How bookies make your money
Earlier this summer my research on UK soccer gambling advertising was published (link to .pdf). With an observational study of gambling advertising over the 2014 soccer World Cup in Brazil, I found that two features predict whether a bookie will invest resources advertising a specific bet to bettors (primarily through either TV or shop window advertising).
Bookies advertise bets that combine complexity and salience. Compared to simple bet categories (e.g Germany to win), complex bet categories have a greater number of bet outcomes (e.g. Germany to win 2-0, 2-1, etc…). This allows bookies to offer alluringly high potential wins to customers, despite the fact that the true actual odds have actually become less fair. For example, although the probability of Germany winning should equal the sum of probabilities for all Germany winning scoreline bets (since the two sets of events are identical), the latter probability will be systematically higher as derived from bookies’ odds, indicating that the odds are less fair for the scoreline bets.
The more complex a bet category is, the harder bettors will find it to make normative probability judgments. This will especially be the case for salient events (cf the representativeness heuristic). Over the 2014 soccer World Cup bookies systematically advertised salient sub-events within complex bet categories (e.g. Germany to win 3-1, Thomas Muller to score the first goal, or Germany to win 3-1 and Thomas Muller to score the first goal). This suggests that bookies have learned how to nudge bettors towards alluring but bad bets.
The 2015-2016 soccer season has just started, so I took a few pictures of local bookies’ shop window advertising to see how this pattern holds up out-of-sample. Below are shop window adverts from Betfred, Ladbrokes and William Hill for the Manchester City versus West Bromwich Albion match yesterday. As you can see, the patterns hold up. The adverts are predominately for the favorite team to win (Manchester City, who coincidentally won 3-0), but the odds are made more attractive by requiring additional conditions to come off for the bet to win.
But the odds for these complex bets can actually be easily demonstrated to be less fair than a simple bet on “Manchester City to win” (see the link to my paper above).