individual matches:

This next shot shows a summary of the 31 matches 'traded' in this way:

Some interesting observations can be drawn from looking at these 31 matches, and I'll return to this over the next couple of days, but for now I'll close with a screenshot showing the breakdown of Premier League fixtures from the last season:

For now, just look at roughly how close a sample the 31 matches traded are in terms of home and away wins, draws and 0-0 games to a whole season of English top flight football....