Sentiment in the betting market on Spanish football

Abstract

We employ a sample of over 3000 bets available on matches from the top tier of Spanish football in an examination of the efficiency of betting odds offered in the on-line betting market. Odds appear to be influenced by the relative number of fans of each club in a match, with supporters of the more popular team offered more favourable terms on their wagers. We report similar findings for a sample of games from Scotland. The results contrast with studies of American sports betting markets but are consistent with competitive behaviour by profit maximizing bookmakers in a market where bettors can choose between several operators.

Notes

¹ In the NFL, each team is quoted at odds of 10/11 to beat the spread. With equal money wagered on each side, bookmakers would earn 10% of amount won by bettors (this is termed ‘vigorish’) or 4.55% of total stakes.

² An access fee is payable.

³ To give sharper focus to the analysis, we include only home or away win bets, ignoring wagers on a match being drawn. Our principal hypothesis is concerned with the influence of fan betting. The draw has no fans.

⁴ Let decimal odds for home win, draw and away win be d_H , d_D and d_L , respectively. BOOKPROB, for a home win for example, is then (1/d_H )/(1/d_H  + 1/d_D  + 1/d_L ).

⁵ Away wins occurred in 0.197 of the matches while 0.323 were drawn. The proportions reflect greater home advantage than in, for example, the English Premier League.

⁶ As is customary in European soccer betting, and in contrast to the Las Vegas market on American sports, odds are available about three days before a game and are generally held fixed until the start of the match. There is therefore no distinction to be made between opening and closing values of BOOKPROB.

⁷ For example, for two teams with similarly poor recent form, it may be more probable that it will be the better supported- and therefore likely better resourced club- that will recover by winning its next game because reversion to mean effects would apply in its case. It is conceivable that odds setters focus on form and not on this mean reversion effect.

⁸ All references to significance relate to the 5% level.

⁹ Strumpf found that bookmakers also price discriminated by further shifting the odds against clients (making telephone bets) who had a record of regularly backing those teams.

¹⁰ Similar (small negative) returns would typically have accrued to each of these strategies in each of the seasons 2001/02 to 2004/05.