Tweet
Bookies, biased punters and calculating odds
At a bookies, you usually see an original assessed probability which is continually corrected as punters stakes are placed. What happens though, when stakes are placed without due regard for realistic probabilities, due to the bias of the people betting?
For example, in England you may expect to see more people, say 60:40, to bet on England to win than to lose even though the probability may be 50:50. I think it can be shown (my maths may be dodgy) that a bookie can make a profit over the longer term of about 1% of stakes by offering odds of 55:45 in such a scenario. If such a scenario existed a person betting against England would be onto a winner as well. Of course, bookies change odds to make total probabilities around 105-110%. This may mean odds of 58:48 are offered, but even then the person who bets against is getting better odds than true probabilities would suggest.
What I would like to know is does this happen empirically. Are there occasions when bias is significant enough to make bets better than they should be?
For example, in England you may expect to see more people, say 60:40, to bet on England to win than to lose even though the probability may be 50:50. I think it can be shown (my maths may be dodgy) that a bookie can make a profit over the longer term of about 1% of stakes by offering odds of 55:45 in such a scenario. If such a scenario existed a person betting against England would be onto a winner as well. Of course, bookies change odds to make total probabilities around 105-110%. This may mean odds of 58:48 are offered, but even then the person who bets against is getting better odds than true probabilities would suggest.
What I would like to know is does this happen empirically. Are there occasions when bias is significant enough to make bets better than they should be?
Especially as the team are going to have a manager they aren't going to be able to communicate with 
Comment