Analysing the favourite-longshot bias: Why it exists and how you should respond to it

There are many things that are common knowledge in the world of sports betting, but few can match the tendency to overvalue outside bets and undervalue favourites. This phenomenon is called the favourite-longshot bias, and in this article, we will provide some recommendations for how you can use it to your benefit and keep your winnings intact. When all is said and done, bookmakers are businesses. It is for that very reason that each bookmaker will add a margin to all provided odds so that he can turn a profit. This is done by decreasing the odds relative to the fair expectation associated with each outcome. Let us take a two-player match as an example. In the following equation, the odds of player A being victorious are ‘a’ and the odds of player B emerging triumphant are ‘b’: Margin = [(1/a) + (1/b)] x 100 per cent. However, the figure of 100 per cent – reflecting as it does the sum of probabilities of all possible outcomes – will only be such in a fair book. A profit-seeking bookmaker will look to alter the equation to his advantage, choosing a figure higher than 100 per cent. This creates an excess, which is known in the industry as an overround, vig or juice. This is all intuitive, but it is not quite as obvious how the bookmaker loads his margins – he can do so all on player A, all on player B or spread evenly between the two.

How margin is added to the odds

In normal circumstances, the wisest choice would seem to be spreading the margin evenly between the two players. This would mean that if both were of a similar calibre, their fair odds would be 2.00. A 2.5% margin increase spread evenly between the two would see their odds drop to 1.95. For a two-player contest in which the odds of player A winning are ‘a’ and the odds of player B winning are ‘b’, the margin can be expressed by the following equation: Margin = [(1/a) + (1/b)] x 100%.It is rare, however, that two players, teams or competitors are completely evenly matched. Take a football match between Manchester United and Torquay United – this pair cannot be said to be of comparable quality, so the above does not really apply.Let us take a game which has a clear favourite and an obvious underdog. If the odds for the former are 1.2 and the odds for the latter are 6.00, an even distribution of the margin would see these change to 1.17 and 5.85. As experience bettors will be aware, this is seldom what happens.More likely odds in this scenario would be 1.19 and 5.41. As you can see, the odds have not been shortened at the same rate – rather, the underdog’s odds have decreased by a larger amount. The margin for the favourite is just one per cent compared to 11 per cent for the underdog. This is the favourite-longshot bias in action.

Examples of the favourite-longshot bias

We have plenty of proof that the favourite-longshot bias exists, including in football, tennis, horse racing and many other spots. In 1997, research conducted by Leighton Vaughan Williams and David Paton for The Economic Journal found that a strong favourite-longshot bias existed in 4,689 runners in 481 races during the 1992 UK flat racing season. Those who were available to back at a price lower than 2.00 experienced losses of seven per cent, which was far less than the 40 per cent for those with odds of over 40/1. Similar findings also came out of a paper by Michael Cain, David Peel and David Law which was published in the Scottish Journal of Political Economy three years later. The trio documented evidence of the favourite-longshot bias in football in England and Scotland during the 1991/92 campaign. The favourites (defined in this case as odds shorter than 1.66) had just two per cent losses, whereas the outsiders (available at 5.00 or higher) had 15 per cent losses. Similar results came out of Squares and Sharps, Suckers and Sharks: The Science, Psychology & Philosophy of Gambling, a book written by Joseph Buchdahl which uses European football to explore the bias.Yet another instance of the bias can be found in Pinnacle’s tennis market. Using the odds for ATP and WTA matches between 2011 to 2015, we see a clear difference in terms of the margin. Favourites with a win expectancy of 91 to 100 per cent (of which there were 2574 bets) brought returns of -0.46 per cent; contrast this with the underdogs with a win expectancy of zero to 10 per cent, which brought losses of 22.45 per cent. At each interval of 10 percentage points, the margin got bigger and bigger according to the principle that the outsider’s odds will shorten by a larger amount. The research has also revealed that bookmakers with larger margins have stronger biases, with the extra margin placed on the underdog. Novak Djokovic was expected to beat Ante Coric in the 2016 Madrid Masters, for example, with Pinnacle offering respective odds of 1.06 and 13.00. Another bookmaker, Interwetten, offered odds of 1.05 and 8.00. As you can see, there is very little difference between the two prices for Djokovic, but Coric’s price differs by 5.00.

Betting long-shots

We have now seen the favourite-longshot bias in action, but whydoes it exist and why is it so widespread? There have been several suggested answers to these two questions: some argue bookmakers are reacting to insider information, while others believe it is down to a misjudgement of low and high probabilities. There is a correlation between bookmakers with larger margins and their strength of their biases: the larger the margin, the stronger the bias. In virtually every case, the extra margin goes almost entirely to the outsider. It is thought that bettors express risk-seeking utility towards outsiders and risk aversion towards favourites, which means that the favourite-longshot bias is in effect a typical cognitive bias. Bookmakers make it their job to respond to bettors’ habits, which is why they shorten their prices. Some say this is to help manage liabilities, but others – including David McDonald from Southampton University’s Centre for Risk – insist that this action is a deliberate attempt by the bookmakers to take advantage of bettors’ biases. If this reading is accurate, bookmakers shorten those prices purely because they have the power to do so. Conversely, there is a larger demand for elasticity when it comes to favourites, so the prices across bookmakers do not vary too much.