Media Briefings


  • Published Date: March 2014

MISPRICED BETS ON TENNIS MATCHES: Evidence from Wimbledon illustrates the trader’s problem of processing information

The mispricing of bets on the Wimbledon tennis championships reveals the limitations of the human mind in processing information. Research by Dr Alasdair Brown, which is published in the March 2014 issue of the Economic Journal, analyses how betting prices on who will win and by what score change both before and during matches.

His study shows that bets are frequently mispriced while a match is in play, an outcome that he attributes to people’s inability to process large amounts of information accurately under time pressure.

Dr Brown uses the Betfair betting exchange as a laboratory for his study. He examines the evolution of prices in pre-match and in-play betting on the Wimbledon men’s singles in 2011 and 2012. Punters on Betfair can place bets on a player to win – the ‘win market’ – and also the specific score in sets by which they will win, for example, three sets to two – the ‘set market’.

This means that the win market may imply, for example, a 65% chance that Roger Federer will beat Andy Murray while the corresponding set market may imply a 70% chance of a Federer win. As we are talking about the same event – a Federer win – any discrepancy between the prices in the two markets must mean that at least one of the bets is mispriced.

Dr Brown examines mispricing between the two markets both prior to matches – when little or no new information arrives – and while they are in play – when information on the progress of the match (and the fundamental value of the bets) arrives continuously. He describes three main patterns in the data:

· Mispricing is approximately ten times higher during matches, with an average discrepancy of 5.3% between the two markets during play.

· Prices change much more frequently in the win market during matches. This suggests that bettors periodically neglect the peripheral set market when assessing information on the outcome of each point as it arrives.

· In-play price changes in the win market accurately predict subsequent price changes in the set market. This implies that information on the progress of matches is predominantly processed for win bets first and set bets second.

Put together, these results suggest that mispricing arises, at least in part, because of constraints on people’s abilities to process information. It would appear that participants in the market are able to incorporate information accurately into the prices of all of the bets that are traded only when there is a lull in play.

Dr Brown comments:

‘My results have implications beyond the world of betting. Betting and financial markets share many characteristics.

‘Just like the trading of bets, successful trading of stocks and bonds depends on the accurate forecasting of future events. Many of the same types of individuals – highly numerate and risk-taking – gravitate towards betting and financial markets.

‘If cognitive limitations can lead to mispricing in a tennis betting market – where information arrives in a simple binary form (a player wins or loses a point) – then the potential for such an effect in large financial markets – where information arrives in complex quantitative and qualitative forms – is substantial.’

Economists have long been interested in the efficiency of asset markets – that is, the extent to which asset prices reflect all available information. If markets are inefficient and assets are mispriced, this has repercussions for the value of investor portfolios, the cost of financing for governments and the amount of capital at the disposal of firms.

This new research demonstrates that people’s cognitive limitations – specifically, constraints on the abilities of traders to process information – play a prominent role in the mispricing of bets (assets).

The advantage of studying a betting market is that we know when information is arriving (matches are televised) and, by extension, when information processing constraints may matter.

In financial markets, the timing of information arrival may vary across market participants (perhaps a trader sits on a firm’s board) and may not be observed by the researcher.


Notes for editors: ‘Information Processing Constraints and Asset Mispricing’ by Alasdair Brown is published in the March 2014 issue of the Economic Journal.

Alasdair Brown is at the University of East Anglia.

For further information: contact Alasdair Brown via email:; or Romesh Vaitilingam on +44-7768-661095 (email:; Twitter: @econromesh).